Representation by Timothy J. Harland Representation Same Picture, Different Lens Book Three Timothy J. Harland Contents One · Four Versions Two · The View Three · The Third Case Four · The Label Five · Persistence Six · The Decision Seven · Available Eight · The Gap Nine · The Tape Ten · Enough Eleven · The Threshold Twelve · The Distinction Thirteen · The Difference Fourteen · The Aggregate Fifteen · The Baseline Sixteen · The Boundary Seventeen · The Handoff CHAPTER ONE Four Versions At 6:47 a.m., Lena had already counted four representations of the same person. The first was mounted outside Room 614. MORGAN, ELLIS — 58 The second was on the workstation screen. Ellis Morgan. Male. 58. Admitted 8/10. Observation. The third was moving quietly between two nurses at the desk. “Six-fourteen. Morgan. Fifty-eight-year-old male. Came in yesterday afternoon for dizziness and weakness. CT negative. Labs unremarkable except mild dehydration. Fluids overnight. Orthostatics improved. Ambulating independently. Likely discharge after rounds.” The fourth was behind the closed door. Lena looked at it. Jonah looked at her. “No,” she said. “I didn’t say anything.” “You were going to.” “I was going to ask whether you wanted coffee.” “You have coffee.” “That doesn’t answer the question.” Lena returned her attention to the nurses. The outgoing nurse was named Priya Desai. The incoming nurse, Megan Cole, stood beside her with one hand resting on the workstation and the other holding a folded report sheet. Neither appeared rushed. That had surprised Lena. Not because she expected healthcare workers to be frantic, but because nearly every description they had received before arriving at Bellweather Regional Medical Center had included some variation of the word busy. Busy emergency department. Busy medical floor. Busy discharge process. Busy morning handoff. At 6:49, the floor was quiet enough that Lena could hear the wheels of a medication cart somewhere beyond the corner. Priya continued. “Walked him twice overnight. No dizziness the second time. Ate about half his dinner. No nausea. Slept most of the night.” Megan checked something on the screen. “Any pain?” “Denied.” “Fall risk?” “Low now. He was moderate on admission.” “Family?” “Wife called around nine. She’ll pick him up if he’s discharged.” Megan nodded. “Anything else?” Priya thought for a moment. “No.” The exchange moved on to Room 615. Lena wrote nothing. Jonah did. She waited until the nurses were discussing a woman recovering from pneumonia. “What?” Jonah looked up. “What what?” “You wrote something.” “I frequently do.” “During six-fourteen.” “Yes.” “What?” He angled the notebook away from her. Lena stared at him. “That is childish.” “I’m preserving observer independence.” “You wrote that after I asked.” “I thought of it before.” “Convenient.” Jonah smiled. A woman in navy scrubs approached from the opposite end of the desk. “Good morning.” Dr. Elena Vasquez had met them the previous afternoon. Hospital medicine, fourteen years. She had volunteered to let them shadow morning rounds provided neither interfered with patient care and both left immediately if anyone asked them to. Lena liked her. Jonah knew this because Lena had described her as specific. “Morning,” Jonah said. Vasquez glanced toward the nurses. “Everything going okay?” “So far,” Lena said. “That sounds ominous.” “It wasn’t intended to.” “Give it time,” Jonah said. Lena looked at him. Vasquez smiled. “Rounds in ten. Morgan will probably be first. Straightforward discharge if he still looks good.” “Dizziness?” Lena asked. “Resolved.” “Cause?” “Most likely dehydration. He’d been working outside, wasn’t drinking enough, got lightheaded. Emergency department did the appropriate workup because of his age and symptoms. Nothing concerning turned up.” “Nothing?” Jonah asked. Vasquez shifted her attention to him. “Nothing that suggests he needs to remain in the hospital.” Jonah nodded. “Different answer.” A beat passed. “Yes,” Vasquez said. “It is.” Lena looked at Jonah. He wrote something else. This time she let him. At 7:11, Ellis Morgan was sitting on the edge of his bed wearing jeans and a faded green T-shirt. The hospital gown had been folded on the chair. His shoes were already tied. Lena noticed that first. Jonah noticed the plastic bag beside his feet. Dr. Vasquez entered with Megan, followed by a resident named Dr. Patel. Lena and Jonah remained near the wall, exactly where they had been told to remain. Five people entered the room. Ellis looked at all five. “Either I’m getting discharged or things got dramatically worse.” Vasquez laughed. “Hopefully the first one. How are you feeling?” “Better now that someone finally listened.” Megan stopped beside the monitor. Dr. Patel looked at Vasquez. Lena looked at Ellis. Jonah looked at Lena. Nobody said anything for perhaps half a second. Then the normal machinery of the visit resumed. “Good,” Vasquez said. “Any dizziness when you got up this morning?” “No.” “Chest pain?” “No.” “Shortness of breath?” “No.” “Headache?” “No.” “Any weakness?” Ellis considered it. “Not like yesterday.” Vasquez nodded. “Your testing has been reassuring. Your blood work looks better after fluids, your blood pressure is more stable when you stand, and you walked last night without symptoms. Assuming everything stays that way this morning, I think you can go home.” “That’s what Priya said.” Megan smiled. “She beat us to it.” “She also found my glasses.” Megan looked at the bedside table. “They were missing?” “Apparently.” “Where were they?” “In my shoe.” Jonah looked down. Lena did not. Megan laughed. “How did they get in your shoe?” “No idea.” “Were you wearing the shoe?” “Thankfully, no.” Vasquez moved to the next question. “Do you feel safe going home?” “Absolutely.” “Your wife picking you up?” “If she answers her phone.” “We’ll make sure you have a ride before discharge.” Ellis nodded. Dr. Patel began reviewing follow-up instructions. Hydration. Primary-care appointment. Return precautions. Stand slowly. Call if symptoms returned. Ellis listened. He asked whether coffee counted toward fluids. Patel said water would be better. Ellis looked disappointed but accepted the ruling. The visit lasted seven minutes. At the door, Jonah stopped. “Mr. Morgan?” Ellis looked at him. “You said you felt better now that someone finally listened.” “Yes.” “Who listened?” Vasquez remained at the door. Megan looked from Jonah to Ellis. Ellis shrugged. “The nurse last night.” “Priya?” “Yeah.” “What did she listen to?” Ellis glanced toward the others as though checking whether the question was more complicated than it sounded. “My dizziness.” Jonah waited. Ellis frowned. “No. That’s not right.” Lena’s attention sharpened. “What is right?” she asked. Ellis looked at her. “The dizziness wasn’t the problem.” Dr. Patel lowered the tablet in his hand. “What was?” Ellis rubbed one palm across his knee. “I couldn’t explain it yesterday.” Nobody interrupted. “I kept telling everyone I was dizzy because that’s what it felt like when I stood up. But that wasn’t why I came in.” Vasquez stepped back into the room. “Why did you come in?” Ellis looked toward the window. “Because I couldn’t figure out my route.” Lena said nothing. Jonah did not move. “What route?” Megan asked. “My route.” Ellis gestured vaguely. “I deliver parts. Same places every Tuesday and Thursday. Have for years. Yesterday morning I got in the van and knew where I was supposed to go, but I couldn’t figure out which stop came next.” Dr. Patel looked at Vasquez. Ellis noticed. “No, no. I knew where I was. I knew the addresses. I knew all the customers. I just…” He searched for the words. “Couldn’t put them in order.” “How long did that last?” Vasquez asked. “I don’t know. Ten minutes? Maybe fifteen.” “Did you tell the emergency department that?” “I tried.” “What did you tell them?” “That something was wrong.” “And the dizziness?” “That happened later. When I got out of the van.” Vasquez’s expression had changed. Not dramatically. Enough. “Mr. Morgan, I want to ask you a few more questions.” Ellis looked at the shoes he had tied because he thought he was leaving. “I’m not going home, am I?” “I don’t know yet.” He sighed. “Should’ve kept the gown on.” At 7:34, the discharge order had not been entered. At 7:41, Dr. Vasquez was on the phone with neurology. At 7:46, Dr. Patel was reviewing the emergency-department documentation. “I want the record,” Lena said. “Whatever they actually wrote, not what anyone remembers writing.” Jonah was already moving toward the door. “I want Priya.” “She’s off shift.” “She’s in the building. People don’t stop knowing things when they clock out.” Lena looked at him. It wasn’t a disagreement exactly — they rarely disagreed about what mattered, only about where to stand while looking at it. “The record will still be there in twenty minutes.” “So will she. But she won’t stay reachable.” Neither of them argued the point further. That was new enough that Lena noticed it and didn’t say so. At 7:52, she was reading emergency-department documentation alone on a workstation borrowed with permission from the hospital’s quality director. Jonah was somewhere in the building being someone else’s problem for a while. Lena scrolled. Chief complaint: dizziness. She continued. Patient reports acute dizziness and generalized weakness while working. Denies syncope. Denies focal neurological deficit. Alert and oriented x4. Another section. Patient states, “Something felt wrong.” Lena stopped. By the time Jonah came back, she had already read it four times. He dropped into the chair beside her, still holding two coffees he’d apparently acquired somewhere between the locker room and here. “She wasn’t there yet. Change of shift.” “So you got coffee instead.” “I got coffee while I waited. There’s a difference.” He set one cup down near Lena’s keyboard without asking, then looked at her screen. “What.” “There.” She turned the monitor slightly. He read it once. “There.” “I see it.” She read farther. No route. No sequence problem. No inability to order familiar stops. Nothing suggesting confusion beyond the phrase something felt wrong. “Maybe he didn’t tell them,” Jonah said. “Maybe.” “Maybe he did.” “Maybe.” “You hate maybe.” “No. I hate using maybe as a conclusion.” Jonah nodded. “Fair.” Megan approached the workstation. “Dr. Vasquez asked me to pull Priya back before she leaves.” “She’s changing shift,” Jonah said. “I missed her by about four minutes.” Megan looked at the screen. “You found something?” Lena moved slightly so Megan could see. “Not yet.” Megan read the note. “That looks normal.” “Yes.” Megan looked at Lena. “That sounded like a trap.” “It wasn’t.” Jonah made a small sound. Lena ignored him. “What does alert and oriented times four establish?” she asked. “Person, place, time, situation.” “Would Mr. Morgan have passed?” “From what he just described? Probably.” “So the documentation could be accurate.” “Yes.” “And incomplete.” Megan hesitated. “Yes.” Jonah spoke from behind them. “Different question.” Megan turned. “What is?” “Whether the chart was wrong and whether the chart represented what happened.” Lena looked at him. Megan looked back at the screen. “Those sound like the same question.” “No,” Lena said. Megan read the note again. “No,” she said slowly. “I guess they don’t.” Priya returned three minutes later carrying her coat. “I was told not to leave.” “You can leave,” Megan said. “We just need to ask you something about Morgan.” Priya’s expression changed immediately. “What happened?” “Nothing yet,” Megan said. “He told Dr. Vasquez he had trouble sequencing his delivery route before the dizziness started.” Priya nodded. “Yes.” Everyone looked at her. Lena spoke first. “You knew?” “He told me around two.” “Two this morning?” “Yes.” “Where is it documented?” Priya looked toward the screen. “I put it in.” Megan pulled up the nursing notes. They searched. 2:03 a.m. Patient awake, conversational. Denies dizziness while supine. Expresses concern regarding events preceding admission. Emotional support provided. Megan stared at it. Priya stared at it. “I wrote more than that.” “Where?” Priya moved closer. She navigated to another section. Nothing. Then another. Nothing. “I documented it.” “I believe you,” Megan said. Priya’s eyes narrowed. “No. I mean I remember documenting it.” Lena stepped away from the keyboard. “What did you write?” “That he said he couldn’t organize his delivery route. I asked whether he forgot where he was going. He said no. I asked whether he forgot the customers. He said no. He said he could see the whole route in his head but couldn’t tell which stop should come next.” Jonah opened his notebook. Lena didn’t. “Why did you ask those questions?” he said. Priya shrugged. “Because what he was saying didn’t sound like dizziness.” There it was. Not dramatic. Not technical. Not even particularly unusual. An observation. Lena looked at the screen again. “Where did the rest go?” Priya pulled a chair closer. “I don’t know.” At 8:26, they had an answer. Mostly. Priya had entered the information in a free-text field attached to an overnight neurological reassessment. The reassessment had been saved. The structured values had transferred into the longitudinal nursing view. The free-text comment had not. It still existed. Technically. Three clicks down, inside the individual assessment record, the paragraph was intact. Megan read it aloud. “Patient reports transient difficulty determining sequence of familiar delivery stops prior to onset of dizziness. Denies inability to identify locations or customers. States, ‘I knew all the places. I just couldn’t make the route happen.’” Jonah looked at Lena. She was already looking at the screen. Megan continued. “Patient concerned because this has never occurred previously. Encouraged patient to discuss with physician during morning rounds.” Priya folded her arms. “I knew I wrote it.” “You did,” Lena said. “Then why didn’t anyone see it?” No one answered immediately. The quality director, Carla Mendoza, had joined them during the search. She stood on the opposite side of the workstation. “It isn’t hidden,” Carla said. Lena looked at her. Carla corrected herself. “It isn’t intentionally hidden.” “Different answer,” Jonah said. Carla glanced at him. “Yes.” She pulled another chair over. “The handoff view aggregates selected fields. It can’t display everything.” “Of course not,” Lena said. Carla seemed surprised. “You agree?” “Completely.” Jonah smiled. Carla noticed. “Why do I feel like there’s a second half?” “There is,” Lena said. “Of course there is.” Lena pointed to the screen. “The aggregation isn’t the problem yet.” “Yet?” “We don’t know enough.” Jonah closed his notebook. Carla looked between them. “What do you need?” Lena considered the question. “How does the system decide what survives into the handoff view?” Carla nodded toward the workstation. “Structured fields mostly. Changes in status. Flags. Orders. Certain nursing assessments. There are rules.” “Who wrote them?” “Clinical informatics, nursing, physicians, quality. Probably others.” “Based on what?” Carla paused. “Relevance.” Jonah looked at Lena. Lena looked at Carla. “Define relevance.” Carla leaned back. “Oh, I see why they hired you.” “No,” Jonah said. “Not yet.” Carla looked at him. “What does that mean?” “Give it time.” At 9:03, neurology had evaluated Ellis. At 9:28, additional imaging had been ordered. At 9:31, Lena and Jonah were sitting in an empty family consultation room with two coffees between them. Neither belonged to Lena. Jonah opened his notebook. “You want to know what I wrote.” “No.” “Yes, you do.” “I already know.” “You don’t.” “You wrote his sentence.” “Which one?” “Better now that someone finally listened.” Jonah turned the notebook around. She was wrong. He had written: Megan asked: Anything else? Below it: Priya: No. And underneath that: Priya knew. Lena read it twice. “She wasn’t lying.” “I know.” “She answered the question she was asked.” “I know.” “The information was in the record.” “Yes.” “The system preserved it.” “Yes.” “The handoff omitted it.” “Yes.” “Because it wasn’t represented in the view being used.” “Yes.” Lena looked toward the closed door. “No individual component failed.” Jonah took a drink. “No.” “The emergency department documented what they understood.” “Yes.” “Priya documented what she understood.” “Yes.” “The software retained her documentation.” “Yes.” “The handoff accurately summarized the information it was configured to surface.” “Yes.” “Megan accurately communicated what she received.” “Yes.” “Vasquez made a reasonable decision based on the information available to her.” Jonah waited. Lena looked at him. “That bothers you.” “No.” “It bothers me.” “I know.” “Why doesn’t it bother you?” “It does.” “You just said no.” “You asked whether that bothers me.” Lena stared at him. Jonah pointed toward the notebook. “What bothers me is that everyone can be right.” Lena’s expression changed. Only slightly. Jonah continued. “And the picture can still be wrong.” Silence. Not disagreement. Not agreement either. Lena reached for the coffee nearest her. Jonah watched. She took a drink. Her face immediately tightened. “This is yours.” “Yes.” “Why didn’t you stop me?” “I wanted to see what would happen.” She pushed it back across the table. “What is in that?” “Hazelnut.” “It tastes like furniture.” “I don’t know what furniture tastes like.” “You do now.” Jonah smiled. Lena opened her notebook. At the top of a clean page, she wrote: REPRESENTATION Underneath it: Nothing missing. She paused. Crossed it out. Wrote again. Nothing lost. Another pause. She crossed that out too. Jonah watched without speaking. Finally she wrote: Everything preserved. Not everything present. She looked at the sentence. Then added: Present where? Jonah leaned back. “There she is.” “Don’t.” He picked up his coffee. Lena looked through the small window in the consultation- room door. Across the hall, a transporter was pushing an empty wheelchair toward the elevators. “What did Priya notice?” she asked. “That his dizziness wasn’t the problem.” “No.” Jonah waited. Lena tapped the pen once against the page. “She noticed that his representation of the problem was wrong.” Jonah considered that. “He said dizziness because dizziness was the part he could describe.” “Yes.” “And everyone investigated dizziness.” “Reasonably.” “Yes.” Lena looked again at the page. “The patient was the first proxy.” Jonah frowned. “For himself?” “For whatever happened to him.” He thought about it. “He experienced it.” “That doesn’t mean he could represent it accurately.” “No.” “Then the emergency department represented his representation.” “And Priya made another one.” “The software made another.” “The handoff made another.” “Vasquez received another.” Jonah looked toward Room 614, though they could not see it from where they sat. “And Ellis is still Ellis.” Lena stopped tapping the pen. “Yes.” Jonah smiled slightly. “What?” “Nothing.” “You’re doing the thing.” “What thing?” “The thing where you think you’ve found something and refuse to say it because you haven’t verified it.” “That is not a thing.” “It absolutely is.” Lena closed the notebook. “We have one case.” “Lucky guess?” She looked at him. Jonah waited. Something almost resembling amusement crossed her face. “Insufficient sample.” He laughed. She stood. “Where are you going?” “To find out whether this was an exception.” Jonah picked up his notebook. “And if it was?” “Then we learned something about an exception.” “And if it wasn’t?” Lena opened the door. “Then we have the wrong map.” Jonah followed her into the corridor. Neither noticed the discharge board update behind them. CHAPTER TWO The View By 10:14, Lena had found seven versions of Ellis Morgan. She did not count the man himself. That felt important. The emergency department had one. Nursing had another. Hospital medicine had another. Neurology had started a fifth. The discharge workflow had produced a sixth. “Do what?” Jonah asked. “Whatever this is.” Lena said, “Yes.” Jonah looked at her. “Apparently.” Carla turned back to the monitor. The quality dashboard had Ellis listed beneath: DISCHARGE DELAY — ADDITIONAL CLINICAL EVALUATION Carla pointed. “That is accurate.” “Yes,” Lena said. “The discharge was delayed because neurology ordered additional imaging.” “Yes.” “So what’s wrong with it?” “Nothing.” Carla waited. Jonah leaned toward her. “This is usually the part where waiting helps.” “I’ve noticed.” Lena opened the underlying event history. At 07:18, Ellis had been identified as a likely discharge. At 07:34, discharge preparation began. At 07:52, status changed to pending physician review. At 08:01, neurology consult requested. At 09:28, imaging ordered. At 09:36, discharge readiness automatically changed from likely to delayed. Carla looked at the sequence. “Still accurate.” “Yes.” “Then why are we looking at it?” “Because yesterday this patient would have appeared in your throughput analysis as a routine discharge delay.” “He still will.” Lena looked at her. Carla corrected herself. “He might.” “What determines the category?” “Primary delay reason.” “Who assigns it?” “Usually automatically. Sometimes case management updates it.” “Based on?” “Orders, status changes, timestamps, disposition.” Lena nodded. Jonah wrote something. Carla noticed. “I’ve started to hate that notebook.” “You’ve known it six hours.” “Long enough.” A knock sounded at the open office door. A woman in a gray cardigan stepped inside carrying a tablet. “Carla?” “Perfect timing. Lena Vale, Jonah Reed. This is Samantha Brooks, director of clinical informatics.” Samantha looked at Lena. Then Jonah. Then the notebook. “I was told you were asking why the handoff view doesn’t show everything.” “We’re asking how it decides what to show,” Lena said. Samantha relaxed slightly. “Better question.” Jonah looked at Carla. Carla pointed at him. “No.” “I didn’t say anything.” “You were going to.” Samantha sat. “The short version is that it doesn’t decide anything. People decided. Over years.” “Which people?” Lena asked. “Everyone.” Lena stopped. Samantha smiled. “I thought you’d like that.” “Define everyone.” “Clinical informatics. Nursing councils. Medical staff. Pharmacy. Quality. Risk. IT. Regulatory. Case management. Individual departments. Vendors. Sometimes state requirements. Sometimes federal requirements.” Jonah looked at the monitor. “So nobody built the view.” Samantha thought. “Lots of people built the view.” “Different answer,” he said. “Yes.” Lena opened the handoff screen again. “Show me the design logic.” Samantha pulled her chair closer. “There isn’t one document.” Lena looked at her. “There are requirements.” “Yes.” “Specifications.” “For portions.” “Governance records.” “Some.” “Change requests.” “A lot.” “Then there is a design logic.” Samantha nodded. “Distributed.” Jonah said, “Inherited.” Samantha looked at him. “That too.” She opened a configuration panel. The handoff view looked simple from the clinical side. Patient name. Age. Diagnosis. Code status. Diet. Mobility. Fall risk. Lines. Medications due. Pending tests. Discharge plan. A few recent assessments. Underneath it was not simple. Each field came from somewhere else. Some were calculated. Some were copied. Some were manually entered. Some appeared only when values crossed thresholds. Some disappeared when values normalized. Some were retained for hours. Others persisted for the whole admission. Some were role-specific. Some were specialty-specific. Some depended on what unit the patient occupied. Jonah stopped writing. “That is a lot of rules.” Samantha laughed. “This is the simple screen.” Lena looked at him. “Careful.” “I didn’t say anything.” “You were going to say architecture.” “I was going to say nightmare.” Samantha nodded. “Also valid.” Lena pointed to a field. “Neurological status.” Samantha opened it. “Structured assessment.” “Source?” “Nursing flowsheet.” “Frequency?” “Depends on unit and indication.” “What survives here?” “Latest relevant values.” “Relevant according to?” Samantha looked at her. “Here we go.” Carla sat back. “I told you.” Samantha opened another panel. A long list appeared. Alertness. Orientation. Pupil response. Speech. Strength. Sensation. Glasgow Coma Scale. Several others. Lena scrolled. “No sequencing.” “No.” “Executive function?” “Not on this view.” “Memory?” “Limited.” “Why?” “Because this isn’t a neuropsychological assessment.” “I know.” Samantha paused. “Then what are you asking?” Lena looked at the screen. “I’m trying to determine what kind of patient this view is designed to represent.” The room became quiet. Not dramatically. Enough. Samantha looked at the fields again. “A hospitalized one.” Jonah tilted his head. “That sounds obvious.” “It is.” “Which usually means it isn’t.” Carla looked at him. “You two are exhausting.” “Occupational hazard,” Jonah said. Lena ignored them. “What does the hospital need this representation to do?” Samantha answered more carefully. “Support safe care transitions. Help clinicians rapidly understand current state. Reduce omission. Prioritize high- risk information. Avoid forcing them to reconstruct the entire chart every time responsibility changes.” “Good.” “That’s good?” “Yes.” Samantha looked suspicious. Lena continued. “So completeness would make it worse.” “Absolutely.” Jonah looked at her. “There it is.” She nodded. The problem was not that the view omitted information. It had to. A useful handoff depended on omission. At 10:42, Samantha demonstrated how Ellis’s overnight note had disappeared from the default handoff. Not deleted. Not suppressed. Not filtered by anyone who knew what it said. The note lived under: NURSING ASSESSMENT → NEUROLOGICAL → ADDITIONAL COMMENTS The handoff pulled structured neurological values. Alert and oriented. Pupils equal. Speech clear. Strength equal. No dizziness while supine. All normal. The free-text comment remained attached to the original assessment. The handoff did not display free-text neurological comments by default. “Why?” Jonah asked. Samantha answered immediately. “Because nurses write everything in those fields.” “Everything?” “Everything.” Carla laughed. “She means everything.” Samantha nodded. “Patient upset about breakfast. Daughter called. Patient says room is cold. Patient wants chaplain. Patient reports chronic numbness present twenty years. Patient doesn’t like the television. Patient says left foot feels weird. Patient says the doctor reminds him of his brother.” Jonah smiled. “That last one feels operationally significant.” “No.” “Depends on the brother.” Lena looked at him. “Don’t.” Samantha continued. “If we surfaced every free-text comment during handoff, nobody would see anything. We tried variations years ago. Too much information. Clinicians stopped reading the section.” “So it was removed,” Lena said. “From the default view.” “Because comprehensive representation reduced usable representation.” “Yes.” Lena wrote: MORE COMPLETE ≠ MORE USEFUL Jonah leaned over. “Too broad.” She looked at the sentence. “Agreed.” She added: FOR A SPECIFIC DECISION. Jonah nodded. “Better.” Samantha watched them. “You’ve done this before.” Carla laughed. “Don’t ask.” A pager sounded from Samantha’s tablet. She silenced it. “Let me show you something.” She searched the change-request archive. Hundreds of entries appeared. Add anticoagulation status to handoff. Approved. Add last bowel movement. Approved for some units. Add family communication preference. Rejected. Add isolation status. Approved. Add pain goal. Modified. Add preferred pharmacy. Rejected from handoff, retained elsewhere. Add fall history. Approved with conditions. Add social barriers. Too broad. Revised. Approved for discharge planning view only. Jonah read down the list. “This is argument history.” Samantha looked at him. “What?” “The screen.” He pointed. “It’s the residue of every argument about what mattered enough to survive.” Lena stopped writing. Samantha stared at the configuration. Carla said, “That’s annoyingly accurate.” Jonah smiled. “Welcome.” Samantha looked at him. “To what?” Carla said, “Don’t encourage him.” Lena opened one rejected request. REQUEST: Display complete nursing narrative in handoff panel. The rationale was six years old. A committee had proposed expanding narrative visibility after a series of communication complaints. Pilot implemented on two units. Handoff duration increased. Clinician satisfaction decreased. Critical-item acknowledgment decreased. The change was reverted. Lena read the final note twice. Narrative content remains accessible in source documentation. Users are expected to review detailed notes when clinically indicated. “There,” she said. Jonah looked. “Expected by whom?” “Committee.” “No. I mean who knows when it’s clinically indicated?” Samantha answered. “The clinician.” “How?” “Context.” Jonah nodded. “So the system compresses the record because clinicians can’t read everything.” “Yes.” “And relies on clinicians to know when the compressed view isn’t enough.” “Yes.” “Based on signals that may only exist in the information that was compressed out.” Samantha stopped. Carla stopped. Lena looked at Jonah. He looked back. Neither spoke. Samantha said, “That’s not always true.” “No,” Lena said. Jonah nodded. “No.” Samantha looked relieved. Lena continued. “It only has to be true occasionally to matter.” Relief gone. Samantha, still beside them, looked confused. Carla sighed. “I knew eventually there would be vocabulary.” Lena considered it. “Possibly.” Jonah looked offended. “Possibly?” “We don’t know whether the dependency materially affects outcomes.” “It affected Ellis.” “One case.” “Right.” “Bounded side quest.” Jonah smiled. “There she is.” Samantha looked at Carla. “Do I want to know?” “No.” They returned to the sixth floor. Ellis was gone from Room 614. The bed had been stripped. Jonah stopped. “Discharged?” Megan was charting nearby. “No. MRI.” “Anything new?” “Neurology thinks possible transient ischemic event. They’re working it up.” “Possible?” Lena asked. “Yes.” “Any diagnosis?” “Not yet.” “Good.” Megan looked at her. “Good?” “Good that they’re not pretending certainty.” “Oh.” Megan turned back to the chart. “His wife is here.” Jonah glanced down the hall. “Where?” “Family room.” Lena looked at him. “No.” “I didn’t say anything.” “You’re going to talk to her.” “I was going to ask whether we should.” “Different.” “Barely.” Megan said, “She already asked who you are.” Jonah smiled. “See?” Lena looked at Megan. “What did you tell her?” “That you’re reviewing communication.” “Accurate.” “Approximately,” Jonah said. Lena looked at him. He raised both hands. “Sorry.” Dana Morgan was sitting alone with a paper cup she had not touched. She stood when they entered. Jonah introduced himself first. Lena noticed. That had become normal. They explained their role. Quality review. Not clinical care. Not there to make decisions. Dana nodded. “So you’re the reason he’s still here?” “No,” Lena said. Jonah looked at her. She added, “We asked about something he had already said. His physicians decided it required more evaluation.” Dana sat again. “Good.” Neither expected that. Jonah took the chair across from her. “You were worried?” “I’ve been worried since yesterday.” “About the dizziness?” Dana stared at him. “No.” Jonah glanced at Lena. Lena remained still. “What worried you?” he asked. “He called me from the van.” “What did he say?” “He asked me what day it was.” Lena’s attention shifted. “Did he know?” “Yes.” Dana shook her head. “That was what was weird. He knew it was Tuesday.” Jonah waited. “He said, ‘It’s Tuesday, right?’ I said yes. Then he asked why he was on Saginaw.” “Is Saginaw part of his route?” “Every Tuesday.” “Did he know that?” “Yes.” Dana looked frustrated. “He knew everything. That’s what I’m trying to say.” Lena said, “Keep trying.” Dana looked at her. No offense. Just surprise. “He knew where he was. He knew what day it was. He knew where he was supposed to go.” “But?” Jonah asked. Dana looked between them. “But none of it connected.” Silence. Lena felt the sentence before she understood why. Dana continued. “He kept saying, ‘I know this.’ Like he was angry at himself.” “Did you tell anyone at the hospital?” Lena asked. “Yes.” “Who?” “The woman at the desk downstairs.” “Emergency department registration?” “I don’t know. Scrubs.” “Did she document it?” Dana’s expression hardened slightly. “I don’t know.” Jonah said, “What did you tell her?” “That he wasn’t acting right.” “Those words?” “Probably.” “Anything else?” “I said he knew everything but couldn’t use it.” Lena looked at Jonah. He had stopped writing. “Why did you stop?” she asked. Dana looked confused. Jonah answered. “Because I heard it.” Lena nodded. Dana watched them. “Am I missing something?” “No,” Lena said. That answer came too quickly. She corrected herself. “We don’t know yet.” Dana almost smiled. “That’s more believable.” They found the emergency department note at 11:39. Not the physician note. Not the triage note. Registration. A short field labeled: FAMILY CONCERN The entry read: Wife reports patient not acting like himself. Nothing else. Dana’s longer explanation had become six words. Accurate. Lena hated how often that word remained applicable. Jonah read the field. “That’s not wrong.” “No.” “It’s also almost useless.” “Not useless.” “Fine. Low-resolution.” Lena nodded. “Better.” Megan stood beside them. “Family concern is a flag.” “Where does it go?” Lena asked. “Emergency department dashboard.” “After admission?” “It stays in the ED encounter.” “Does inpatient nursing see it?” “If they open the ED documentation.” “Is it in handoff?” “Not usually.” “Hospitalist admission note?” “If they incorporate it.” “Did they?” They checked. Collateral history obtained from spouse: patient experienced dizziness while working and appeared unusually anxious/confused. Jonah read it. “Confused.” Megan said, “That’s stronger.” Lena shook her head. “No.” They both looked at her. “Different.” Megan frowned. “How?” “Dana didn’t describe lack of orientation. She described preserved facts without usable sequence.” Jonah nodded slowly. “Confused collapses the shape.” Megan looked at the notes side by side. Not acting like himself. Anxious/confused. Dizziness and weakness. Difficulty sequencing familiar route. Four descriptions. Same event. Maybe. At noon, Dr. Vasquez joined them in a small conference room. Samantha returned from informatics. Carla brought lunch she claimed nobody had time to eat. Megan stayed. Five active speakers. Jonah noticed Lena notice. She noticed him noticing. Neither commented. The speaker-state problem did not become a problem. Carla opened. “Are we saying the system missed a neurological event?” Vasquez answered first. “We don’t know whether there was a neurological event.” “Fair.” “Imaging may be normal. Symptoms may not fit a TIA. Neurology is still evaluating.” Samantha said, “And the documentation did contain the relevant information.” “Eventually,” Jonah said. Samantha looked at him. “That’s not what I meant.” “I know.” Lena intervened. “The current question isn’t whether the diagnosis was missed.” Vasquez nodded. “Good.” Carla looked around. “Then what is the question?” Lena turned the monitor toward them. Four representations appeared. Dizziness. Not acting like himself. Anxious/confused. Difficulty sequencing familiar route. She said, “Are these equivalent?” “No,” Vasquez said immediately. Samantha shook her head. “No.” Megan said, “Related.” Jonah said nothing. Carla looked at him. “And you?” “I want to know which one changes what people do.” Lena looked at him. “Keep it.” He smiled. Vasquez leaned toward the screen. “Dizziness gets one workup.” “Confusion?” Lena asked. “Potentially another.” “Difficulty sequencing a familiar route?” “Definitely makes me think differently.” “Not acting like himself?” “Depends heavily on context.” “So each representation changes the decision space.” “Yes.” Samantha said, “But we can’t preserve every phrase exactly as spoken.” “No,” Lena said. “Then compression is unavoidable.” “Yes.” Carla pointed at the monitor. “So what are we fixing?” Lena looked at her. “I don’t know.” Carla leaned back. “That’s becoming less alarming.” Jonah smiled. “Welcome.” Carla ignored him. Megan studied the four lines. “What if the problem isn’t compression?” Lena turned toward her. Megan continued. “We compress constantly. Every handoff. Every note. Every call. Nobody can repeat every conversation.” Vasquez nodded. “Correct.” “So maybe the question is what gets preserved when we compress.” Samantha said, “That’s what the rules are for.” “Rules built around known categories.” Lena looked at Megan. Megan stopped. “Too broad?” “No.” Jonah leaned back. Lena added, “Keep going.” Megan pointed at Ellis’s phrases. “If the patient’s concern fits a known category, it survives better.” “Pain,” Vasquez said. “Fall risk,” Samantha added. “Medication allergy,” Carla said. “Suicidal ideation,” Megan said. “Chest pain,” Vasquez added. Jonah said, “Because somebody already decided those signals matter.” “Yes,” Lena said. Megan looked again at: I knew all the places. I just couldn’t make the route happen. “And if it doesn’t fit?” Nobody answered. Jonah finally said, “It gets translated.” Lena looked at him. “Into?” “Something the system already knows how to carry.” Samantha stared at the screen. “Dizziness.” “Confusion.” “Anxiety.” “Family concern.” Megan sat back. “Oh.” Carla looked at her. “What?” Megan ignored them. “Last week we had a woman who kept saying her medication made her feel ‘far away.’” Vasquez’s expression changed. “What happened?” “She was documented as drowsy.” “Was she?” “Not exactly.” Jonah opened the notebook. Lena looked at him. He paused. “Relevant?” “Possibly.” He wrote. Megan continued. “She was answering normally. Walking normally. Not sleepy. She just kept saying ‘far away.’” “What medication?” Vasquez asked. “New pain medication after surgery.” “What happened?” “Dose changed later because she started refusing it.” “Why did she refuse it?” “Because it made her feel far away.” Samantha said, “Was that in the chart?” “Probably.” Lena looked at her. Samantha corrected herself. “We can find out.” Carla sighed. “We’re doing this now.” Jonah looked at Lena. She looked back. “Relevance checkpoint,” she said. He nodded. They found the case in six minutes. The phrase was there. Once. Buried in a nursing note. Later notes called it: sedation then: intolerance then: patient preference The medication was changed. The outcome was fine. No harm. No complaint. No incident. Carla looked relieved. “So nothing happened.” Lena said, “Something happened.” Carla looked at her. “Clinically?” “I don’t know.” “Operationally?” “Yes.” “How?” Lena put the sequence on screen. FAR AWAY → DROWSY → SEDATION → INTOLERANCE → PATIENT PREFERENCE Jonah stared. “That drifted.” Samantha frowned. “The meanings are related.” “Are they?” Jonah asked. Vasquez answered. “Not necessarily.” Megan nodded. “‘Far away’ could mean sedation. Could mean dissociation. Could mean dizziness. Could mean cognitive fog. Could mean something else.” Carla said, “But the medication still got changed.” “Yes,” Lena said. “So the outcome was correct.” Lena looked at her. “Maybe.” Carla laughed once. “Of course.” Lena drew a line beneath the sequence. Then another. Above the first she wrote: EVENT Below it: REPRESENTATIONS OVER TIME Jonah watched. She added: Does representation drift while the underlying event remains unresolved? Samantha said, “That’s hard to measure.” “Yes.” “Possibly impossible at scale.” “Yes.” Carla looked suspicious. “You’re agreeing a lot today.” Lena ignored her. Jonah said, “We don’t need scale yet.” Lena looked at him. “Correct.” Megan said, “One floor?” “Smaller,” Lena said. “One type of event?” “Maybe.” Vasquez looked at the examples. “What are you actually looking for?” Lena considered. Not a flaw. Not a diagnosis. Not a universal pattern. She looked at Ellis’s phrase again. Then at the medication example. “We need situations where the original description changes form as it moves.” Jonah added, “And where the change could alter what somebody does.” Samantha nodded slowly. “Representation drift with decision consequence.” Lena looked at her. “Potential decision consequence.” “Right.” Carla stood. “How many?” Lena said, “Enough.” Carla stared. Jonah smiled. “She hates when other people do that.” “I do not.” “You absolutely do.” Vasquez looked between them. “How many is enough?” Lena thought. “Five independent cases to start.” Jonah shook his head. “Too many.” She looked at him. “Three.” “Two.” “Three.” “Why?” “Because two establishes recurrence. Three gives us somewhere to be wrong.” Jonah considered. “Keep it.” At 12:48, the group dispersed. Carla returned to quality. Samantha went back to informatics. Vasquez returned to rounds. Megan went to give medication. Lena and Jonah remained. On the whiteboard: MORGAN route sequence → dizziness → confusion → family concern → discharge delay POST-OP PATIENT far away → drowsy → sedation → intolerance → patient preference Two paths. Not enough. Jonah stood with his hands in his pockets. “You know what bothers me?” Lena looked at him. “That is an extremely broad question.” “Keep it.” She waited. “The system isn’t changing what happened.” “No.” “It’s changing what happened means.” “Maybe.” He looked at her. “Skepticism?” “Precision.” “Same interface.” “No.” He smiled. She looked back at the board. “Each translation may be reasonable in isolation.” “Yep.” “And each one may be more useful to the next person than the original wording.” “Yep.” “Which means preserving the original wording everywhere could actually make the system worse.” “Yep.” “And translating it may make the system better.” “Yep.” She frowned. “And still introduce error.” Jonah nodded. “There.” She looked at him. “Don’t.” He pointed at the board. “So what’s the invariant?” Lena did not answer. He waited. She looked at Ellis. Then the postoperative patient. Different patients. Different complaints. Different clinicians. Different outcomes. Different language. Same problem? Maybe. She erased EVENT. Jonah watched. “What are you doing?” “Wrong reference point.” She wrote: WHAT MUST SURVIVE TRANSLATION? Jonah stopped smiling. Lena stared at the question. “That,” he said. “Maybe.” “Come on.” “One case.” “Two.” “One confirmed, one candidate.” He sighed. “Skepticism has ruined you.” “No.” “Lucky guess, bro.” Lena slowly turned toward him. He waited. She pointed at the door. “Go find the third case.” Jonah smiled. “Welcome.” He left. Lena looked at the board again. WHAT MUST SURVIVE TRANSLATION? She added one word. WHY? Then crossed it out. CHAPTER THREE The Third Case “I don’t know. Recognition. Appreciation. Perhaps a brief acknowledgment that I successfully investigated an anomaly without adopting it as a research specialty.” “You checked a lunch tray.” “I followed the protocol.” “You invented the protocol.” “And it worked.” Lena capped the marker. “Do you want a certificate?” Jonah considered it. “No.” “Good.” “A sticker might be nice.” “Absolutely not.” He smiled. Lena pointed at the whiteboard. MORGAN route sequence → dizziness → confusion → family concern → discharge delay POST-OP PATIENT Carla was waiting in a quality-review room on the second floor. A large monitor displayed an incident report. Not a patient chart. Lena noticed that before she sat down. “What happened?” Carla pointed at the screen. “Nothing.” Jonah pulled out a chair. “You called us down here for nothing?” “Yes.” He looked at Lena. She ignored him. Carla continued. “Which is why I almost didn’t.” The report concerned a seventy-one-year-old woman named Ruth Keller who had been admitted three weeks earlier after a fall at home. No fracture. Mild bruising. Observation overnight. Physical therapy evaluation the following morning. Discharged home the next afternoon. No readmission. No formal complaint. No documented injury during hospitalization. Jonah looked at Carla. “What’s the incident?” “Her daughter called patient relations two days later.” “About?” Carla opened the complaint summary. Family expressed concern that patient was repeatedly described as noncompliant with mobility instructions despite reported difficulty understanding use of assistive device. No injury reported. Family requested clarification of discharge instructions. Lena read it twice. “Noncompliant.” “Yes.” “Where did that originate?” “That’s why I called you.” Carla opened the chart. The first note came from the emergency department. Patient reluctant to ambulate secondary to pain. The admission note said: Patient hesitant with ambulation. Requires encouragement. An overnight nursing note: Patient reminded to call for assistance prior to ambulation. Verbalizes understanding. At 3:18 a.m.: Patient found attempting to ambulate independently to bathroom. Reinforced fall precautions. At 6:42: Patient continues to require reinforcement regarding safety instructions. Physical therapy, 9:16: Patient demonstrates inconsistent carryover of walker sequencing despite repeated instruction. Jonah leaned forward. “Walker sequencing?” Carla nodded. Lena kept reading. The next nursing note: Patient noncompliant with walker use. Education provided. There it was. One word. Lena pointed. “That’s the first occurrence?” “First one I found.” “Search it.” Carla did. Seven occurrences followed. Noncompliant with walker. Noncompliant with fall precautions. Continues noncompliant behavior despite education. Family notified of noncompliance. Jonah looked at the first physical therapy note again. “Inconsistent carryover.” “Yes,” Lena said. “What does that mean?” Carla shrugged. “I know what the words mean.” “That isn’t what I asked.” Lena glanced at him. He caught it. “What?” “Nothing.” “That was a look.” “It was an observation.” Carla sighed. “I’m going to get the therapist.” Her name was Allison Grant. She arrived ten minutes later with a water bottle and the cautious expression of someone summoned by Quality without explanation. Carla introduced everyone. Allison sat. “Am I in trouble?” “No,” Carla said. Jonah said, “That sounded reassuring.” “It was intended to.” Lena looked at him. He looked innocent. Carla turned the monitor toward Allison. “Do you remember Ruth Keller?” Allison read the name. “Yes.” “What do you remember?” “Fall at home. No fracture. Sweet woman.” Lena looked up. “Anything about the walker?” Allison’s expression changed. “Oh.” Jonah noticed. “What?” “She hated that thing.” “Why?” “She couldn’t get the sequence.” Lena leaned forward. “Describe that.” Allison put down the water bottle. “It was a standard rolling walker. Nothing unusual. I showed her the pattern. Walker first, then the weaker leg, then the stronger leg.” “Could she repeat the instruction?” “Yes.” “Could she demonstrate it?” “Sometimes.” “How often?” Allison thought. “If I stood beside her and talked her through it, she was okay. Then we’d stop, talk for a minute, start again, and she’d do something different.” “Different how?” Jonah asked. “Move her feet before the walker. Pick the walker up. Push it too far ahead. Once she tried to leave it behind.” “Did she understand why she needed it?” “Absolutely.” “Did she refuse it?” “No.” Lena looked at the screen. Patient noncompliant with walker use. “Did she choose not to follow the instruction?” Allison hesitated. “I don’t know.” “Did you think she was choosing not to?” “No.” “Then why ‘inconsistent carryover’?” Allison looked at her. “Because that’s what it was.” Lena waited. Allison continued. “She could learn it in the moment. She just didn’t reliably retain the sequence.” Jonah looked at Lena. She had already turned toward the screen. “Was cognition assessed?” Carla asked. “Basic screen was normal.” “Orientation?” “Normal.” “Memory?” “Nothing obvious.” “Confusion?” “No.” Jonah said, “She knew everything.” Allison looked at him. “What?” He shook his head. “Different patient.” Lena understood immediately. Ellis knew the route. Ruth knew the instructions. Both had difficulty turning known information into ordered action. Similar shape. Different context. Maybe. She resisted the connection. One step at a time. “Did you document concern about sequencing?” Lena asked. Allison pointed at her note. “Right there.” Inconsistent carryover of walker sequencing despite repeated instruction. “Yes.” “That’s the concern.” “Did anyone contact you after that?” “No.” “Did you recommend anything?” “Continued supervision. Family training. Home health.” “Was that done?” “Family training was. Home health referral too.” Carla checked. “Yes.” “So discharge planning responded appropriately,” Lena said. “Looks like it.” Jonah pointed to the later nursing notes. “Then where did noncompliant come from?” No one knew. They found the nurse forty minutes later. Not because she had done anything wrong. Carla repeated that twice before calling. Danielle Price joined by video from home, still wearing a sweatshirt. She had worked the overnight shift and was trying to sleep before returning that evening. Carla apologized. Danielle waved it off. “What do you need?” Lena introduced herself. “We’re trying to understand a word.” Danielle looked suspicious. “What word?” “Noncompliant.” Danielle laughed. “That could take a while.” Jonah looked at Lena. She ignored him. Carla displayed Ruth’s chart. Danielle recognized the patient. “Oh, Ms. Keller.” “What do you remember?” “Nice lady. Stubborn.” Jonah asked, “Stubborn how?” “She kept getting up without the walker.” “After being told not to?” “Repeatedly.” “Did she understand the instruction?” “She said she did.” “Did you think she was refusing?” Danielle paused. “At the time? Kind of.” “Why?” “Because I’d explain it, she’d say okay, and twenty minutes later the bed alarm would go off.” “So you documented noncompliance.” “Yes.” Lena asked, “Did you read physical therapy’s note?” “I would have reviewed the recommendations.” “Do you remember the phrase ‘inconsistent carryover of walker sequencing’?” Danielle thought. “No.” Carla opened the nursing handoff from that shift. Physical therapy status appeared as: AMBULATE WITH WALKER / ASSIST X1 No narrative. No sequencing concern. Danielle nodded. “That’s what I would have seen.” Jonah looked at the screen. “What did ‘noncompliant’ mean when you wrote it?” Danielle answered quickly. “Not following the safety instructions.” “Why wasn’t she following them?” “I didn’t know.” “So the word described the behavior.” “Yes.” “Not the reason.” Danielle stopped. “I guess.” Lena looked at Jonah. He had found it. Not an answer. A distinction. She turned back to Danielle. “Would ‘unable to consistently follow walker sequence’ have changed anything for you?” “Maybe.” “How?” “I probably would have watched her differently.” “Differently?” Danielle thought. “If I think someone won’t do something, I keep reminding them.” “And if you think they can’t?” “I figure out why.” Silence. Danielle looked uncomfortable. Carla spoke immediately. “We’re not evaluating your care.” “I know.” “You acted on the information you had.” “I know.” Lena said, “And your documentation was accurate.” Danielle looked at her. “That doesn’t sound like you think it was.” “It accurately described an observed behavior.” Danielle waited. Lena looked at the word again. “It also permitted an explanation that had not been established.” Jonah turned slightly toward her. Danielle frowned. “Noncompliant means she wasn’t following instructions.” “Yes.” “That’s what she was doing.” “Yes.” “Then what’s the problem?” Lena started to answer. Jonah spoke first. “If I don’t open a door because I don’t want to, and Lena doesn’t open it because it’s locked, did we do the same thing?” Danielle looked at him. “You both didn’t open the door.” “Right.” “But for different reasons.” “Right.” Danielle looked back at the chart. “Oh.” Lena looked at Jonah. He did not look back. That was deliberate. She let it stand. After the call, Carla closed the chart. “So is this three?” “No,” Lena said. Jonah said, “Maybe.” Carla looked between them. “I hate that I know why those are different answers.” Lena stood. “We have a candidate.” “Another candidate?” “Yes.” “What would make it a case?” “Evidence that the representation changed the available interpretation in a meaningful way.” Jonah looked at the closed chart. “Danielle just told us it did.” “She told us it might have changed how she approached the patient.” “That sounds meaningful.” “It is retrospective.” “So?” “So people reconstruct reasoning.” Jonah nodded. “Fair.” Carla said, “You two realize normal people occasionally accept evidence.” “We accept evidence,” Lena said. Jonah added, “She just makes it fill out paperwork first.” Lena looked at him. “Metaphorically.” “Obviously.” She paused. “Mostly.” At 2:26, they returned to the sixth-floor conference room. The whiteboard still held the first two paths. Lena added a third. KELLER difficulty carrying over walker sequence ↓ requires reinforcement ↓ not following safety instructions ↓ noncompliant Jonah stood beside her. “That one bothers me more.” “Why?” “The others got vague.” Lena looked at the board. He was right. Ellis’s experience had been compressed into familiar categories. The postoperative patient’s description had drifted through increasingly administrative language. Ruth’s had done something else. “What changed?” Lena asked. Jonah stepped closer. “‘Inconsistent carryover’ describes what she did.” “So does ‘noncompliant.’” “Not the same way.” Lena waited. Jonah pointed to the first phrase. “This one makes me wonder what happened.” Then the last. “This one makes me think I know.” Lena looked at him. There. She uncapped the marker. Under WHAT MUST SURVIVE TRANSLATION? she wrote: UNCERTAINTY? Jonah read it. “Maybe.” She looked at him. He smiled. “Sorry.” “No, you aren’t.” “No.” She looked back at the board. “Inconsistent carryover leaves cause unresolved.” “Yes.” “Noncompliant allows cause to be inferred.” “More than allows.” Lena turned. “What?” “It feels like a reason.” “It isn’t.” “I know.” She stared at the word. NONCOMPLIANT A behavioral description. A classification. And, without anyone explicitly saying so, an explanation. She wrote: DESCRIPTION → INTERPRETATION Then beneath it: WHEN DOES A LABEL BEGIN ANSWERING A QUESTION THE DATA DIDN’T? Jonah read it slowly. “That one.” “Too broad?” “No.” “Why?” “Because it isn’t about hospitals.” Lena stopped. That was precisely why she had asked. Jonah continued. “It’s about what happens when we name something.” She looked at the three cases. Dizziness. Sedation. Noncompliance. All useful words. All defensible in context. None necessarily wrong. That was becoming irritating. “Words have to simplify,” she said. “Yes.” “Categories have to collapse variation.” “Yes.” “Otherwise categories are useless.” “Yes.” “So the problem cannot be simplification.” Jonah waited. Lena looked at him. “You’re doing that intentionally.” “What?” “Agreeing until I object to myself.” “I’ve found it efficient.” She almost smiled. Almost. “What if the distinction is reversibility?” Jonah frowned. “Meaning?” “If a representation simplifies something, can the next person still recover what was simplified?” He looked at the paths. “With Ellis, yes. Eventually.” “Three clicks.” “And a nurse who remembered.” “Yes.” “The medication patient?” “One original phrase.” “Ruth?” “Physical therapy note.” Jonah nodded slowly. “So nothing was lost.” Lena stared at the board. That handoff returned to her. Everything preserved. Not everything present. “No,” she said. “What?” “Wrong.” “What’s wrong?” “‘Nothing was lost.’” Jonah waited. “The source survived. That isn’t the same as the information surviving the transition.” He looked at the board again. “Oh.” Lena underlined SURVIVE. “Storage is not transmission.” Jonah nodded. “And transmission isn’t understanding.” She looked at him. “Keep that.” He smiled. “Gladly.” At 2:51, Megan entered carrying two cups. She handed one to Jonah. Then one to Lena. Lena looked suspiciously at hers. “What is it?” “Coffee.” “What kind?” “Black.” She looked at Jonah’s. “Hazelnut.” “Thank you.” Megan sat. “What did you find?” Jonah pointed to the third path. She read it. “Noncompliant.” “You’ve seen that before?” he asked. “Everyone has.” “What does it mean?” Megan shrugged. “Patient isn’t doing what they’re supposed to.” “Why?” “Depends.” Lena turned toward her. “Does the word tell you that?” “No.” “Does it feel like it does?” Megan started to answer. Stopped. Read the word again. “A little.” Jonah looked at Lena. She nodded. Not proof. Useful. Megan noticed the question on the board. WHEN DOES A LABEL BEGIN ANSWERING A QUESTION THE DATA DIDN’T? “That’s uncomfortable.” “Why?” Lena asked. “Because we label everything.” “Necessary?” “Absolutely.” “Useful?” “Yes.” “Dangerous?” Megan looked at her. “Sometimes.” Lena nodded. “Better.” Megan pointed at the three paths. “So what happens now?” “We test whether they’re actually the same pattern,” Lena said. Jonah took a drink. “And if they aren’t?” “We separate them.” “And if they are?” “We define what they have in common.” Megan looked at Jonah. “Does she ever answer a question without creating three more?” “No.” “Yes,” Lena said. Megan waited. Lena drank her coffee. Jonah laughed. At 3:17, Ellis Morgan returned from imaging. Jonah saw the transporter first. Lena saw Ellis looking irritated. They met him outside his room. “How’d it go?” Jonah asked. Ellis pointed at the wheelchair. “They wouldn’t let me walk.” “Probably policy.” “I know.” “You don’t sound convinced.” “I understand it. I still don’t like it.” Lena looked at him. Different. Understanding did not imply agreement. She stored that. Not now. Ellis looked at her. “What?” “Nothing.” “That didn’t look like nothing.” Jonah said, “You get used to it.” “I heard that.” “You were supposed to.” Megan approached from the nurses’ station. “Dr. Vasquez will come talk to you when the final read is back.” “Can I go home after?” “She’ll tell you.” Ellis sighed. “My wife is going to start charging parking to the hospital.” Megan smiled. “I’ll warn administration.” He stood from the wheelchair. Then paused. Not physically. His face changed. Jonah noticed first. “What?” Ellis looked toward the hallway intersection. “Where’s the elevator?” Megan pointed. “Left, then right.” “I know.” No one moved. Ellis frowned. “I know where it is.” Lena felt the room narrow. Ellis pointed left. “Down there.” “Yes,” Megan said. “Then right.” “Yes.” He looked at them. “I know that.” Jonah spoke quietly. “But?” Ellis looked angry now. Not at them. At himself. “But I had to build it.” Lena said nothing. She had stopped taking notes without deciding to. Megan stepped closer. “Build what?” “The route.” He tapped his temple. “Usually it’s just there.” Jonah looked at Lena. She was already looking at him. Not dizziness. Not confusion. Not memory. Something else. Maybe. Megan reached for the phone clipped to her waistband. “I’m calling Dr. Vasquez.” Ellis sighed. “I’m definitely not going home.” Nobody answered. He looked at Lena. “Was that the thing you were looking for?” “No,” she said. He waited. She corrected herself. “Not exactly.” Jonah looked at her. Approximally. At 3:31, Lena stood alone at the whiteboard. Jonah had gone with Megan to find Vasquez. She erased nothing. That mattered too. Under Morgan she added: I had to build it. Usually it’s just there. Then she stepped back. The three cases no longer looked the same. Good. Morgan involved sequence. The postoperative patient involved subjective experience translated into clinical terminology. Keller involved behavior translated into a label that suggested motive. Different mechanisms. Potentially different consequences. The temptation was to force them together. She knew that temptation. A pattern was satisfying. Satisfaction was not evidence. The door opened. Jonah returned. “Well?” “Vasquez is with him. Neurology’s coming back.” Lena nodded. He looked at the board. “You changed it.” “I added data.” “You didn’t make them match.” “They don’t.” He stood beside her. “So no pattern?” “I didn’t say that.” He smiled. “Of course not.” She pointed at Morgan. “Sequence became symptom.” Then the postoperative patient. “Experience became clinical category.” Then Keller. “It’s shaping it,” he said. “Yes.” Jonah looked again at the three cases. “So what do we call that?” Lena capped the marker. “We don’t.” He looked at her. “Not yet?” “Not until we know what we mean.” Jonah nodded. “Fair.” He opened his notebook. Lena pointed at him. “What are you writing?” He turned it toward her. One sentence. Don’t name it before you understand what the name will do. Lena read it twice. CHAPTER FOUR The Label By 7:18 the next morning, Ellis Morgan had acquired a diagnosis. Not a final one. A working one. Lena noticed the distinction before she noticed the words. TRANSIENT NEUROLOGICAL EVENT — ETIOLOGY UNCLEAR Jonah stood beside her at the nurses’ station. “That feels better.” “Than what?” “Confusion.” “Yes.” Lena said, “Barely.” Jonah turned toward her. She was already reading the chart. He smiled. Ellis’s MRI had been unrevealing. Neurology had not identified an acute stroke. No seizure had been observed. No major metabolic abnormality explained the episode. His symptoms had improved. Not disappeared. Improved. The new note said: Patient remains neurologically intact on standard examination. Reports intermittent difficulty with sequencing familiar tasks. Further outpatient cognitive evaluation recommended if symptoms persist. Lena read the sentence twice. Then a third time. Jonah waited. “What?” “Nothing.” “That’s never nothing.” She pointed. “Neurologically intact.” “Yes.” “Difficulty sequencing familiar tasks.” “Yes.” He looked at the screen. “Both can be true.” “I know.” “You don’t like it.” “I like it very much.” “That sounded hostile.” “It is precise.” Megan leaned over the counter. “What are you two arguing about?” “We’re not,” Lena said. Jonah answered at the same time. “Definitions.” Megan looked between them. “Same thing?” “No,” Lena said. “Yes,” Jonah said. Megan walked away. “Good morning to both of you.” At 7:42, Dr. Vasquez joined them outside Ellis’s room. “He may go home today.” Jonah looked through the glass. Ellis was sitting in the chair, eating oatmeal. “Still having trouble?” “He says less often.” “Meaning?” Vasquez considered. “He had difficulty sequencing the steps to use the television remote last night.” Lena looked at her. “Documented?” “Yes.” “Where?” Vasquez smiled. “Already checked.” “Where?” “Nursing note. Free text.” Lena looked toward the station. Vasquez followed her gaze. “It also made it into my progress note.” “Why?” “Because now I know to look for it.” Jonah said, “That sounds important.” “It is.” “Also slightly dangerous.” Vasquez looked at him. “How?” “You knew to look because someone already found the first one.” “Yes.” “And if nobody had?” Vasquez didn’t answer immediately. Lena did. “We already know that answer.” Jonah nodded. Ellis greeted them with the remote in his hand. “I have evidence.” Lena looked at it. “Of?” “That television manufacturers hate people.” Jonah sat in the visitor chair. “That’s a broad claim.” “Look at this thing.” The remote had forty-three buttons. Jonah counted. “Forty-one.” Ellis looked offended. “You counted?” “No.” Lena said, “He did.” Jonah ignored her. Ellis pointed at the cluster of buttons in the center. “I knew what I wanted. Volume. Channel. Input. Whatever. But last night I had to stop and think about what order to press things.” “You usually don’t?” Jonah asked. “No.” “Did you forget what the buttons meant?” “No.” “Forget what you wanted?” “No.” “Could you use it eventually?” “Yes.” “How?” Ellis frowned. “I broke it apart.” Lena looked up. “What does that mean?” “Instead of just doing it, I had to say it to myself.” He pointed. “Power. Input. Cable. Channel.” Jonah looked at Lena. She understood why. Different task. Same shape. Maybe. Ellis looked between them. “That face again.” “What face?” Lena asked. “The one where both of you stop talking.” Jonah said, “That’s usually a good sign.” “It isn’t.” “Fair.” Ellis put the remote down. “So what’s wrong with me?” Lena did not answer. Jonah did not either. Vasquez stepped closer. “We don’t know yet.” Ellis nodded. He looked unhappy. But not misled. Lena noticed that too. After rounds, Lena returned to the whiteboard. She added: REMOTE CONTROL known goal known controls known outcome sequence requires conscious reconstruction Under Morgan’s earlier route problem. Jonah watched her. “Same?” “Possibly.” “Same person.” “Yes.” “Same kind of failure?” “Unknown.” “Useful?” “Yes.” He nodded toward the board. “So what survives translation?” Lena didn’t answer. Instead, she drew two columns. OBSERVED INFERRED Under OBSERVED: • knew destination • knew route stops • could not order stops automatically • knew remote functions • required conscious step-by-step reconstruction Under INFERRED: • confusion • cognitive impairment • executive dysfunction • transient neurological event • anxiety Jonah read the second column. “You don’t like those.” “I didn’t say that.” “They’re all crossed out.” She had not noticed she was holding the marker over them. She lowered it. “They may be valid.” “But?” “They are not observations.” Jonah nodded. “So keep them separate.” “Yes.” He looked at the board. “Could’ve saved us a lot of trouble yesterday.” “Yesterday produced the distinction.” “Convenient.” “Accurate.” At 8:31, Carla arrived with Samantha. Carla had coffee. Samantha had a tablet. Neither looked happy. Jonah noticed. “You found something.” Carla pointed at him. “See? This is why I dislike you.” “That’s fair.” Samantha put the tablet on the table. “We searched the last six months.” Lena looked at her. “For?” “Terms that might indicate representation drift.” Jonah said, “You did that overnight?” Samantha looked tired. “Clinical informatics is a poor life choice.” Carla opened the search results. Words appeared in clusters. NONCOMPLIANT CONFUSED DROWSY ANXIOUS AGITATED REFUSING UNCOOPERATIVE POOR HISTORIAN Lena stopped at the last one. “What does that mean?” Samantha gave her a look. “You know what it means.” “I know the phrase.” Carla answered. “Patient gives incomplete, inconsistent, or unreliable history.” “According to whom?” “Usually whoever wrote it.” Jonah looked at Lena. She had gone still. “What?” “Nothing.” He waited. She pointed at the phrase. “Open one.” Samantha did. A seventy-nine-year-old man admitted for abdominal pain. The note read: History limited. Patient is a poor historian. Another: Patient unable to provide reliable medication list. Another: Daughter contacted for collateral information. Jonah read the entries. “What was wrong with him?” “Nothing cognitive documented,” Samantha said. “Why poor historian?” Carla searched. The admission note clarified: Patient uncertain of medication names and doses. States wife organizes medications at home. Jonah looked at the screen. “That’s it?” “That’s what I found.” Lena said, “He knew he didn’t know.” Carla looked at her. “What?” “He did not provide false information.” “No.” “He identified the boundary of his knowledge.” “Yes.” “And the representation became ‘poor historian.’” Samantha shifted in her chair. “That phrase is common.” “I know.” “It’s not necessarily judgmental.” “I didn’t say it was.” Jonah read the note again. “It sounds like one.” Carla looked at him. “Poor?” “Yes.” “Historian?” “That too.” Samantha said, “It’s clinical shorthand.” Lena nodded. “Useful shorthand?” “Sometimes.” “Precise?” Samantha hesitated. “Sometimes.” “Dangerous?” Carla sighed. “We’re doing this again.” They opened another chart. A woman with chronic heart failure. Poor historian regarding medication adherence. Source note: Patient reports taking medications ‘most days’ but cannot identify which pill is the diuretic. Daughter fills weekly organizer. Jonah said, “That one is different.” “How?” Lena asked. “She may actually be giving incomplete information about what she takes.” “Yes.” “Still doesn’t tell us why.” “No.” Another. Poor historian secondary to language barrier. Lena stopped. “No.” Samantha looked at her. “What?” “That label contains two different things.” “Which?” “Information quality and communication access.” Carla read the note. Interpreter had been requested. Not immediately available. The patient answered through a family member for twenty minutes before professional interpretation began. Jonah leaned closer. “Open the later note.” Samantha did. With an interpreter, the medication history was complete. Jonah looked at Lena. She didn’t need him to say it. Still, he did. “Same patient.” “Yes.” “Different historian?” “No.” Samantha looked uncomfortable. Carla crossed her arms. “So what do you want the label to say?” Lena shook her head. “That is not the question.” “What is?” “What decision does the label cause?” Carla stopped. Jonah smiled slightly. Lena ignored him. They traced the phrase through the admission. Poor historian. Then: Medication reconciliation incomplete. Then: Home medication list unreliable. Then: Pharmacy requested external fill history. No harm. No obvious failure. The system compensated. Carla looked relieved again. Jonah noticed. “You keep doing that.” “Doing what?” “Relaxing when nobody got hurt.” “That is generally considered good.” “It is.” Carla waited. He didn’t continue. Lena did. “But absence of harm does not establish absence of distortion.” Carla looked at her. “You two coordinated that.” “No,” Lena said. Jonah said, “Not consciously.” Lena looked at him. “Don’t.” Samantha opened another result. Patient refusing discharge. A fifty-four-year-old woman with pneumonia. Discharge medically appropriate. Transportation arranged. Prescriptions ready. Patient would not leave. Jonah read the case. “Why?” Carla scrolled. Patient expresses concern about returning home alone. Later: Patient continues refusing discharge despite reassurance. Later: Social work notified. Later: Patient states electricity disconnected at residence. Silence. Jonah looked at Samantha. “How long between those?” “Three hours.” Lena read the sequence. Concern. Refusal. Social work. No electricity. She pointed at the first note. “What did ‘refusing’ represent?” Carla answered. “Behavior.” “Correct.” Jonah looked at her. “That sounded ominous.” “It wasn’t.” She pointed to the final note. “What changed?” “The reason became available.” Samantha said, “And the plan changed.” “Exactly.” Social work arranged temporary placement. At 9:22, they moved to a conference room. Megan joined them. So did Dr. Vasquez. Lena chose the corner. Jonah noticed. She placed the search terms on the monitor. Carla opened. “We are not banning words.” “No,” Lena said. Samantha added, “Good.” Vasquez read the list. “Some of these are clinically useful.” “Yes,” Lena said. “Some are necessary.” “Yes.” Megan pointed at AGITATED. “That can be important.” “Absolutely,” Lena said. Jonah looked at her. “You’re agreeing again.” She ignored him. Vasquez said, “So what are we actually testing?” Lena turned the monitor. On the whiteboard: OBSERVATION INTERPRETATION LABEL DECISION She drew arrows. Observation to interpretation. Interpretation to label. Label to decision. Then another arrow back. Decision to observation. Jonah watched. “The label changes what people look for next.” “Yes.” Vasquez nodded. “That’s normal.” “Necessary,” Lena said. Samantha relaxed. Then Lena added: “But once the label exists, later observations are interpreted through it.” Samantha stopped relaxing. Megan looked at NONCOMPLIANT. “If I’m told someone is noncompliant, I probably notice the times they don’t follow instructions.” Jonah asked, “And the times they do?” Megan frowned. “I’d notice those too.” “Same amount?” She thought. “Probably not.” Carla looked at Lena. “Confirmation bias.” “Possibly.” Jonah said, “Too broad?” Lena looked at him. “Yes.” He smiled. “Good.” She pointed to the diagram. “We don’t need to explain the cognitive mechanism yet. We need to establish whether the representation changes subsequent documentation.” Samantha nodded. “That we can test.” “How?” “Compare sequences before and after the first label appears.” Lena uncapped the marker. “Now we have a measurable question.” They chose noncompliant first. Not because it was the most important. Because it was easy to find. Samantha pulled twenty anonymized cases. Lena reduced them to ten. Jonah asked why. “Time.” He nodded. “Clinical constraint acknowledged.” They reviewed the first. Patient repeatedly removed oxygen. Initial note: Patient removes nasal cannula while eating. States tubing interferes with chewing. Later: Noncompliant with oxygen therapy. Later still: Continues removing oxygen despite education. Then: Patient requests shorter tubing during meals. Respiratory therapy adjusted equipment. Removal stopped. Jonah looked at the sequence. “That one changed when the equipment changed.” “Yes.” Megan said, “So the behavior wasn’t refusal.” “Maybe partly,” Vasquez said. Megan nodded. “Right.” Lena liked that correction. The second case: Diabetic patient missed insulin doses before admission. Noncompliant with medication regimen. Later history: Prescription had not been filled for six weeks. Why? Cost. Different. Third case: Repeated missed dialysis. Transportation failure. Fourth: Dietary noncompliance. Patient knowingly ate restricted foods despite counseling. Jonah looked at Lena. “That one fits.” “Possibly.” Vasquez laughed. “Can she ever just say yes?” Jonah answered. “No.” Lena said, “Yes.” Everyone looked at her. She looked back. “The documentation supports deliberate deviation from the recommended plan.” Jonah smiled. “Growth.” “Don’t.” By the seventh case, a pattern was beginning to form. Not the one Jonah expected. Lena saw it first. “Stop.” Samantha stopped scrolling. “What?” Lena pointed at the notes. “Look at what happens after the reason appears.” They did. Case one: Noncompliant with oxygen → equipment interference identified → equipment adjusted → label disappears Case two: Medication noncompliance → cost barrier identified → social work/pharmacy assistance → later documentation becomes access barrier Case three: Dialysis noncompliance → transportation instability identified → case management involvement → later notes say transportation barrier Case four: Dietary noncompliance → deliberate preference persists → label remains Jonah leaned closer. “So the label is temporary.” “Sometimes,” Lena said. “Until a better explanation exists.” “Sometimes.” Samantha said, “That may be exactly what it’s doing.” Everyone turned toward her. “A placeholder,” she continued. “Until cause is known.” Lena thought. “Or until a more useful representation becomes available.” Jonah looked at her. “That’s different.” “Yes.” Carla frowned. “Why?” “Cause may never be known,” Lena said. “But the representation can still improve.” Megan nodded slowly. “We don’t need to know why someone can’t afford medication to know cost is the constraint.” “Correct.” Vasquez said, “And we don’t need a complete explanation of why someone prefers restricted food to document that they understand the recommendation and choose otherwise.” “Yes.” Jonah looked at Lena. She knew what he was about to say. “Proxy,” he said. She paused. “Explain.” He pointed to the labels. “They’re standing in for a more complete explanation.” Lena considered. “Sometimes.” “And useful until we get something better.” “Sometimes.” Jonah nodded. Lena continued. “A proxy cannot be mistaken for the person.” Jonah nodded. “Or for the cause.” There. Lena wrote: LABEL = PROXY FOR CURRENT UNDERSTANDING Then immediately crossed out the equals sign. Replaced it. LABEL MAY FUNCTION AS A PROXY FOR CURRENT UNDERSTANDING Jonah laughed. “Of course.” “What?” “You couldn’t let the equals sign survive.” “It exceeded the evidence.” “Obviously.” Carla looked at the sentence. “So the solution is better labels?” “No,” Lena said. “Better documentation?” “Maybe.” “More context?” “Sometimes.” Carla stared at her. “This is deeply unsatisfying.” Jonah nodded. “Welcome.” Carla pointed at him. “I knew that was coming.” He looked pleased. Lena did not. Outwardly. At 10:41, they took a break. Not because the work was finished. Because Vasquez had patients. Megan had medications. Samantha had a meeting. Carla had lost control of her calendar hours ago. Jonah followed Lena toward the elevator. “You know what I think?” “Yes.” “That’s rude.” “You’re going to say the labels aren’t the problem.” He stopped. “That was exactly what I was going to say.” “I know.” He looked at her. She pressed the elevator button. “Welcome.” Jonah stared. Lena looked straight ahead. The elevator arrived. She stepped inside. Jonah followed. “You did that on purpose.” “No.” “You absolutely did.” “Your confidence exceeds the available evidence.” He laughed hard enough that the elevator doors almost closed before Carla reached them. She stepped inside. “What happened?” “Nothing,” Lena said. Jonah was still smiling. Carla looked from one to the other. “I don’t believe either of you.” “Reasonable,” Jonah said. Back upstairs, Ellis was dressed. Again. Shoes tied. Again. This time no discharge order had been started. Vasquez stood beside him. “Neurology wants outpatient follow-up. No evidence of acute stroke. No reason to keep you here if your wife is comfortable and you’re not alone.” Ellis looked at Lena. “So I’m fine?” Vasquez answered. “No.” Ellis raised an eyebrow. “That was quick.” “You’re stable enough to go home. That is not the same as saying we know what happened.” He nodded. “That I believe.” Jonah stood near the door. “You okay with not having a name for it?” Ellis looked at him. “No.” Jonah waited. “But I’d rather have no name than the wrong one.” Lena stopped. Not visibly. Enough. Ellis noticed. “That one going in the notebook?” Jonah looked at Lena. She looked back. “No,” she said. Ellis smiled. “Why not?” “Because I heard it.” Jonah’s expression changed. Lena knew exactly what he was thinking. “Don’t.” “I didn’t say anything.” “You were going to.” Ellis looked between them. “I’m going home before this gets worse.” At 11:18, Lena returned to the whiteboard alone. She erased none of the labels. Instead, beside each one she added a small mark. NONCOMPLIANT → ? CONFUSED → ? DROWSY → ? ANXIOUS → ? REFUSING → ? POOR HISTORIAN → ? Not wrong. Not right. Provisional. She wrote beneath them: A LABEL CAN BE USEFUL WITHOUT BEING COMPLETE. Then: A USEFUL LABEL BECOMES DANGEROUS WHEN ITS PROVISIONAL STATUS DISAPPEARS. She stared. Too broad? Maybe. She left it. Jonah returned with coffee. He read the board. “That one staying?” “For now.” He handed her the cup. Black. She took it. She considered. “Closer.” He smiled. Lena capped the marker. “Don’t make that a thing.” “No promises.” “I know.” They left the room. Behind them, the labels remained on the board. Useful. Incomplete. Waiting to see what happened next. CHAPTER FIVE Persistence At 8:06 the following morning, someone had erased the whiteboard. Lena stopped in the doorway. Jonah nearly walked into her. “What?” She moved aside. He looked into the conference room. “Oh.” The board was clean. Not completely. A faint gray residue remained where yesterday’s questions had been: NONCOMPLIANT → ? CONFUSED → ? DROWSY → ? REFUSING → ? And beneath them, barely visible: A USEFUL LABEL BECOMES DANGEROUS WHEN ITS PROVISIONAL STATUS DISAPPEARS. Jonah stepped into the room. “Housekeeping?” “Probably.” “You look offended.” “I’m not.” “You stopped walking.” “The board contained work.” “We have pictures.” “I know.” “Notes.” “Yes.” “Carla probably has three copies.” “Probably.” Jonah looked at the erased board. “So nothing was lost.” Lena turned toward him. He smiled. “Sorry.” “No, you aren’t.” “No.” She put her bag on the table. “Nothing was lost from storage.” Jonah nodded. “But?” “The working state disappeared.” His smile faded. He looked at the board again. “Oh.” Lena uncapped a marker. Then stopped. Jonah noticed. “You’re not rewriting it.” “No.” “Why?” “Because I want to know what I remember.” “That seems dangerous.” “It’s bounded.” “Still dangerous.” She handed him another marker. “You too.” “Why?” “Independent reconstruction.” Jonah looked at the marker. Then at her. “You’ve made remembering a test.” “Yes.” “That is deeply on brand.” “Write.” They divided the board vertically. Lena took the left. Jonah took the right. Neither looked at the other’s side. Lena wrote: OBSERVATION → INTERPRETATION → LABEL → DECISION Then: LABEL MAY FUNCTION AS A PROXY FOR CURRENT UNDERSTANDING Then: PROVISIONAL She stopped. Something else had been there. She could reconstruct the concept. Not the wording. She left space. On the other side Jonah wrote: WHAT CAN’T WE AFFORD TO LOSE? Then: LABELS CHANGE WHAT PEOPLE LOOK FOR Then: THE NAME ISN’T THE THING Lena glanced sideways. “You looked.” “I did not.” “That last one is suspiciously concise.” “I contain multitudes.” “Unfortunately.” He continued. TEMPORARY ANSWERS HAVE A WAY OF BECOMING PERMANENT Lena stopped writing. “What?” “What?” “Say that again.” “Temporary answers have a way of becoming permanent.” “That wasn’t on the board.” “No.” She looked at him. “Why did you write it?” Jonah shrugged. “Because that’s what yesterday felt like.” Lena looked at her own word. PROVISIONAL Then at his sentence. Different representation. Same— She stopped herself. “Maybe,” she said. Jonah looked over. “I didn’t ask.” “I know.” He smiled. Carla arrived at 8:19. She carried coffee and a folder. She stopped when she saw the divided whiteboard. “No.” Jonah turned. “Good morning.” “I left you alone for thirteen minutes.” “Fourteen,” Lena said. Carla looked at her watch. “That does not improve this.” She put the folder down. “What are you doing?” “The board was erased,” Lena said. “I see that.” “We reconstructed it independently.” “Why?” Jonah pointed at the clean portions of the board. “Apparently because nothing can simply happen anymore.” Carla looked at Lena. “Accurate?” “Approximally.” Jonah turned slowly toward her. Lena saw him. “Don’t.” Carla looked between them. “I’m going to regret asking, but what does that mean?” “Optimal approximation,” Jonah said. Carla waited. “That did not help.” “It will.” “No.” Lena pointed to the folder. “What’s that?” Carla looked down. Her expression changed. “Right.” She opened it. “I brought you something.” The document was twelve pages long. PATIENT MOBILITY SAFETY IMPROVEMENT INITIATIVE Carla flipped to page seven. A table listed common contributing factors in inpatient falls. Medication. Environmental obstruction. Improper footwear. Failure to request assistance. Confusion. Noncompliance. Lena pointed to the last two. “When was this written?” “Four years ago.” “Current?” “Yes.” “Used for?” “Fall reviews. Training. Some quality reporting.” Jonah looked at Carla. “You came here because of noncompliance.” “Yes.” “Yesterday?” “Yes.” “No.” Carla frowned. “What?” “You didn’t come because of yesterday.” “I absolutely came because of yesterday.” “You found something.” Carla looked at Lena. Lena was already reading page eight. Jonah smiled. “There it is.” Carla sighed. “Fine.” She turned the page. A decision tree. PATIENT FAILED TO FOLLOW ESTABLISHED FALL- PREVENTION PROTOCOL Below it: Was patient cognitively capable of understanding instruction? YES → NONCOMPLIANCE NO → COGNITIVE CONTRIBUTOR UNKNOWN → FURTHER REVIEW Lena stared at it. “How is cognitive capability determined?” “That’s the problem.” “By whom?” “Usually the reviewer.” “Based on?” “Chart documentation.” Jonah leaned closer. “And if the chart already says noncompliant?” Carla nodded. “Exactly.” Silence. Lena traced the tree with one finger. Patient behavior generated documentation. Documentation generated a label. The label entered a review. The review used the documentation to determine which label best explained the behavior. And then— “What happens to the result?” she asked. “Falls database.” “And?” “Quarterly reporting. Training priorities. Unit-level interventions.” “Does it return to the chart?” “Not directly.” “Does it affect future practice?” “Yes.” “How?” “If a unit has a pattern of noncompliance-related falls, education may focus on patient adherence.” Jonah sat back. Samantha arrived at 8:37. Carla had called her without asking. She read the decision tree. “Oh.” Jonah looked at her. “That seems to be happening a lot.” “It’s a useful syllable.” Lena asked, “Can you trace the field?” “Which field?” “Noncompliance. From chart to falls database.” Samantha sat. “Probably.” “Probably?” “There may be transformations.” “What transformations?” “Mapping. Normalization. Category consolidation.” Lena stared at her. Samantha sighed. “Yes, I hear it.” “Good.” Jonah looked between them. “I don’t.” Samantha pointed at the decision tree. “The word in the chart may not be the same field in the reporting database.” “But it means the same thing?” “Hopefully.” Lena said, “Show us.” The first transformation was simple. Clinical review: NONCOMPLIANCE became: PATIENT ADHERENCE CONTRIBUTOR Jonah pointed. “Different words.” “Same reporting category,” Samantha said. Lena asked, “Why change it?” “Standardization. Different units use different terms.” The second transformation combined several categories. REFUSAL NONCOMPLIANCE FAILURE TO FOLLOW INSTRUCTION became: ADHERENCE / BEHAVIORAL FACTOR Jonah frowned. “Those aren’t the same.” “No,” Samantha said. “But for aggregate reporting, they’re considered similar.” “Why?” “Otherwise the categories become too small to analyze.” Lena nodded. Necessary compression. Again. Samantha continued. At the enterprise level: ADHERENCE / BEHAVIORAL FACTOR joined: COGNITIVE FACTOR COMMUNICATION FACTOR OTHER PATIENT-SPECIFIC FACTOR under: PATIENT FACTOR Jonah stared. “Now they’re definitely not the same.” “They aren’t supposed to be.” He looked at her. Samantha pointed to the reporting hierarchy. “This level isn’t asking why an individual patient fell. It’s asking how many falls had staff, environmental, equipment, medication, or patient contributors.” Lena nodded. “Different decision.” “Yes.” “Different required resolution.” “Yes.” Jonah looked at the hierarchy. “So each level is accurate for what it’s trying to do.” “Approximally,” Lena said. Jonah looked at her again. She ignored him. Samantha smiled. “I’m starting to like that word.” “Don’t encourage them,” Carla said. Lena drew the transformations on the board. RUTH KELLER ↓ NONCOMPLIANT WITH WALKER ↓ PATIENT ADHERENCE CONTRIBUTOR ↓ ADHERENCE / BEHAVIORAL FACTOR ↓ PATIENT FACTOR Jonah stared at it. “By the bottom, Ruth is gone.” Lena looked at him. “Yes.” Carla said, “She’s supposed to be.” Neither answered. Carla continued. “That’s aggregate data. We cannot retain individual narratives in enterprise reporting.” “I know,” Lena said. “Then what’s wrong?” “Nothing.” Carla waited. Lena pointed at the hierarchy. “Yet.” Carla closed her eyes. Jonah smiled. “She really hates that.” “I know.” They pulled the previous quarter’s fall report. PATIENT FACTORS: 31% Under recommendations: Continue patient education regarding fall-prevention protocols. Reinforce adherence to call-light and assistive- device expectations. Lena read it once. Then again. “Why education?” Carla answered. “Because adherence was a major contributor.” “According to which level?” Carla stopped. Samantha opened the underlying data. Twenty-six incidents had been classified under patient factors. Nine involved adherence or behavioral factors. Seven cognitive. Four communication. Six other. Jonah looked at the recommendation. “All twenty-six became education?” “Not exactly,” Carla said. “There are other recommendations.” “But this one came from the category.” “Yes.” Lena asked, “What did the nine adherence cases contain?” Carla knew where this was going. “You want all nine.” “Yes.” “That’s going to take time.” “How much?” “An hour. Maybe two.” Lena looked at Jonah. He looked at the clock. “Bounded?” “Quarter only.” “Relevance?” “Whether the aggregate recommendation remains supported by the source cases.” He nodded. “Clinical constraint acknowledged.” Carla stared at him. “What does that mean?” Jonah smiled. “It means she didn’t ask for four years.” Carla considered that. “Fine.” They reviewed nine. Three were straightforward. Patients understood instructions and deliberately chose not to follow them. Two involved cognitive impairment that had not yet been recognized at the time of the incident. One involved a patient who could not reach the call light. One involved language access. One involved urgency related to diarrhea. The last involved Ruth Keller. Jonah counted. “Three.” Carla looked at him. “Three what?” “Three where education about adherence clearly matches the source problem.” Lena corrected him. “Three where the available documentation supports that interpretation.” “Right.” “Different.” “I know.” Samantha looked at the remaining six. “The reporting classification isn’t necessarily wrong.” “No,” Lena said. “Because the patient’s behavior contributed to the fall risk.” “Yes.” “But the recommendation assumes the behavior is modifiable through education.” Lena looked at her. “Yes.” Carla sat back. “Oh.” Jonah pointed at her. “That one.” She ignored him. Lena returned to the whiteboard. Under the hierarchy she wrote: REPRESENTATION 1 → supports decision A REPRESENTATION 2 → supports decision B REPRESENTATION 3 → supports decision C Then: WHAT HAPPENS WHEN REPRESENTATION 3 IS USED TO MAKE DECISION A? Samantha stood. “That.” Carla looked at her. “What?” “That may be the problem.” Lena waited. Samantha pointed at the hierarchy. “Each abstraction level is valid for its intended use. But once the data exists, people reuse it.” “Reports,” Carla said. “Dashboards. Benchmarking. Staffing. Training. Policy.” Jonah added, “Questions it wasn’t built to answer.” Nobody spoke. Lena looked at the board. A representation did not have to be wrong. It only had to travel. At 10:02, they moved to the larger conference room. Megan joined them. Vasquez arrived five minutes later. Lena entered first. She chose the chair at the end nearest the whiteboard. The monitor sat directly ahead of her. Everyone else fell within the same field of view. Jonah stopped before sitting. Lena looked at him. “What?” “You picked an anchor.” “I picked a chair.” He looked at the board. Then the monitor. Then the room. “Approximally.” Lena followed his glance. Realized what she had done. Looked back at him. “Don’t.” He sat. “I didn’t say anything.” “You were going to.” Megan whispered to Carla, “What just happened?” Carla whispered back, “I stopped asking.” Samantha presented the fall-report hierarchy. Vasquez understood immediately. “You’re mixing levels of abstraction.” Lena looked at her. “Possibly.” Vasquez smiled. “Of course.” Carla said, “The enterprise category is valid.” “Yes.” “The unit-level classification is valid.” “Potentially.” “The individual documentation is valid.” “Potentially.” Carla looked frustrated. “Then where exactly is the failure?” Jonah looked at Lena. She waited. He understood. This one was his. He pointed to the quarterly recommendation. “What question was this report built to answer?” Carla looked at it. “What contributes to inpatient falls.” “And what question did the recommendation answer?” “How do we reduce them.” “Same question?” Carla started to answer. Stopped. Vasquez leaned back. “No.” Jonah nodded. “The first is descriptive.” Samantha added, “The second is intervention.” Lena watched the room. Nobody needed her yet. Good. Megan said, “Knowing something contributed doesn’t tell you how to change it.” “Exactly,” Vasquez said. Carla looked at the nine cases. “So we had valid data.” “Yes,” Lena said. “A valid category.” “Yes.” “A valid report.” “Yes.” “And still reached a recommendation the source data doesn’t fully support.” Lena considered the wording. “Approximally.” Carla stared. Jonah laughed. “Now she’s doing it.” At 10:39, the room had stopped discussing noncompliance. That was another sign. The question had moved. Samantha pulled a readmission dashboard. “Same architecture.” Medication-related readmission. Access. Jonah pointed around the room. “We’re about to follow every place this could happen.” Samantha closed the readmission dashboard. Vasquez smiled. “He’s right.” Lena looked at the board. Falls. Readmissions. Infections. Pain. Satisfaction. Five branches already. Diminishing returns. She erased four. Left falls. Carla watched. “That physically hurt me.” “Why?” “You just erased potentially useful investigations.” At 11:06, the working group had a bounded question. Not: Are labels bad? Not: Is aggregation dangerous? Not: Does healthcare oversimplify patients? Lena rejected all three. Instead: CAN A REPRESENTATION REMAIN VALID WHILE BECOMING UNSUITABLE FOR A NEW DECISION? Vasquez read it. “Yes.” Lena looked at her. “That was not a request for an answer.” “I know. But yes.” Jonah smiled. Lena ignored both of them. Samantha photographed the board. Carla said, “If the answer is obviously yes, why are we investigating it?” Jonah looked at Lena. She nodded. He answered. “Because we don’t know when people notice the decision changed.” Carla went still. Lena added beneath the question: WHERE IS THE DECISION BOUNDARY? Samantha stared at it. “That’s measurable.” “How?” Megan asked. “Trace where data changes purpose.” Lena nodded. “Not where it changes format.” Samantha looked at her. “Right.” Jonah added, “The words can stay exactly the same and the purpose can change.” Lena turned toward him. “Yes.” He smiled. She did not tell him to stop. At 11:34, Carla received an email. She read it. Then read it again. Jonah noticed. “Problem?” “I don’t know.” Lena looked over. “Relevant?” “Possibly.” Carla turned the laptop. A department administrator had requested the current fall- risk dashboard for a staffing review. Samantha frowned. “Why?” “Budget planning.” Lena looked at the email. “What does the dashboard measure?” “Fall rates. Contributing factors. Interventions.” “What staffing decision?” Carla read further. “They’re reviewing whether overnight sitter utilization can be reduced on two units.” “You can’t just walk into Finance.” Lena stopped at the door. “Why?” “Because—” Carla paused. Jonah waited. Carla sighed. “I don’t actually have a good answer.” “Good.” Jonah stood. Carla looked at him. “Where are you going?” “With her.” “I assumed that. Why?” He picked up his notebook. “Different decision.” Lena opened the door. Jonah followed. Carla called after them. “Do not interrogate Finance.” Lena stopped. Turned. “What would you prefer?” Carla thought. “Ask questions like normal people.” Lena looked at Jonah. He looked at her. Neither spoke. Carla closed her eyes. “Just go.” The elevator doors closed. Jonah leaned against the wall. “You know what this means.” “No.” “We found the next layer.” “Possibly.” “Same data.” She looked toward the administrative corridor. “Reuse.” Jonah stopped. The word was simple. Almost disappointingly so. He smiled. “Keep it?” Lena continued walking. “For now.” Behind them, the elevator doors closed. Upstairs, the whiteboard still held the question: CAN A REPRESENTATION REMAIN VALID WHILE BECOMING UNSUITABLE FOR A NEW DECISION? For the first time, they were about to follow the representation somewhere the patient never went. CHAPTER SIX The Decision Finance was quieter than Lena expected. Not silent. Quiet in the way spaces became quiet when most conversations happened through spreadsheets. Jonah noticed her looking. “What?” “Nothing.” “You expected something.” “No.” “That was fast.” Lena kept walking. Carla had called ahead despite saying she wouldn’t. The department administrator was waiting for them in a glass-walled office overlooking the parking structure. Daniel Cho. Director of Resource Planning. He stood when they entered. “Carla said you had questions about the fall dashboard.” “Yes,” Lena said. Daniel looked at Jonah. “And you are?” “Jonah Reed.” A beat. Daniel waited. Jonah waited back. Lena said, “He’s with me.” Daniel nodded. “That somehow created more questions.” “It usually does,” Jonah said. They sat. Daniel turned his monitor. “Which part of the dashboard?” Lena remained standing. “The part you intend to use.” Daniel looked at her. “All of it?” “For what decision?” “Sitter utilization.” “Specifically?” Daniel leaned back. “You really do start there.” “Where?” “With the decision.” “Yes.” Jonah took the chair nearest the door. Daniel glanced at him. Jonah shrugged. “She does.” Daniel opened a staffing model. Two inpatient units. Four quarters. Fall rates. Patient acuity. Sitter hours. Overtime. Agency utilization. Labor cost. Lena moved closer. “What are you trying to determine?” “Whether current overnight sitter utilization is justified.” “Define justified.” Daniel looked at her. Jonah smiled. Daniel noticed. “Is that funny?” “No.” “Yes,” Jonah said. Lena ignored him. Daniel turned back to the model. “We’ve increased sitter hours seventeen percent over eighteen months. Falls haven’t decreased proportionally.” “So the question is whether sitters work.” “No.” Lena looked at him. Daniel continued. “The question is whether the additional sitter hours are producing enough benefit to justify the additional cost.” Jonah stopped smiling. Lena noticed. “Good,” she said. Daniel raised an eyebrow. “Good?” “That is a different question.” “From what?” “Whether sitters work.” “Yes.” He waited. “So?” Lena sat. “Show me the model.” Daniel started with labor. Average sitter hours per night. Hourly cost. Overtime exposure. Agency coverage. Vacancy rates. Then fall outcomes. Falls per thousand patient-days. Falls with injury. Falls during sitter coverage. Falls without sitter coverage. Jonah watched the columns appear. “This is more complicated than the dashboard.” “Of course.” “Why?” Daniel looked at him. “Because dashboards are for people who don’t want to look at this.” Jonah nodded. “I respect the honesty.” Daniel scrolled. “We’re not trying to eliminate sitters.” “What are you trying to do?” Lena asked. “Determine whether we can reduce baseline overnight coverage and deploy sitters more selectively.” “Based on?” “Risk.” “Defined how?” Daniel opened another tab. A fall-risk score. Age. Mobility. Medication. Cognition. Prior falls. Assistance requirements. Lena recognized it. “Where does patient behavior enter?” Daniel pointed. “Here.” ADHERENCE / BEHAVIORAL FACTOR Jonah leaned forward. “There it is.” Daniel looked at him. “What?” “Nothing.” Lena said, “Not nothing.” Jonah looked at her. “Growth.” “Don’t.” Daniel waited. Lena pointed at the field. “Where does that value come from?” “The safety data.” “Which comes from?” “I assume clinical review.” “You assume?” Daniel paused. “Yes.” “Have you verified it?” “No.” “Why not?” “Because I don’t own the source system.” Lena stopped. That answer mattered. Not because it was wrong. Because it was ordinary. Daniel folded his arms. “Is the data bad?” “No.” “Then I’m missing something.” Jonah looked at Lena. She nodded once. He took it. “The data was built to answer a different question.” Daniel looked at him. “What question?” “What contributed to a fall.” “And?” “You’re using it to predict who needs a sitter.” “Yes.” “Same question?” Daniel started to answer. Stopped. Lena watched him reconstruct it. “No.” Jonah nodded. Daniel turned toward the screen. “But contribution should still have predictive value.” “Possibly,” Lena said. “There it is.” “What?” “That word.” “Which?” “Possibly.” Jonah leaned back. “You get used to it.” “No, you don’t,” Carla said from the doorway. Everyone turned. She held up a coffee. “I decided this required supervision.” “You called ahead,” Lena said. “Yes.” “That was supervision.” “It was insufficient.” She sat beside Jonah. Daniel looked at her. “Did you know about this?” “I knew enough to become uncomfortable.” “That’s reassuring.” “It wasn’t intended to be.” Lena returned to the model. “How much weight does adherence carry?” Daniel opened the scoring logic. “Not much.” “How much?” “Four percent of the composite risk score.” Carla relaxed slightly. Jonah saw it. He said nothing. Lena did too. She also said nothing. Daniel noticed both of them noticing. “What?” “Nothing,” Carla said quickly. Jonah looked at Lena. Lena continued. “If we remove it?” Daniel changed the weight to zero. The model recalculated. Almost nothing happened. Jonah looked disappointed. “So it doesn’t matter.” Daniel nodded. “Not materially.” Lena looked at the rankings. Two patients changed position. Neither crossed the staffing threshold. “Correct.” Carla exhaled. “There.” Lena looked at her. “There what?” “The questionable category isn’t affecting the decision.” “Correct.” Carla waited. Lena returned to the screen. Jonah smiled. Carla pointed at her. “I know that look.” “I don’t have a look.” “You absolutely have a look.” Daniel said, “What am I missing?” Lena pointed at the model. “The field isn’t the question.” Daniel looked at the screen. “What is?” “The threshold.” He opened the staffing threshold. Patients above a certain composite score qualified for recommended overnight sitter coverage. Below it, standard fall precautions. Jonah read the number. “Why forty-two?” Daniel answered immediately. “Because that’s where the model showed the best balance between fall reduction and staffing cost.” Lena looked at him. “When?” Daniel paused. “During implementation.” “When was that?” “Three years ago.” “Using what data?” “The prior two years.” “So the threshold represents conditions from three to five years ago.” “Yes.” “Has it been recalibrated?” Daniel hesitated. “No.” Carla looked at him. “Why not?” “The model has been performing adequately.” Lena asked, “Defined how?” Daniel looked at her. Then laughed once. “Right.” He opened another report. Overall falls had remained within target. Injury falls had decreased slightly. Sitter cost had increased. Overtime had increased. Agency utilization had increased. Patient acuity had also increased. Jonah pointed. “That seems relevant.” “It is,” Daniel said. “How much?” “That’s what we’re trying to determine.” Lena looked at the model again. “Then why are you reducing sitter utilization?” “We haven’t decided to.” “The email said staffing review.” “Yes.” “That isn’t the same thing.” “No.” Carla looked at Jonah. Jonah whispered, “She likes him.” “I heard that,” Lena said. Daniel looked between them. “I have no idea what’s happening.” “Normal,” Carla said. Daniel pulled the original implementation analysis. Three years earlier, the model had estimated that targeted sitter coverage would reduce falls enough to offset a portion of the increased labor cost. Not eliminate the cost. Justify it. Expected: 11% reduction in falls 7% reduction in falls with injury 9% increase in sitter labor cost Actual: 6% reduction in falls 8% reduction in falls with injury 17% increase in sitter labor cost Jonah read the figures. “So it failed.” Daniel shook his head. “No.” Jonah looked at him. Daniel pointed. “Injury falls improved more than expected.” “But total falls improved less.” “Yes.” “And cost almost doubled the expected increase.” “Yes.” “So?” Daniel leaned back. “So the answer depends on what return you care about.” Lena looked at him. There. Jonah looked at her. He had heard it too. Carla hadn’t. “What?” she asked. Lena pointed at the original model. “What return was this designed to optimize?” Daniel answered. “Fall reduction relative to labor cost.” “Total falls?” “Primarily.” “Not injury severity?” “It was included.” “How?” Daniel opened the old workbook. A weighted cost estimate. Emergency imaging. Extended length of stay. Treatment. Liability exposure. Jonah frowned. “You put a price on someone getting hurt?” Daniel didn’t react. “We put an estimated cost on the consequences associated with injury falls.” “Different.” “Yes.” Lena glanced at Jonah. He nodded. Correction accepted. Daniel continued. “The model needs comparable variables. That doesn’t mean the variables are equivalent.” Lena looked at him again. “You should come upstairs.” Carla nearly choked on her coffee. Daniel laughed. “I’ll pass.” Lena wrote on Daniel’s whiteboard. EXPECTED RETURN Below it: ACTUAL RETURN Then: VARIANCE Daniel stood beside her. “That’s what we’re doing.” “No.” He looked at her. “You’re comparing current cost to current outcomes.” “Yes.” “That’s not the same thing.” “Why?” “What did you expect to happen?” Daniel pointed at the original analysis. “That.” “What actually happened?” “The newer report.” “What assumptions explain the difference?” He stopped. Jonah looked at the board. “That’s the missing step.” Daniel nodded slowly. Carla said, “You mean they shouldn’t decide whether the program is worth continuing until they know why actual performance diverged from expected performance.” “Not exactly,” Lena said. Carla closed her eyes. “Of course not.” Lena continued. “They may still need to make a decision before they know why.” Daniel nodded. “Budget cycle.” “So uncertainty does not remove the decision.” “No.” “It changes the confidence.” “Yes.” Jonah looked at the three words. Expected. Actual. Variance. “So you don’t need the perfect answer.” Lena turned. “No.” “You need the best estimate after incorporating what happened.” “Yes.” Daniel said, “A revised forecast.” Lena considered. “An updated approximation.” Jonah smiled. Carla pointed at him before he could speak. “Whatever you’re about to say, don’t.” “I wasn’t going to say anything.” “Now you understand how I feel,” Lena said. Carla stared at her. Jonah laughed. At 12:17, Samantha joined by video. Her face appeared on Daniel’s monitor. “Why am I in Finance?” Carla answered. “Poor choices.” “Fair.” Lena showed her the original model. Samantha read the assumptions. Then the current results. “Patient acuity changed.” Daniel nodded. “Significantly.” “Documentation changed too.” Lena looked at her. “How?” “Two years ago we changed fall-event review templates.” Daniel sat forward. “What changed?” “More structured contributing-factor fields.” “Would that affect the risk model?” “Maybe.” Lena waited. Samantha sighed. “Possibly.” Jonah smiled. “She’s learning.” Samantha ignored him. “If the staffing model imports those fields, the distribution could shift even if patient behavior didn’t.” Daniel opened the interface specification. Carla looked at Lena. “Do you know what any of that means?” “Yes.” “That was a mistake. I shouldn’t have asked.” The interface had changed. Not dramatically. Three fields had been consolidated. One renamed. One newly required. The staffing model still accepted the data. No errors. No warnings. No broken imports. Jonah looked at the mapping. “So everything worked.” “Yes,” Samantha said. “And that might be a problem.” “Possibly.” Lena leaned toward the screen. “Does the model know the source representation changed?” “No.” “Does Finance?” Daniel answered. “No.” “Does clinical quality know Finance uses it?” Carla answered. “I didn’t.” Silence. Nobody had made a mistake. That made it more interesting. Daniel stood and walked to the whiteboard. He added beneath Lena’s words: EXPECTED RETURN ACTUAL RETURN VARIANCE Then: CHANGED CONDITIONS Lena nodded. He added: CHANGED INPUTS Samantha said through the monitor, “Good.” Then Daniel wrote: CHANGED DECISION? Jonah looked at Lena. She didn’t move. Carla noticed. “Why did everyone stop?” Jonah answered. “Because that’s the first question on the board that isn’t about whether the old model was right.” Daniel looked at it. “What is it about?” Lena answered. “Whether it is still suitable.” They spent the next forty minutes not solving the staffing problem. That was deliberate. They identified what they did not know. How much of the sitter-cost increase came from wage changes? How much from increased acuity? Had patient mix shifted? Had the relationship between sitter coverage and injury severity changed? Did the altered clinical documentation affect the model’s inputs? Were there unit-specific differences hidden by the hospital- wide aggregate? Carla looked at the growing list. “This is getting bigger.” “Yes,” Lena said. “That usually worries me.” “It should.” Jonah looked at her. “Too broad?” “Not yet.” Carla sighed. “How do you know?” Lena pointed at the questions. “Each one can change the decision.” Daniel nodded. “Material variables.” “Yes.” Jonah pointed to another question he had written: Does overnight sitter presence reduce total falls or primarily injury severity? “This one?” Daniel considered. “Yes.” “Why?” “If the return is primarily reduced severity, then total fall count is the wrong primary measure for evaluating the program.” Jonah nodded. “And if it doesn’t?” “Then the original measure may still be useful.” Lena looked at the board. They were not collecting everything. They were collecting what could change the decision. Different boundary. Same discipline. At 1:08, Daniel ordered lunch. Carla looked surprised. “You’re keeping them?” Daniel looked at Lena. “I want to know whether I’m about to cut something that’s working because I’m measuring the wrong return.” Lena nodded. Jonah looked at Carla. “That seems reasonable.” Carla pointed at him. “You’re enjoying this.” “A little.” Daniel opened a new spreadsheet. “Let’s start with wages.” Lena stopped him. “No.” He looked up. “What?” “Before you calculate anything.” She pointed at the board. “What decision are you actually authorized to make?” Daniel stared at her. Then at Carla. Carla stared back. “I don’t know,” he said. Jonah slowly put down his pen. Daniel opened the budget directive. They read it together. The language was simple: Each department shall identify opportunities to reduce variable labor expenditure by 5% while maintaining established quality and safety targets. Jonah read it again. “You aren’t required to reduce sitter hours.” Daniel shook his head. “No.” “You’re required to identify opportunities to reduce variable labor.” “Yes.” Carla leaned forward. “So why sitters?” Daniel looked at the staffing report. “Because they’re one of our fastest-growing variable labor categories.” Lena nodded. Reasonable. Then: “Fastest-growing by percentage or dollars?” Daniel stopped. Carla looked at the ceiling. Jonah whispered, “Here we go.” Daniel opened another report. Agency nursing. Overtime. Sitters. Temporary clerical coverage. Transport. The largest percentage increase was sitters. The largest dollar increase was agency nursing. By a lot. Daniel stared at the screen. Carla said nothing. Jonah said nothing. Lena waited. Daniel finally spoke. “We sorted the report by percentage.” “Yes.” “And started reviewing from the top.” “Yes.” He looked at her. “That isn’t necessarily wrong.” “No.” “But if the objective is five percent of variable labor expenditure—” “Then percentage growth and absolute expenditure answer different questions.” Daniel leaned back. “Right.” Nobody laughed. At 1:26, the staffing question changed. Not because anyone discovered new data. The data had been there. The representation had been there. The directive had been there. The error—if it was an error—was smaller than that. They had begun with: Which growing labor category should we reduce? The directive had actually asked: Where can variable labor expenditure be reduced while maintaining safety and quality? Jonah looked at both sentences on the board. “Those feel almost identical.” “Yes,” Lena said. “But they aren’t.” “No.” Daniel stared at them. “The first assumes reduction begins with growth.” “Yes.” “The second doesn’t.” “No.” Carla looked at Lena. “So what now?” Lena looked at the board. Expected return. Actual return. Variance. Changed conditions. Changed inputs. Changed decision. Then the two questions. “We don’t evaluate the sitter program yet.” Daniel nodded. “We evaluate whether it belongs in the candidate set.” “Yes.” “And if it does?” “Then we evaluate it against the correct return.” Jonah added, “Which we haven’t defined.” Daniel looked at him. “Not completely.” Carla laughed. “What?” “Nothing.” “No, tell me.” She shook her head. “I thought you came down here to ask whether a dashboard was being misused.” Lena looked at her. “We did.” “And now?” Jonah looked at the whiteboard. “Now we know we were asking the wrong question.” Lena considered that. “No.” Jonah looked at her. She pointed to the original dashboard. “It was the right question to get us here.” He smiled. “Different?” “More precise.” “I’ll take it.” At 1:41, they left Finance. Daniel remained behind with a new candidate-analysis request and a much shorter list of questions than the one they’d almost built. In the elevator, Jonah leaned against the wall. “You know what’s bothering me?” “Yes.” “Stop doing that.” Lena waited. He continued. “Nothing failed.” “No.” “The dashboard worked.” “Yes.” “The import worked.” “Yes.” “The staffing model worked.” “Yes.” “The fall report worked.” “Yes.” “The budget directive was clear.” “Reasonably.” “And we still almost evaluated the wrong intervention against the wrong return.” “Yes.” The elevator descended. Jonah looked at the numbers above the door. “So where was the failure?” Lena thought about it. Not because she lacked an answer. Because she had several. “The handoff.” “Between systems?” “Between questions.” Jonah looked at her. She continued. “Each representation remained usable long enough to reach the next decision.” “But nobody checked whether the next decision was the one it had been built to support.” “Yes.” The doors opened. Neither moved immediately. Jonah looked back at her. “What do we call that?” Lena stepped into the hallway. “We don’t.” He followed. “Why not?” “Not yet.” “Because?” She kept walking. “We have one example.” Jonah smiled. “Lucky guess?” Lena stopped. Turned. He looked entirely too pleased with himself. “That doesn’t mean what you think it means.” “It means we need another case.” CHAPTER SEVEN Available At 8:14 the next morning, Samantha had another case. She did not call it that. She sent Lena an email with a subject line: THIS MAY BE NOTHING Jonah was standing behind Lena when she opened it. He read the subject. “I like her.” “You already liked her.” “I like her more.” Lena opened the attachment. A single-page report. No patient names. No clinical notes. Just timestamps. Medication order entered: 14:03 Order verified: 14:07 Medication available: 14:09 Administration documented: 14:41 Jonah looked at it. “Thirty-eight minutes.” “Yes.” “Is that bad?” “Unknown.” “Then why did she send it?” Lena scrolled. There was one sentence beneath the table. All individual systems report completion within expected parameters. Jonah read it twice. “Oh.” Lena looked at him. “What?” “Nothing.” “That wasn’t nothing.” “No.” He pointed at the screen. “Yesterday.” “Yes.” “Everything worked.” “Yes.” “And something still took thirty-eight minutes.” “Yes.” Jonah pulled out a chair. “What was the medication?” Lena looked at the attachment. “Not listed.” “Patient?” “Not listed.” “Unit?” “Not listed.” “Expected administration time?” “Not listed.” He looked at her. “We don’t have enough.” “No.” “What do we do?” Lena closed the attachment. “Ask.” Samantha was waiting in clinical informatics. So was Megan. Megan looked annoyed. Not angry. Annoyed. Jonah noticed. “You know what happened.” “Yes.” “Bad?” “No.” “Then why do you look like that?” “Because everybody did their job.” Jonah looked at Lena. Lena ignored him. Samantha turned her monitor. The medication was an antibiotic. Routine order. Not STAT. Not emergent. No adverse outcome. No missed dose. No patient harm. Lena looked at Megan. “Then why are we here?” “Because the nurse called pharmacy.” “Why?” “The medication wasn’t there.” Samantha said, “Except pharmacy showed it as available.” Jonah looked at the timestamps again. “At 14:09.” “Yes.” “Where?” Megan pointed at the screen. “That’s the question.” The medication order began in the electronic health record. 14:03. Physician signed. Order transmitted. Pharmacy verification queue received it. 14:04. Pharmacist reviewed. No allergy conflict. Dose appropriate. Route appropriate. Verified at 14:07. Pharmacy dispensing system received the order. Medication already stocked on the unit. No physical dispense required. System marked: AVAILABLE 14:09. Jonah nodded. “So pharmacy is done.” “Yes,” Samantha said. Megan shook her head. “Pharmacy’s system is done.” Jonah looked at her. “Different?” “Yes.” Lena looked at the timeline. “Where was the medication physically?” “In the automated dispensing cabinet.” “Same unit?” “Yes.” “Accessible?” Megan hesitated. “Eventually.” Lena looked up. “What happened?” “The cabinet hadn’t received the patient assignment.” Samantha opened another log. The medication existed. The order existed. The patient existed. The cabinet existed. The nurse had access to the cabinet. But at 14:09, the cabinet did not associate that medication order with that patient. Jonah looked at the screen. “Broken interface?” “No.” “Network?” “No.” “Bad data?” “No.” “Then?” Samantha opened another timestamp. Patient transfer completed: 13:58. Megan pointed. “He had moved units.” The transfer had worked. Bed management showed the patient on the new unit at 13:58. The electronic health record showed the patient on the new unit at 13:58. Nursing assignment updated at 14:00. Medication order entered at 14:03. Pharmacy verified at 14:07. But the dispensing cabinet synchronized patient-location data on a separate cycle. Jonah looked at Samantha. “How separate?” “Five minutes.” “That doesn’t sound like thirty-eight.” “It isn’t.” Lena looked at the logs. “Show the synchronization.” Samantha did. 14:00. Then 14:05. Then 14:10. At 14:05, the patient was not present. At 14:10— Also not present. Lena pointed. “Why?” Samantha opened the interface queue. The patient-location message had been transmitted. Accepted. Processed. No error. Jonah leaned closer. “So where did he go?” Samantha looked at Megan. Megan crossed her arms. “Apparently nowhere.” They reconstructed the transfer. Original unit: 4 WEST Destination: 6 SOUTH Bed management: 6 SOUTH — 612B Electronic health record: 6 SOUTH — 612B Nursing assignment: 6 SOUTH — 612B Interface message: 6S — 612B Dispensing cabinet: SOUTH 6 — 612B Jonah looked at the screen. “That seems obvious.” Samantha nodded. “It is to us.” “Not to the system.” “No.” Lena pointed at the interface message. “Translation table?” “Yes.” Samantha opened it. 6S → SOUTH 6 Active. Correct. Jonah frowned. “Then it should work.” “Yes.” “But it didn’t.” “No.” “Why?” Samantha scrolled. There were two entries. 6S → SOUTH 6 and 6 SOUTH → SOUTH 6 Both active. Both valid. Lena waited. Samantha kept scrolling. Then stopped. “Oh.” Megan leaned over. “What?” Samantha highlighted another field. Effective date: 06/01 “Unit rename.” Megan looked at her. “We renamed the unit two months ago.” “From?” “6S to 6 SOUTH.” Jonah pointed at the translation table. “But both work.” “They should.” “Again with should.” Samantha ignored him. She opened the transfer message. The location field said: 6 SOUTH The interface accepted it. The translation table mapped it. The dispensing system received: SOUTH 6 Everything appeared correct. Lena looked at the timestamps. “Then this isn’t the failure.” Samantha nodded slowly. “No.” Megan looked frustrated. “So we’re back to nowhere.” “No,” Lena said. Megan looked at her. “We eliminated one branch.” Jonah nodded. “Useful.” Megan looked at both of them. “You two have a very strange definition of useful.” At 9:02, Samantha found the next boundary. The dispensing cabinet knew the patient was on SOUTH 6. But it still did not display the medication. Why? Because the medication profile was associated with the patient’s encounter. The encounter number in the cabinet was different. Jonah looked at the two strings. One ended in 47. The other in 74. “That looks like a typo.” “It isn’t,” Samantha said. “What is it?” “Two encounters.” “For the same patient?” “Yes.” “Why?” Megan answered. “He came through observation.” Then inpatient admission. Two billing encounters. One clinical stay. Lena looked at the screen. “Which encounter did pharmacy use?” “Inpatient.” “Cabinet?” “Observation.” “Why?” Samantha opened another log. The patient had entered the hospital under observation status. The dispensing cabinet received that encounter. When status changed to inpatient, the electronic health record created the new encounter and linked the two. Pharmacy received the new encounter. The cabinet received the new encounter too. But it did not replace the old one. It retained both. Jonah looked at Samantha. “Is that allowed?” “Yes.” “Why?” “Historical continuity.” Lena looked at him. He understood. “Useful.” “Yes.” “Until?” Lena didn’t answer. Samantha did. “Until another system has to decide which one is current.” The cabinet had a rule. When multiple active encounters existed for the same patient, prioritize the encounter associated with the current location. Reasonable. Both encounters had the same location. Second rule: Prioritize the earliest active encounter. Also reasonable. Observation came first. So the cabinet displayed the patient under the observation encounter. The medication order belonged to inpatient. Pharmacy marked the medication available because it was stocked and verified against the inpatient encounter. The cabinet correctly showed no medication order under observation. Megan looked at the screen. “So every system is right.” Samantha nodded. “Within its own rules.” Jonah looked at Lena. She was already writing. EHR — RIGHT PHARMACY — RIGHT CABINET — RIGHT LOCATION INTERFACE — RIGHT Underneath: PATIENT STILL DOES NOT HAVE MEDICATION Megan read it. “He did get it.” “Eventually.” “How?” Megan smiled slightly. “I called pharmacy.” Jonah looked at her. “And?” “They said it was available.” “Helpful.” “Very.” “What did you say?” “That it wasn’t.” “Then?” “They checked the cabinet.” “And?” “They said it should be there.” Jonah looked at Samantha. “There it is again.” Samantha nodded. Megan continued. “Then I pulled the medication under override.” Lena stopped writing. “Allowed?” “Yes.” “For this medication?” “Yes.” “Documented?” “Yes.” “Why thirty-eight minutes?” Megan looked at the timeline. “Because I tried to solve it the normal way first.” Lena turned. “Define normal way.” Megan counted. “Refresh the patient list.” One. “Log out and back in.” Two. “Check the other cabinet.” Three. “Check the medication room.” Four. “Call pharmacy.” Five. “Wait while pharmacy checks.” Six. “Call again.” Seven. “Then override.” Jonah looked at the clock. “And every one of those steps made sense.” “Yes.” “Did anyone own the thirty-eight minutes?” Megan looked at him. “What do you mean?” “Pharmacy owned verification.” “Yes.” “You owned administration.” “Yes.” “Informatics owns interfaces?” Samantha grimaced. “Depending on which interface.” “Cabinet?” “Pharmacy technology.” “Patient assignment?” “Nursing operations.” “Encounter logic?” “Revenue cycle and clinical systems.” Jonah looked at Lena. “That’s a lot of owners.” “Yes.” “Who owns the space between 14:09 and 14:41?” Nobody answered. At 9:26, Carla arrived. She looked around the room. “No.” Jonah smiled. “You don’t even know what happened.” “I know all four of you are staring at timestamps.” “Fair.” Lena handed her the timeline. Carla read it. “No harm?” “No,” Megan said. “Medication given?” “Yes.” “Within acceptable window?” “Yes.” “Then why—” She stopped. Jonah looked at her. Carla sighed. “I heard it.” Lena said nothing. Carla continued reading. “So what’s the concern?” Samantha pointed at the system states. “Every subsystem reports successful completion.” Carla looked at the thirty-eight-minute interval. “But the process didn’t.” “Correct.” “Is thirty-eight minutes clinically significant?” “For this medication?” Megan said. “Probably not.” Carla looked relieved. Then caught herself. Jonah smiled. “Learning.” “Shut up.” Lena drew a line across the board. At the left: ORDER At the right: ADMINISTRATION Between them she marked: EHR PHARMACY INTERFACE CABINET NURSING Then she drew small gaps between each. Jonah watched. “The gaps?” “Yes.” “What are they?” Lena looked at Samantha. Samantha answered. “Handoffs.” Lena shook her head. “Some.” She pointed between EHR and pharmacy. “Interface.” Between pharmacy and cabinet. “Interface.” Between cabinet and nursing. Megan answered. “Workflow.” Between nursing and pharmacy. “Communication.” Jonah looked at the line. “Different kinds of boundaries.” “Yes.” “But the medication crosses all of them.” “No,” Lena said. He looked at her. She pointed. “The medication doesn’t.” Jonah looked at the diagram again. “The representation does.” “Yes.” That changed the room. Samantha stood. “The order is represented differently at each layer.” Lena nodded. “EHR?” “Clinical order.” “Pharmacy?” “Verified medication order.” “Cabinet?” “Dispensable item attached to patient encounter.” “Nursing?” Megan answered. “Medication due to this patient.” Lena pointed at the patient transfer. “Bed management?” Samantha answered. “Patient occupying a location.” “Revenue cycle?” Carla looked at her. “Billable encounter.” Jonah stared at the board. “Same patient.” “Yes.” “Same medication.” “Yes.” “Different representations.” “Yes.” “And each one is valid.” “Within its decision.” He looked at her. “Until the handoff.” Lena considered. “Closer.” At 10:03, Samantha pulled three months of cabinet exceptions. Carla saw the number. “Absolutely not.” Lena looked at her. “What?” “Four thousand.” “We’re not reviewing four thousand.” “Thank God.” Samantha filtered. Multiple active encounters. Ninety-two. Same location. Thirty-one. Medication unavailable despite pharmacy verification. Eight. Megan leaned closer. “Eight?” “Yes.” “Same problem?” “Unknown.” Lena said, “Sample.” Carla looked suspicious. “How many?” Lena looked at Jonah. He considered. “Enough to determine whether the mechanism repeats.” Carla waited. “That isn’t a number.” “No.” Lena said, “Start with five.” Carla pointed at her. “Thank you.” “For what?” “A number.” Case one. Same mechanism. Observation to inpatient. Duplicate active encounter. Cabinet selected earlier encounter. Medication attached to later encounter. Nurse used override. Case two. Different. Patient name mismatch after registration correction. Resolved automatically. Case three. Same mechanism. Case four. Different. Medication physically absent despite inventory showing stock. Case five. Same mechanism. Jonah counted. “Three.” “Yes,” Lena said. “Enough?” “To establish recurrence.” “Not prevalence.” “No.” “Not clinical significance.” “No.” “Not cause of every delay.” “No.” He nodded. “Good.” Carla looked at him. “You’re checking her now?” Jonah looked surprised. “I always check her.” Lena continued reading. Carla looked at Lena. “You allow that?” “Yes.” Jonah smiled. Carla looked between them. “I continue to regret joining this.” They called Daniel at 10:28. He answered on speaker. “Why am I involved?” Lena said, “You’re not.” “Good.” “We have a question.” “That sounds like involvement.” Jonah leaned toward the phone. “Hypothetically—” “No.” Jonah looked offended. Daniel continued. “Every time someone says hypothetically, I end up with a spreadsheet.” Lena took over. “If a process delay occurs between independently successful systems, where would its cost appear?” Silence. Daniel answered more slowly. “Depends on the delay.” “Nursing time.” “Labor.” “Where?” “Unit productivity, maybe overtime if it accumulates.” “Pharmacy time?” “Pharmacy labor.” “Interface investigation?” “IT or informatics.” “Patient harm?” “Quality, treatment cost, length of stay, depending.” “And if there is no harm?” “Then probably nowhere as one event.” Jonah looked at the board. “What if it happens repeatedly?” Daniel paused. “Then the cost exists.” “Where?” “Distributed.” Lena wrote the word. DISTRIBUTED Daniel continued. “You may never see it as one line item.” “Could each department remain within target?” “Yes.” “And the total process still carry unnecessary cost?” “Yes.” Carla leaned toward the phone. “I hate that answer.” Daniel laughed. “Welcome to Finance.” After the call, Lena left DISTRIBUTED on the board. Jonah looked at it. “Cost?” “Yes.” “Ownership?” “Possibly.” “Failure?” “Unknown.” “Risk?” “Possibly.” Carla shook her head. “I’ve never heard two people say so little with so many words.” Megan smiled. “I think I’m starting to understand them.” “Don’t.” At 11:04, Dr. Vasquez arrived. Megan had asked her to come. Vasquez read the timeline. “Routine antibiotic?” “Yes.” “Thirty-eight minutes?” “Yes.” “No harm?” “No.” “Then clinically, this case is not very interesting.” Lena nodded. “Agreed.” Everyone looked at her. Vasquez laughed. “That was easier than expected.” Jonah said, “Give it a second.” Lena pointed at the three recurring cases. “What if the medication were time-sensitive?” Vasquez looked at the encounter logic. “Then it becomes interesting.” “Does the mechanism know which medication will encounter it?” “No.” “Does the system prevent the same condition from occurring with a time-sensitive medication?” Vasquez looked at Samantha. Samantha shook her head. “Not that I’ve found.” Vasquez’s expression changed. Carla saw it. “So now it’s a safety problem.” “No,” Lena said. Carla looked at her. “What?” “It is a mechanism capable of producing a safety problem.” Vasquez nodded. “Different.” Jonah looked at Carla. Carla pointed at him. “Don’t.” They returned to the eight cases. Medication classes. Timing. Patient status. Outcome. None had caused documented harm. Two had involved medications where a longer delay could plausibly matter. Vasquez refused to generalize further. “That’s not enough to quantify clinical risk.” Lena nodded. “Agreed.” Jonah looked at Vasquez. “You two agreeing is unsettling.” Vasquez smiled. “We know what we don’t know.” Lena looked at her. That was better. At 11:32, Samantha asked the question nobody had asked yet. “Why does the cabinet prioritize the earliest encounter?” Nobody knew. She opened the configuration documentation. The rule had been installed seven years earlier. Reason: Preserve medication continuity during encounter transitions. Megan read it. “That makes sense.” “Yes,” Samantha said. “Was observation handled the same way seven years ago?” Carla asked. “Why?” Samantha was already searching. The hospital had changed observation workflow four years earlier. Then again two years ago. The cabinet rule had not changed. Jonah looked at Lena. “Old solution.” “Yes.” “New environment.” “Possibly.” Samantha found the implementation note. The original rule solved a real problem. When patients transferred between care settings, medication profiles sometimes disappeared briefly while encounters changed. Prioritizing the earliest active encounter preserved access. Megan nodded. “So it fixed missing medications.” “Yes.” Jonah pointed at the current cases. “And now it’s hiding new ones.” Samantha looked at him. “Sometimes.” He smiled. “I’m learning too.” Lena wrote: RULE CREATED TO PRESERVE CONTINUITY Then: ENVIRONMENT CHANGED Then: RULE REMAINED Then: CURRENT EFFECT ≠ ORIGINAL EFFECT She stared at the last line. Erased it. Jonah noticed. “Too strong?” “Yes.” She replaced it. CURRENT EFFECT MAY DIFFER FROM ORIGINAL EFFECT Jonah nodded. “Better.” Carla looked at him. “You know that’s contagious, right?” At 12:06, the group had three choices. Change the cabinet rule. Change encounter handling. Add an alert. Nobody liked any of them. Megan wanted the cabinet fixed. Samantha said changing encounter priority could reintroduce the original continuity problem. Carla wanted an alert. Megan pointed out nurses already had too many alerts. Vasquez asked whether the issue could be resolved upstream. Samantha said maybe. Jonah watched the discussion move faster. Solution. Counterexample. New solution. New constraint. Another solution. Another constraint. Lena watched too. Then: “Stop.” Everyone stopped. Jonah smiled slightly. Carla did not. Lena pointed at the board. “What are we deciding?” Samantha answered first. “How to fix the encounter problem.” “No.” Megan tried. “How to prevent medication delays.” “Too broad.” Vasquez looked at the three recurring cases. “How to prevent the cabinet from selecting an encounter that cannot display a current medication order.” Lena considered. “Closer.” Jonah looked at the original rule. “Without breaking the continuity the rule was designed to preserve.” Lena looked at him. “Yes.” Samantha leaned back. “That changes the solution space.” “How?” Carla asked. “We don’t have to eliminate earliest-encounter priority.” Megan said, “We need a condition where it shouldn’t apply.” “Maybe.” Carla looked at Lena. “Can we please appreciate that someone else said maybe?” “No.” Samantha sketched a possible rule. If multiple active encounters exist: retain earliest encounter priority unless a later encounter has: current unit match active medication orders active nursing assignment Then prioritize the later encounter for medication display. Megan looked at it. “That sounds obvious.” Samantha nodded. “Which worries me.” Jonah pointed. “I like her even more.” Lena asked, “Feasible?” Samantha shook her head. “I don’t know.” “Why?” “Vendor configuration. Interface dependencies. Regulatory validation. Testing. I know what I want the system to do. I don’t know whether this cabinet can do it safely.” Lena nodded. “What do you need?” “Pharmacy technology. Vendor documentation. Probably revenue cycle. Maybe the vendor.” Jonah looked at the board. “So the answer is?” Samantha closed her laptop. “My map doesn’t contain enough terrain.” Lena looked at her. “Good.” Carla stared. “How is not knowing good?” “Because she knows where the boundary is.” At 12:41, the meeting ended without a fix. Megan had a workaround. Samantha had a bounded technical question. Vasquez had a clinical-risk question. Carla had three departments to notify. Jonah had two pages of notes. Lena had the board. Carla stood at the door. “So what exactly did we accomplish?” Lena pointed. The medication delay was not owned by one system. The representation changed at each boundary. Each local decision remained reasonable. A legacy rule remained useful for its original purpose. The environment changed around it. The current solution cannot be selected without external technical constraints. Carla read them. “So we found the problem.” “No.” Carla closed her eyes. Jonah smiled. Lena continued. “We found a repeatable mechanism.” “Isn’t that the problem?” “Part of it.” “What else is there?” Lena looked at the first line. “The system can detect local failure.” “Yes.” “It can detect interface failure.” “Yes.” “It can detect medication error.” “Yes.” She pointed at the thirty-eight-minute gap. “What detects this?” Nobody answered. At 1:03, Lena and Jonah were alone again. The room was quiet. Jonah stood at the board. “Yesterday you said the failure was between questions.” “Yes.” “Today it’s between representations.” “Possibly.” “Are those the same?” “No.” “Related?” “Probably.” He looked at her. She looked back. “What?” “You said probably.” “Yes.” “I just wanted to appreciate it.” “Finished?” “No.” She waited. He turned back to the board. “Every system had an owner.” “Yes.” “Every field had a definition.” “Yes.” “Every handoff had a protocol.” “Mostly.” “Every local decision made sense.” “Yes.” “And nobody owned whether the whole thing still represented the same thing after it crossed all of them.” Lena stopped. Jonah looked at her. “What?” She didn’t answer. He waited. She walked to the board. Under everything else, she wrote: WHO OWNS THE REPRESENTATION BETWEEN SYSTEMS? Jonah read it. “That feels bigger.” “Yes.” “Too big?” “Maybe.” He smiled. She looked at him. “What?” “Nothing.” “That wasn’t nothing.” “No.” She capped the marker. “Don’t.” “I didn’t say anything.” “You were going to.” He opened the door. “Lucky guess.” Lena looked at the board. One case yesterday. Three recurring cases today. Still not enough. But no longer one. She followed him into the hallway. Behind them, every system remained operational. Every owner still owned their part. And the space between them remained unassigned. CHAPTER EIGHT The Gap At 8:22 the next morning, Carla Mendoza had an answer. Lena knew because Carla entered the conference room carrying no coffee. Jonah noticed too. “That bad?” Carla put a folder on the table. “No.” He looked at her empty hands. “Then where’s the coffee?” “I forgot it.” Jonah looked at Lena. Lena looked at Carla. Carla pointed at both of them. “Don’t.” Lena opened the folder. On top was a printed organizational chart. Clinical informatics. Pharmacy. Nursing operations. Revenue cycle. Quality. Information technology. Patient safety. Risk management. Facilities. Finance. More beneath them. Lines connected departments. Boxes nested inside larger boxes. Some lines were solid. Others dotted. A few had arrows. Jonah leaned over the table. “You found the owner.” “No,” Carla said. Lena kept reading. “That’s why she forgot the coffee.” Carla sat. “Yes.” The question Lena had left on the board remained behind them. WHO OWNS THE REPRESENTATION BETWEEN SYSTEMS? Carla pointed to the organizational chart. “I asked.” “Who?” Lena said. “Everyone who seemed remotely plausible.” “And?” Carla opened the second page. Clinical informatics owned configuration inside the electronic health record. IT integration owned message transport. Pharmacy technology owned dispensing-system configuration. Revenue cycle owned encounter definitions. Nursing operations owned assignment workflows. Patient safety owned event review. Quality owned process improvement. Each answer was clear. None answered the question. Jonah read the list. “So everyone owns a boundary.” “Parts of one,” Carla said. “Who owns the transition?” Carla looked at him. “That depends on the transition.” Lena nodded. “Good.” Carla stared. “No. Not good.” “It’s more precise.” “That is becoming your response to everything I don’t like.” “Not everything.” Jonah said, “Most things.” Lena ignored him. Carla continued. “Integration says they guarantee message delivery.” “Not meaning,” Lena said. “No.” “Pharmacy technology?” “Medication functionality.” “Not patient identity across encounter states.” “No.” “Revenue cycle?” “Encounter logic.” “Not medication availability.” “No.” “Nursing?” “Workflow.” “Not interface behavior.” “No.” Jonah looked at the chart again. “So nobody failed to own what they were assigned.” “Correct.” “And nobody was assigned this.” Carla looked toward the board. “Correct.” Samantha arrived at 8:37. She had coffee. Jonah looked relieved. “For you?” “Yes.” “Good.” Carla looked offended. Samantha sat. “I got your email.” Lena turned the organizational chart toward her. “Anything wrong?” Samantha read it. “No.” “Anything missing?” “Probably.” “Relevant?” Samantha traced the lines. “Yes.” She pointed to a dotted connection between clinical informatics and revenue cycle. “This says shared governance.” Carla leaned closer. “For what?” “Encounter-state changes.” Lena looked at her. “Yesterday’s mechanism?” “Part of it.” “Who chairs it?” Samantha pulled out her tablet. “A physician informaticist and someone from revenue cycle.” Jonah sat back. “Found them.” Samantha shook her head. “Found a committee.” He paused. “Different.” “Yes.” The committee was called the Clinical Encounter Governance Group. It met monthly. Its charter was easy to find. Purpose: ensure consistent implementation of encounter definitions, status transitions, and associated clinical and administrative workflows. Lena read it twice. Carla noticed. “That sounds like ownership.” “Possibly.” Jonah looked at her. She looked back. “What?” “Nothing.” “That wasn’t nothing.” “No.” Samantha opened the meeting archive. Agendas. Minutes. Change requests. Observation-to-inpatient conversions appeared repeatedly. Billing. Documentation. Order continuity. Discharge workflow. Patient-location updates. Authorization. Jonah pointed to one item. “Medication cabinets.” Samantha opened it. Two years earlier. Issue: duplicate patient profiles during encounter transition may create user confusion in automated dispensing systems. Megan entered as Samantha read it aloud. She stopped inside the door. “You already knew?” Samantha looked up. “I didn’t.” “The hospital did.” “Someone did.” Lena looked at the minutes. “Different.” Megan came closer. “What happened?” Samantha scrolled. The issue had been reviewed. Frequency described as low. No documented medication errors. Existing override workflow considered adequate. Vendor limitation noted. Recommendation: MONITOR. NO CONFIGURATION CHANGE AT THIS TIME. Jonah looked at Megan. She read it again. “Monitor how?” Nobody knew. Carla searched the later agendas. The issue appeared once more three months later. No additional reported events. Then disappeared. Megan folded her arms. “We found three yesterday.” Samantha nodded. “None were reported as medication events.” “Why not?” “No wrong medication. No missed administration window. No documented harm.” Jonah looked at Lena. She was already looking at the words. No additional reported events. He said, “Technically true.” “Yes.” Megan looked frustrated. “So the committee did what it was supposed to do, and the problem still disappeared.” Lena shook her head. “The problem didn’t disappear.” “What did?” “The representation.” They reconstructed one of yesterday’s cases. Nursing documentation: Medication delayed while resolving dispensing-cabinet patient profile. Administered via override. Pharmacy: No error. Cabinet log: Successful override. Medication administration record: Completed within acceptable window. Patient safety system: No report. Encounter governance: No new event. Jonah looked at the sequence. “It existed in four places.” “Yes,” Lena said. “And still wasn’t present where monitoring happened.” “Yes.” Samantha went still. Carla looked at her. “What?” Samantha turned the tablet around. “The committee didn’t monitor the mechanism.” Lena waited. “They monitored reports.” “Yes.” “And reports required a threshold.” “Yes.” “Which these cases didn’t cross.” “Yes.” Samantha looked back at the old recommendation. “So ‘monitor’ meant ‘wait for someone to formally report something.’” Carla frowned. “Isn’t that usually what monitoring means?” “No,” Lena said. Carla looked at her. “What does it mean?” “It depends on what you’re trying to detect.” Megan sat. “If nobody reports it, how would they know?” Jonah answered. “They wouldn’t.” “That’s ridiculous.” “Is it?” She looked at him. He continued. “You don’t report every time you have to refresh a screen.” “No.” “Every time a cabinet is slow.” “No.” “Every time you call pharmacy and solve something.” “No.” “Why?” “Because then I’d spend all day reporting things that didn’t matter.” Jonah nodded. “So not reporting is reasonable.” Megan leaned back. “I hate when you do that.” “What?” “Make the thing I’m complaining about make sense.” Lena said, “Both can be true.” Megan looked at her. “That doesn’t help.” “It isn’t supposed to.” At 9:04, Lena wrote on the board: MONITOR Below it: WHAT SIGNAL? Then: WHAT THRESHOLD? Then: WHERE? Jonah read them. “Not who?” “Not yet.” Carla pointed at the original question. WHO OWNS THE REPRESENTATION BETWEEN SYSTEMS? “You’re abandoning that?” “No.” “Looks like it.” “It was too early.” Jonah smiled. Lena saw him. “What?” “You said too early.” “Yes.” “You usually say wrong.” “Those are different.” “I know.” Samantha returned to the old committee minutes. “What if ownership isn’t the missing piece?” Lena turned. “Go on.” Samantha pointed to the charter. “They had owners.” “Of what?” “The transition rules. The workflow. The known issue.” “And?” “They still couldn’t see recurrence because their monitoring signal started after the thing they were trying to observe.” Silence. Jonah looked at the board. “What happened before the report?” Megan answered. “Work.” “Meaning?” “Calls. Overrides. Refreshing. Looking for the medication. Asking pharmacy.” Lena wrote: COMPENSATION Megan read it. “That sounds dramatic.” “Is it inaccurate?” “No.” Carla said, “Workaround?” Lena considered. “Sometimes.” Jonah looked at the word. “Compensation feels broader.” “Yes.” They pulled the three recurring cases again. Case one: Nurse refreshed cabinet twice. Called pharmacy. Override. Case three: Nurse checked two cabinets. Asked charge nurse. Called pharmacy. Override. Case five: Pharmacy reverified order. Nurse logged out. Cabinet resynchronized. Medication appeared. No override. Lena circled the actions. “Every case resolved locally.” “Yes,” Megan said. “Before escalation.” “Yes.” “Before report.” “Yes.” “Before governance.” “Yes.” Jonah looked at the board. “So the system is good at recovering.” Carla nodded. “That’s supposed to be good.” “It is,” Lena said. Everyone waited. She continued. “Recovery can hide recurrence.” Nobody spoke. Jonah looked at the three cases again. “Because every successful workaround removes the reason to escalate.” “Yes.” Megan frowned. “But if we escalate every workaround, we drown.” “Yes.” “Then what do you want?” Lena looked at the cases. “I don’t know.” Carla laughed. Megan looked at her. “You’re getting used to that?” “No.” “Yes,” Jonah said. At 9:31, Daniel joined by video. Carla had invited him. He appeared on the monitor holding coffee. Jonah looked around the room. “Apparently I’m the only one without one.” Lena slid her untouched cup toward him. He looked at it. “Black?” “Yes.” He took it. Carla stared. “What?” “Nothing.” “That was absolutely something.” Daniel said through the monitor, “Should I leave?” “No,” Lena said. Carla answered at the same time. “Probably.” Daniel stayed. Lena showed him the three cases. “What do the workarounds cost?” Daniel looked at the actions. “Minutes.” “Whose?” “Nursing. Pharmacy. Maybe informatics if escalated.” “Where does that cost appear?” “Productivity.” “Specifically?” “It probably doesn’t.” “Why?” “Because each event is too small.” Jonah looked at him. “And together?” “Depends on frequency.” “Which we don’t know.” “No.” Lena pointed at the old governance note. Frequency described as low. “How was frequency measured?” Daniel read. “Reports?” “Yes.” “So actual frequency is unknown.” “Correct.” “Expected cost?” “Unknown.” “Actual cost?” “Unknown.” “ROI of fixing it?” “Unknown.” Daniel smiled. “Excellent business case.” Carla groaned. Daniel leaned closer to his camera. “Can you estimate recurrence from system logs?” Samantha answered. “Partly.” “Which part?” “Multiple active encounters, cabinet mismatch, override, timing.” “Can you identify nursing workarounds?” “Not consistently.” Megan said, “Because half of them happen outside structured documentation.” Lena looked at her. “Phone calls.” “Yes.” “Walking to another cabinet.” “Yes.” “Asking another nurse.” “Yes.” “Refreshing.” “Yes.” “Waiting.” “Yes.” Jonah looked at Daniel. “Unpriced labor.” Daniel nodded. “More accurately: labor already paid for but diverted from something else.” Lena looked at him. “Opportunity cost.” “Yes.” Megan sighed. “So now me standing at a cabinet is a financial model.” “No,” Daniel said. “Good.” “Your time is.” She stared at him. “That was not better.” At 10:02, they had a new diagram. SYSTEM CONDITION ↓ LOCAL FRICTION ↓ HUMAN COMPENSATION ↓ SUCCESSFUL RECOVERY ↓ NO FORMAL EVENT ↓ NO GOVERNANCE SIGNAL ↓ CONDITION PERSISTS Jonah read it from top to bottom. Then bottom to top. “That loops.” “Yes,” Lena said. Carla leaned toward the board. “Not exactly.” Lena looked at her. Carla pointed. “If the condition gets bad enough, something eventually crosses the reporting threshold.” “Maybe.” “There she is,” Jonah said. Carla ignored him. “A medication error. A serious delay. A complaint.” “Yes.” “So the loop breaks.” Lena nodded. “After consequence.” Carla stopped. Megan looked at the board. “That’s what bothers me.” Jonah asked, “What?” “If the workaround keeps succeeding, we call the system fine.” Samantha added, “If it stops succeeding, we call it an incident.” Daniel said, “Which means the same mechanism changes category only after outcome changes.” Lena looked at him. “Keep that.” He smiled. “Do I get a sticker?” Jonah turned toward Lena. She pointed at him. “No.” At 10:19, Dr. Vasquez arrived. She read the diagram. Then read it again. “So what are you proposing?” “Nothing,” Lena said. Vasquez nodded. “Good.” Jonah looked at her. “You’ve adapted quickly.” “I work in medicine. People proposing solutions before defining problems is not new.” Carla said, “I like her.” “You already liked her,” Jonah said. “More now.” Lena looked at Vasquez. “When do clinicians document workarounds?” “When they matter.” “Define matter.” Vasquez smiled. “There it is.” She considered. “If the workaround changes care. Delays care significantly. Creates risk. Requires escalation. Something unusual.” “What if it becomes normal?” Vasquez stopped smiling. Megan answered first. “Then we stop thinking of it as a workaround.” Jonah looked at her. “What is it?” “Work.” The room went quiet. Not dramatically. Enough. Lena wrote one word beneath the loop. NORMAL Then: NORMAL TO WHOM? Megan pointed at the cases. “To nurses.” Samantha said, “Not to informatics.” Daniel said, “Not to Finance.” Carla added, “Not to Quality.” Vasquez looked at the cabinet sequence. “Probably not to physicians.” Jonah looked around. “So the people absorbing the friction know it exists.” “Yes,” Lena said. “And the people measuring the system may not.” “Yes.” Carla looked at the organizational chart. “Because the compensation happens below the threshold that reaches them.” Lena nodded. Samantha looked at the board. “That’s a representation boundary.” “Possibly.” Jonah looked at Lena. She ignored him. Vasquez pointed to NORMAL TO WHOM? “That one matters more.” Lena looked at her. “Why?” “Because every department probably has things like this.” Carla groaned. “Do not say every department.” Vasquez smiled. “I said probably.” Jonah looked at Lena. “Contagious.” “Don’t.” At 10:47, Carla received a message. She read it and frowned. “What?” “Nothing.” Jonah looked at her. She sighed. “Transport wants to know why Quality is asking for cabinet logs.” Lena looked at Samantha. Samantha looked at Carla. Carla raised both hands. “I don’t know.” Megan laughed. Daniel disappeared from the monitor for a moment, then returned. “What did I miss?” “Nothing,” Carla said. Jonah smiled. Lena said, “Tell Transport we aren’t.” Carla typed. Then stopped. “Actually.” Lena looked at her. “What?” “Transport has a workaround problem.” Jonah leaned back. Carla caught herself. “No.” Lena said, “Relevant?” “Probably.” “Necessary to answer the current question?” Carla thought. “No.” “Then leave it.” Carla stared at her. “That is deeply satisfying.” Jonah looked at Lena. “Progress.” “Don’t.” At 11:03, the group reduced the problem again. Not: Who owns the representation between systems? Not yet. Not: How many workarounds exist? Too broad. Not: Should every workaround be reported? Obviously not. Lena wrote: HOW DOES A SYSTEM DISTINGUISH RESILIENCE FROM HIDDEN LOAD? Samantha read it. “That’s better.” Daniel said, “That’s measurable in pieces.” Megan asked, “What pieces?” He counted. “Frequency.” Samantha added, “Duration.” Megan said, “Effort.” Vasquez said, “Clinical consequence.” Carla added, “Escalation.” Jonah looked at the board. “Repeatability.” Lena turned. “Explain.” “If five people independently invent the same workaround, that’s different from one person doing something weird once.” Lena nodded. “Keep it.” He smiled. They built a small capture plan. Not a new reporting system. Not another form. Megan rejected that immediately. Instead, for one week on one unit: Existing cabinet logs. Existing medication administration timestamps. Existing pharmacy call data where available. Short voluntary end-of-shift check from nurses: Did you spend time today resolving a medication-access problem that did not become a formal event? Yes or no. If yes: Approximate minutes. No patient narrative. No diagnosis. No explanation required. Samantha looked at Lena. “You’re leaving out cause.” “Intentionally.” “Why?” “We’re measuring hidden load first.” Daniel nodded. “Then decide whether deeper investigation has enough anticipated value.” Jonah looked at him. “You’re settling in.” “I hate all of you.” Carla smiled. “Welcome.” Everyone looked at her. Carla froze. Jonah slowly smiled. Lena looked at Carla. Carla pointed at Jonah. “No.” Lena returned to the board. Outwardly. At 11:41, they dispersed. Daniel went back to Finance. Vasquez to rounds. Megan to patients. Samantha to informatics. Carla went to find the coffee she had forgotten hours earlier. Jonah stayed. He looked at the original question still faintly visible beneath the newer writing. WHO OWNS THE REPRESENTATION BETWEEN SYSTEMS? “You didn’t erase it.” “No.” “Still relevant?” “Yes.” “Still the question?” “No.” He waited. Lena looked at the loop. SYSTEM CONDITION LOCAL FRICTION HUMAN COMPENSATION SUCCESSFUL RECOVERY NO FORMAL EVENT NO GOVERNANCE SIGNAL CONDITION PERSISTS Then at: HOW DOES A SYSTEM DISTINGUISH RESILIENCE FROM HIDDEN LOAD? “The ownership question assumes the representation exists.” Jonah looked at her. “And sometimes it doesn’t.” “Not where the decision is being made.” He nodded slowly. “So before we ask who owns it—” “We have to ask whether anyone can see it.” Jonah looked at the board. “That feels obvious.” “Yes.” “Which usually means?” Lena capped the marker. “We should verify it.” He smiled. “Good.” At 12:06, Megan returned. Alone. She stood in the doorway. “Do you have a minute?” Lena looked at Jonah. Jonah looked at Megan. “What happened?” “Nothing.” Neither moved. Megan sighed. “Fine. Something.” She stepped inside. “Remember yesterday when we talked about people compensating for the cabinet?” “Yes,” Lena said. Megan nodded toward the hall. “One of the new nurses just asked me why everyone keeps a list of extension numbers taped inside the medication-room cabinet.” Jonah looked at her. “What numbers?” “Pharmacy. Transport. Lab. Bed management. Respiratory. IT help desk.” Lena waited. Megan continued. “I told her because the directory takes too long.” Jonah said nothing. Megan looked at Lena. “Then she asked why nobody fixed the directory.” Lena looked at the board. Local friction. Human compensation. Successful recovery. Normal. Megan followed her gaze. “Do not tell me we’re doing the phone list next.” “No.” “Good.” Lena picked up her notebook. “We’re looking at the tape.” Megan closed her eyes. Jonah stood. “How long?” CHAPTER NINE The Tape The tape had been there long enough to collect dust along the edges. “Yes.” “And this?” 4421 — USE THIS ONE Megan smiled. “That one works.” “For what?” She stopped. “I actually don’t know.” Jonah looked at Lena. She looked back. Neither said anything. Megan sighed. “Fine. I’ll call it.” 4421 rang twice. A man answered. “Central supply.” Megan looked surprised. “This is Megan on Six South. Sorry, quick question. Why do we have your extension taped inside the medication-room cabinet?” Silence. Then: “You have what?” Megan looked at Jonah. Jonah covered his mouth. Lena did not. “A handwritten list of extensions.” “Okay.” “One says ‘4421 — use this one.’” Another pause. “That sounds like Frank.” “Who’s Frank?” “Retired.” “When?” “Three years ago.” Megan closed her eyes. Jonah smiled. The man continued. “What do you use it for?” “That’s what I’m trying to figure out.” “Good luck.” The call ended. Jonah looked at the tape. “Frank.” Lena looked at him. “What?” “Nothing.” “That wasn’t nothing.” “No.” They did not remove the tape. That was Lena’s decision. Megan had already reached for it. “Why not?” “Because it’s still part of the operating system.” “It’s wrong.” “Some of it.” “Which makes it worse.” “Possibly.” Jonah nodded. Megan looked at him. “Don’t encourage her.” “I’m not.” “You are.” Lena pointed at the list. “Who uses it?” Megan looked toward the hall. “Everyone.” “Define everyone.” “Anyone who needs to call somebody quickly.” “Why not use the directory?” “It’s slow.” “How slow?” Megan shrugged. “Depends.” Jonah looked at Lena. She looked at Megan. “Show us.” The official directory lived inside the hospital intranet. Three clicks from the workstation desktop. Then a search box. Megan typed pharmacy. Twenty-eight results appeared. Retail pharmacy. Outpatient pharmacy. Inpatient pharmacy. Pharmacy administration. Pharmacy clinical services. Investigational pharmacy. Medication assistance. Several satellite locations. Jonah leaned closer. “I see the problem.” Megan clicked Inpatient Pharmacy. A department page opened. Three numbers. Main. Clinical. Operations. “Which one?” Lena asked. “Usually main.” “Usually?” “If it’s a medication question.” “And if medication is missing?” “Main.” “Cabinet issue?” “Sometimes operations.” “Order issue?” “Clinical.” Jonah looked at the tape. The tape had one number. PHARM — 3198 He pointed. “Which one is that?” Megan called. Inpatient pharmacy main. Lena wrote nothing. “Now Transport,” she said. Megan searched. Fourteen results. Patient transport. Diagnostic transport. External transport. Transport dispatch. Transport leadership. Equipment transport. Jonah looked at the tape. TRANSPORT — 2770 Megan called. Patient transport dispatch. Lena looked at her. “Directory?” “Technically correct.” “Tape?” “Also correct.” “Difference?” Megan answered immediately. “Time.” Jonah looked at Lena. “Hidden load.” “Maybe.” Megan said, “No. Definitely.” Lena looked at her. Megan raised both hands. “Fine. Probably.” They timed it. Not scientifically. Enough to compare. Using the tape: Find number. Dial. Eight to twelve seconds. Using the directory: Open intranet. Navigate. Search. Interpret results. Choose. Twenty-six to fifty-three seconds. Longer if the department name did not match the term the nurse used. Longer still if the first number was wrong. Jonah looked at the numbers. “That’s not much.” “No,” Lena said. Megan looked at him. “Do it forty times.” He nodded. “Different.” Lena looked at her. “How often?” “No idea.” “Can we find out?” “Not easily.” Jonah smiled. “There it is.” Lena looked at him. “What?” “Nothing.” “That wasn’t nothing.” “No.” At 12:31, Samantha arrived. She stared at the tape. “This?” “Yes,” Lena said. “This is what you pulled me away for?” “Yes.” Samantha looked at Megan. Megan shrugged. “I warned you.” Samantha read the numbers. “These are probably all in the directory.” “Most are,” Lena said. “Then what’s the issue?” Jonah pointed at the workstation. “Find bed management.” Samantha sat. Opened the directory. Typed bed management. No result. She tried patient placement. Five results. She looked at Megan. “Which?” Megan pointed at the tape. BED MGMT — 4012 Samantha called. Patient placement. She looked back at the search results. “Oh.” Jonah smiled. “That syllable again.” Samantha ignored him. “What does staff call the department?” “Bed management,” Megan said. “What does the directory call it?” “Patient placement.” “So the tape translates.” Lena looked at the strip. “Yes.” Not just numbers. Language. Samantha pulled up the directory’s department taxonomy. Official names. Formal organizational structure. Cost centers. Service lines. Jonah looked at Megan. “Do nurses use any of these names?” “Some.” “Which ones?” “Pharmacy. Lab. Respiratory.” “Patient placement?” “Bed management.” “Clinical engineering?” “Biomed.” “Environmental services?” “EVS.” “Information services?” “IT.” “Centralized patient transportation?” “Transport.” Lena looked at Samantha. “The directory represents the organization.” “Yes.” “The tape represents the workflow.” Samantha stopped. Megan looked at the list again. “Oh.” Jonah said, “You can have that one.” “Thank you.” Lena took a picture of the tape. Not because she intended to preserve it permanently. Because somebody eventually needed to clean the cabinet. Megan watched. “Now can I remove it?” “No.” “You just photographed it.” “That preserves the artifact.” “Then why keep it?” “Because we don’t know which functions it still serves.” Megan stared. Jonah said, “She’s right.” “I know.” “You sound unhappy.” “I am.” They tested the unknown entries. LAB — 3320 Correct. RT — 2477 Respiratory therapy. Correct. IT — 6110 Help desk. Correct. 4126 No label. Disconnected. 4421 — USE THIS ONE Central supply. Probably historical. 3894 No label. Answered by a unit clerk on Five East. Megan frowned. “That makes no sense.” “Call history?” Lena asked. “Not available to us.” Samantha looked at the tape. “Could have been reassigned.” “How often do extensions change?” “Not often.” “Department names?” “More often.” Jonah looked at the list. “So the number can stay stable while the meaning around it changes.” Samantha nodded. “Or the opposite.” “Meaning stays. Number changes.” “Yes.” Lena looked at the crossed-out entries. Representation drift. Again. But this time physical. Visible. No software required. Carla arrived at 1:02. She saw the tape. “No.” Megan laughed. “That was my response.” Carla looked at Lena. “Please tell me we are not doing a quality review of masking tape.” “No.” Carla relaxed. “We are reviewing the directory.” Carla stopped relaxing. “Why?” Samantha answered. “Because the tape may be a compensating interface.” Carla looked at her. “You’re one of them now.” “I know.” Jonah looked pleased. Lena did not. Outwardly. They moved to the conference room. The tape stayed in place. On the board Lena wrote: OFFICIAL DIRECTORY Below: ORGANIZATIONAL REPRESENTATION Then: LOCAL TAPE Below: OPERATIONAL REPRESENTATION Jonah looked at the two. “Neither is wrong.” “No.” “Neither is complete.” “No.” “Different purpose.” “Yes.” Carla sat. “So why does this matter?” Megan answered before Lena could. “Because one of them is faster when I’m trying to get something done.” Carla nodded. “That’s fair.” Samantha added, “And because the local version is maintaining translations the official version doesn’t know it needs.” Lena looked at her. “Keep that.” They found three more lists before 2:00. One taped under a nurses’ station desk. One inside a supply-room door. One printed on a badge card. Different units. Different numbers. Different handwriting. Same pattern. Jonah lined the photos up on the monitor. “This is not one person being clever.” “No,” Lena said. “Independent compensation.” “Maybe.” Carla looked at her. “Three units.” “Still maybe.” “Why?” “Because we don’t know whether the mechanism is the same.” Samantha pointed at the lists. “It looks the same.” “Looking the same is not enough.” Jonah nodded. Carla sighed. “Fine.” The second unit’s list solved a different problem. The directory had the right department names. The numbers were right. Staff used the card because the workstations frequently timed out between uses. The third unit’s list existed because some rooms had no workstation nearby. The fourth list included personal extensions for specific people who “actually answered.” Lena stopped there. “Different.” Jonah looked at the last list. “More human.” “Less governed.” “Also more useful?” “Sometimes.” Carla pointed. “That one is dangerous.” “Why?” “People leave. Roles change. Personal numbers bypass formal routing.” Megan said, “And sometimes formal routing takes ten minutes.” Carla looked at her. “Both can be true.” Megan smiled. “Look at you.” “Don’t.” Lena reorganized the cases. Not by unit. By function. TRANSLATION Official name differs from operational name. ACCESS Directory exists but is inconvenient to reach. SPEED Official path is slower than local shorthand. RELIABILITY Staff prefer contacts that consistently answer. PROXIMITY Information placed where the work occurs. Jonah read the list. “So the tape isn’t one workaround.” “No.” “It’s a container.” Lena looked at him. “For?” “Multiple local fixes that happen to use the same medium.” She nodded. “Better.” At 2:24, Daniel joined by video again. He saw the photos. “No.” Jonah laughed. “Everybody says that.” Daniel pointed at one of the lists. “Why am I looking at masking tape?” Carla answered. “Opportunity cost.” Daniel stopped. “Fine.” Lena explained the directory timing. Daniel looked unimpressed. “Forty seconds.” “Per lookup,” Megan said. “How many per shift?” “We don’t know.” “So we can’t price it.” “Not yet.” Daniel looked at Lena. “Anticipated ROI of finding out?” “Low cost.” “What return?” “Unknown.” “That’s not helpful.” Jonah said, “Potentially enough to determine whether we should keep going.” Daniel thought. “Bound the study.” Lena nodded. “One unit. One day.” Megan looked horrified. “You’re going to count phone lookups?” “No.” “Good.” “We’re going to compare existing call behavior against directory use.” Megan looked at Samantha. “Can we do that?” “Maybe.” Lena looked at Samantha. Samantha corrected herself. “Possibly.” Jonah smiled. No one acknowledged it. The cheap study became even cheaper. No surveillance. No new logging. Megan asked ten nurses one question at shift change: When you need a department quickly, what do you use first? Tape. Memory. Saved contact. Coworker. Directory. In that order. The directory came last. Samantha stared at the results. “That’s not statistically useful.” “No,” Lena said. “It’s ten people.” “Yes.” “Single unit.” “Yes.” “One shift.” “Yes.” Daniel said through the monitor, “Then why do I care?” Lena looked at the responses. “Because the official resource is not the default resource.” Jonah added, “Which means staff built another one.” Carla looked at the photos. “Several.” Samantha leaned back. “So if we improve the directory, they might still use memory.” “Possibly.” “Saved contacts.” “Yes.” “The tape.” “Yes.” “Why?” Megan answered. “Because it’s already there.” That changed the question again. Not: Why is the directory bad? Too narrow. Not: Why do staff use unofficial lists? Better. Lena wrote: WHAT MAKES A REPRESENTATION BECOME THE DEFAULT? Jonah read it. “Availability.” Megan said, “Speed.” Samantha added, “Trust.” Carla said, “Habit.” Daniel said, “Cost.” Lena looked at the monitor. “Cost to whom?” Daniel smiled. “Good.” Jonah added, “Context.” Lena wrote them all. Then stopped. “What’s missing?” Nobody answered. Megan looked toward the medication room. “Location.” Lena nodded. PROXIMITY Samantha looked at the word. “The official directory is globally available.” “Yes.” “The tape is locally available.” “Yes.” Jonah smiled. “Closer to the decision.” Lena looked at him. “Keep it.” At 3:11, Carla asked the question. “Should we remove the lists?” “No,” Megan said immediately. Samantha said, “Eventually.” Daniel said, “Depends.” Jonah looked at Lena. Lena said nothing. Carla waited. “Well?” “Wrong decision.” Carla stared. “What is the right one?” “Why do the lists exist?” “We know why.” “No. We know several reasons.” “Fine. What do you want?” “A replacement that preserves the useful properties before removing the compensation.” Samantha nodded slowly. “Otherwise we delete the local solution and restore the friction.” “Yes.” Megan looked at Carla. “Please write that down.” Carla sighed. “I heard it.” They did not design the replacement. Not yet. That was becoming a habit. Instead, Samantha took a bounded task: Compare official directory terminology against common unit terminology. Megan took another: Identify which taped numbers were still actively used. Carla: Determine whether unofficial contact lists appeared in any safety or quality guidance. Daniel: Nothing. He looked relieved. Lena said, “You’re not relevant yet.” “Best meeting I’ve had all week.” Jonah laughed. At 3:42, everyone left except Lena and Jonah. The photographs remained on the screen. Four local lists. Four slightly different solutions. Jonah looked at them. “Yesterday we were asking how a system distinguishes resilience from hidden load.” “Yes.” “Is this resilience?” “Possibly.” “Hidden load?” “Also possibly.” He looked at her. “Both?” “Yes.” “That’s annoying.” “It is.” He walked to the board. WHAT MAKES A REPRESENTATION BECOME THE DEFAULT? He underlined DEFAULT. “Do people choose it?” “Sometimes.” “Does the organization?” “Sometimes.” “Does convenience?” “Sometimes.” “Habit?” “Yes.” “Trust?” “Yes.” He looked back at her. “So default doesn’t necessarily mean official.” “No.” “Or best.” “No.” “Or most accurate.” “No.” “Just the one people reach for first.” Lena stopped. She looked at the tape photo. Then at the official directory. The directory represented the organization more accurately. The tape represented the work more efficiently. Neither could replace the other completely. But only one sat inside the cabinet where the question occurred. She wrote beneath Jonah’s underline: DEFAULT = LOWEST FRICTION ACCESS TO SUFFICIENT INFORMATION Then paused. Crossed out the equals sign. Jonah laughed. “I knew it.” She replaced it. DEFAULT MAY EMERGE FROM THE LOWEST- FRICTION ACCESS TO SUFFICIENT INFORMATION Jonah nodded. “Much better.” “Obviously.” He looked at the sentence. “Not the best information.” “No.” “Enough information.” “Yes.” “Close enough to the decision.” “Yes.” “Fast enough to use.” “Yes.” “Trusted enough to act on.” Lena looked at him. “Keep all four.” He smiled. Megan returned at 4:01. She had the strip of tape in her hand. Lena stood immediately. “You removed it.” “Yes.” “Why?” “Facilities replaced the cabinet.” Lena looked at the tape. Megan held it carefully by the edges. “I remembered.” Jonah smiled. “Good.” Megan placed it on the table. The old cabinet was gone. The local interface had survived only because someone recognized it as information before disposal. Lena looked at the tape. “What did you put in the new cabinet?” Megan smiled. “A copy.” Lena stared at her. “You reproduced an unvalidated information source.” Megan’s smile disappeared. “Oh.” Jonah looked away. Lena looked at him. “What?” “Nothing.” “That was not nothing.” “No.” Megan reached for the tape. “So I should take it down?” Lena looked at the photographs. Then the tasks. Then the new question on the board. “No.” Megan stopped. “Why?” “Because we still haven’t replaced what it does.” “So it stays?” “For now.” “Even though some numbers may be wrong?” “Yes.” Megan frowned. “That feels inconsistent.” “It is constrained.” “Difference?” Lena considered. Jonah waited. “Known imperfect representation with understood limitations.” Megan looked at the tape. “Versus?” “Removing it and restoring a known operational gap.” Megan nodded slowly. “So we label the limitations.” Lena looked at her. “Yes.” Jonah smiled. Megan caught him. “What?” “Nothing.” “That definitely wasn’t nothing.” “No.” Before they left, Megan wrote on the new copy: LOCAL QUICK LIST — VERIFY IF UNCERTAIN Then dated it. Not permanent. Not official. Not invisible. Lena looked at it once. “Better.” Megan smiled. “I’ll take it.” They left the medication room. CHAPTER TEN Enough The tape had changed by the time Lena reached the medication room Monday morning. It was still there, handwriting and all, but beside it now hung a second sheet — typed, laminated, dated Friday. UNIT QUICK CONTACT LIST, it read, and beneath that: For current directory information, consult the hospital directory. “Before you say anything,” Megan said, appearing behind her, “I didn’t make it.” “Who did?” “Administration.” Jonah arrived with two coffees and looked at the cabinet. “Oh no.” “That was approximately my reaction,” Megan said. “I said approximately,” she added, before Lena could say anything. “I know what the word means.” The new list was, by any reasonable measure, better. Nine departments, every one under its official name — Patient Placement instead of Bed Management, Clinical Engineering instead of Biomed, Information Services instead of IT — every number current, every entry verified. Megan looked at it like something had been taken from her. “The translations are gone,” Jonah said. “Yes.” “But the numbers are right.” “Yes.” “And somebody’s responsible for updating it now.” “Apparently.” Carla, arriving a few minutes later, read both lists and delivered her verdict without hesitation: current numbers, a date, clear ownership, consistent terminology. Better, obviously. “Better for what?” Lena asked. Carla stopped. Jonah looked down at his coffee. They didn’t run a study. Lena’s call, and Carla looked relieved to hear it — they already had enough to ask the only question that mattered. Which list did people actually use? Megan’s answer came without a pause: the old one. She knew it. Patient Placement meant nothing to her without a half-second of translation; Bed Management meant something instantly. The new list had corrected everything that had been wrong with the old one and, in doing exactly that, quietly discarded the one thing the old one had been doing right. Samantha, walking through it with them later, put words to what had happened: they’d treated the unofficial terminology as a defect to be corrected, when it had also been serving a function nobody had named. Jonah wrote three words on the board — CORRECTION, PRESERVATION, FUNCTION — and beneath them: Correct what is wrong. Preserve what works. Identify why it works. Carla read it and called it obvious. It was. That, Lena pointed out, was exactly the problem: they’d sorted everything unofficial into the wrong-and-needs-fixing category without checking whether some of it was unofficial and load-bearing. By midmorning, someone on the floor had already started fixing the fix — small handwritten corrections appearing next to the official terms, translating them back. Megan wanted to know if they should leave it. “Yes,” Lena said. “Because it’s data. About what our representation still didn’t contain.” At noon, back in the conference room, Lena added a final line beneath the morning’s work: Enough may be defined as sufficient for the current purpose — until the system tells you otherwise. Jonah thought it sounded almost philosophical. Lena disagreed. It was operational. CHAPTER ELEVEN The Threshold By Tuesday, Daniel’s spreadsheet had grown to twenty-seven rows and he’d highlighted none of them — which Lena took as the correct call before anyone explained why. Samantha had gone back through the eight reported access issues expecting one mechanism repeating itself. She found six that traced to the cabinet, and two that didn’t: a medication filed in the wrong refrigerator bin, and a multidose item someone had relocated to a procedure room without updating anything. “Different mechanisms,” Jonah said. “Same workload,” Daniel added. That reframed the whole count. What looked like eight instances of one problem was actually several different problems that happened to produce the same symptom — nurses spending unlogged minutes solving something the system said didn’t need solving. Three were duplicate-encounter cabinet issues. One was a synchronization delay. One was a lockout. One was a stock discrepancy. One was a plain relocation nobody had documented. Grouped this way, the original question didn’t survive contact with the data: they hadn’t been measuring one hidden problem, they’d been measuring hidden compensatory work in general, which turned out to have several unrelated causes feeding into it. The conversation that followed circled a harder question without landing on it cleanly — when does a decision that makes complete sense to the person making it start costing the system something nobody accounted for? Late in the day, after everyone else had gone, Jonah opened his notebook and wrote it out: When does locally rational action create systemic load? Lena looked at it, took the pen, and inserted one word: When can locally rational action create systemic load? “Can,” Jonah said. “Not does.” “No.” “Because sometimes local rationality produces good system outcomes.” “Yes.” He closed the notebook. Neither of them wrote an answer beneath it. There wasn’t one yet. CHAPTER TWELVE The Distinction Samantha’s test results were already on the screen when Lena arrived Wednesday morning: ACTIVE INPATIENT and ACTIVE OBS, the two new encounter labels, tested against fifty simulated retrievals across five users. Forty-eight correct. Two wrong. Every participant said the new labels felt clearer than the old ones. “So it worked,” Megan said. Lena didn’t answer right away, and Megan read the pause correctly. “Mostly?” Carla, arriving a few minutes later, called ninety-six percent good. No one argued, exactly, but no one agreed either — because good wasn’t the question on the table. Sufficient was, and yesterday’s line was still on the board to prove they’d been here before: Enough may be defined as sufficient for the current purpose. Carla closed her eyes. “I hate continuity.” Vasquez wanted to know what kind of errors the two misses were. Both users, it turned out, had selected ACTIVE OBS when they meant ACTIVE INPATIENT — because both labels started with the same word, and under time pressure, active was the part that registered first. The distinction the label was supposed to carry lived in the second word, and the second word was exactly what a rushed reader skipped. That single failure mode did more work than the ninety-six percent had. It reframed the afternoon’s question away from whether the new labels were an improvement — they clearly were — and toward something narrower and less comfortable: what, specifically, needed to survive every time this information changed hands, so the next person downstream could still trust what they were looking at? Late in the day, Lena put it on the board as a question, deliberately not an answer: What must survive a handoff for the next representation to remain valid? Jonah ran through candidates out loud — data, context, meaning, intent, authority, timing — and got maybe for all six. “You’re enjoying this,” he said. “No.” “You are.” She capped the marker instead of answering. Megan, already at the door, asked if tomorrow they’d figure out what had to survive. Lena shook her head. First they’d need to determine whether survive was even the right model for what they were looking at. By Thursday, Jonah had already revised the question waiting on the board — adding OR beneath it, and under that: What must still be available? Not the same question, even though it looked close enough that Carla, arriving with coffee, initially read it as one more turn of the same screw and said so. It wasn’t. Surviving a handoff and remaining available afterward were different guarantees, and the gap between them was exactly where a call from a confused physician had landed that morning — asking about an encounter number that technically still existed in the record but that nobody could locate quickly enough when it mattered. Samantha, reading both versions of the question, thought the second one might actually be better: not everything needs to travel with the representation. Context usually doesn’t. What has to travel is narrower — enough to tell the next person what the thing in front of them is supposed to mean. That distinction mattered because two failure modes had turned out to look identical from a distance but weren’t: one document had preserved a rule accurately but written it too vaguely to act on; another had kept a perfectly usable condition that nobody downstream had actually followed. Neither was working, but they were failing for opposite reasons — one hadn’t survived the handoff in a usable form, the other had survived fine and simply hadn’t been read. Before noon, they walked back past the medication room, where the revised contact list carried a new handwritten addition at the bottom, in Megan’s writing: Review when unit directory changes. She’d added it herself, unprompted, because the list kept changing under them and someone needed a trigger for checking whether it still matched. Lena told her to leave it. CHAPTER THIRTEEN The Difference Marcus arrived Friday with results, which alone was enough to make the room suspicious. He’d tested two configurations against identical patient scenarios: current production behavior, and the same behavior with the old translation rule — the one still prioritizing whichever encounter came first — removed entirely. Four of five test cases came back unchanged. The fifth was the one that actually generated the complaint calls: with the rule removed, the cabinet displayed the inpatient encounter first, correctly; with it left in place, observation still came first, wrong. “So remove it,” Megan said. Marcus wasn’t ready to. Not because the rule was doing anything useful anymore — the data said plainly that it wasn’t — but because four-fifths of its footprint was neutral, and pulling a rule that old, still active in production, still referenced by systems nobody had fully mapped, deserved more than one clean test result before anyone touched it live. Later that afternoon, alone with Jonah, Lena found the sharper version of the same idea sitting in a different chart entirely — a before-and-after ticket-volume graph Daniel had put together, same rough total both times, completely different categories underneath. “Representation can fail by losing something,” Jonah said, working it out loud. “Or by keeping too much together.” Lena stopped writing and made him say it again. It was the cleaner statement of what the translation-rule test and the aggregate graph were both pointing at from different directions — a total can stay stable while hiding a real shift in what’s actually happening beneath it, and a rule can look completely inert in aggregate while still being the entire explanation for one specific, painful case. She didn’t put it on the board. When Jonah asked where it was going instead, she tapped her notebook. Private, for now. CHAPTER FOURTEEN The Aggregate Daniel brought one number Monday morning: 2.7 hours, the medication-access workload across twenty-three separate incidents, most resolved without ever becoming a formal event. Duplicate encounters, synchronization delays, stock confusion, pharmacy clarification calls, a handful that didn’t sort cleanly into any category at all. “So 2.7 hours is not one thing,” Jonah said. “No. It’s one total.” Daniel’s breakdown split the load by department — nursing carrying most of it, pharmacy less, informatics almost nothing — and Megan objected immediately: wrong. Informatics had spent half an hour with them just Friday, time that didn’t appear anywhere in the capture because the capture hadn’t been built to catch it. Not wrong, Lena corrected. Incomplete. The number was real and the number was still missing part of the picture, and those were different problems requiring different fixes. When Daniel relayed that Finance wanted the 2.7 hours dropped straight into the weekly report, Lena didn’t ask for the category breakdown to go with it. Include it, she wrote back. Add: 23 events, single-unit sample, observed only. Jonah noticed what she’d left out. “Enough for the current purpose,” he said. She didn’t disagree. That same stretch produced the strangest footnote of the week: a cafeteria cash-drawer discrepancy that Megan kept bringing back to the table — three forty-dollar drops instead of the expected two, one of them briefly, incorrectly reversed as a suspected duplicate before turning out to be a separate, valid transaction that happened to land at the same timestamp as another. Once resolved, Megan’s reconciliation note ran four full sentences of careful, specific language — duplicate drop initially suspected, reversal temporarily reconciled register balance, subsequent review identified both drops as valid and separate — rather than the two words that would have been faster to write and wrong. Jonah pointed out the report was longer than “duplicate drop corrected.” It was. That was the whole point: the short version was cheaper and false, and the only way to be both correct and brief was to already know the answer before you wrote the sentence. CHAPTER FIFTEEN The Baseline By Wednesday, every number connected to the removed translation rule looked better. Wrong-encounter-first tickets had dropped from six to zero. Pharmacy callbacks tied to duplicate encounters had fallen from eleven to four. Compensatory minutes were down from ninety-six to thirty-one. Carla approved on sight; Daniel, less convinced by his own good news than anyone else in the room, asked whether they could finally just say the system had improved. “Compared to what?” Lena asked. It wasn’t rhetorical. Samantha’s monitoring data was clean, the before-and-after numbers were real, and none of that answered whether the comparison itself was fair — whether the old measurement method and the new one had ever been counting the same thing. That question got sharper a day later, when Samantha laid two weeks of direct observation against two weeks of the old ticket-based estimate for the same period, same unit, same underlying access issues. The two methods didn’t move together. Sometimes they were close. Sometimes the old estimate ran double the observed figure, because a ticket could stay open logging elapsed time long after the actual nursing work behind it had already finished — the measurement kept moving after the thing it was measuring had stopped. There was no clean conversion factor available between the two methods, and building an average one anyway, Daniel admitted, would only manufacture a confidence the underlying relationship didn’t support. When Finance asked whether three years of historical numbers should be restated to match the new method, Lena’s answer was no — they didn’t have enough to reconstruct the old periods accurately. Instead: mark where the definition changed, document the current method, and leave the old numbers exactly as they’d been recorded. “Repair, not rewrite,” Jonah said. “Exactly.” The harder question surfaced almost as an aside, late in the day, when Jonah noticed that today’s wording had shifted slightly from yesterday’s without anyone flagging it — and if they hadn’t happened to check back, they’d never have known. Not every change needed a marker; marking everything would bury the markers that mattered under the ones that didn’t. The line Lena settled on was narrower and, Jonah thought, annoyingly practical: a change earns a marker when it affects what someone downstream could reasonably conclude from the number. Nothing more, nothing less. CHAPTER SIXTEEN The Boundary At 11:44, everyone left except Lena and Jonah. The board held: BETTER FOR WHOM? BETTER AT WHAT? Jonah looked at the two directory versions. “The old one wasn’t wrong.” “No.” “The new one wasn’t wrong.” “No.” “They represented different things.” “Partly.” “And the problem happened because people treated one representation as though it represented both.” “Yes.” He nodded. “There it is.” Lena looked at him. “What?” “Nothing.” “That wasn’t nothing.” “No.” He picked up the marker. She watched. He wrote beneath the two questions: WHERE DOES THE REPRESENTATION STOP? Lena stared at it. Longer than usual. Jonah waited. “Too much?” “No.” “Too early?” She considered. “No.” He smiled. “Keep it?” “Yes.” They packed up. At the door, Jonah stopped. “Something’s bothering me.” “What?” “The contact list.” “What about it?” “The handwritten review note worked.” “Yes.” “Because the directory changed.” “Yes.” “But the contact list didn’t need to.” “No.” “So review didn’t mean revise.” Lena looked at him. “No.” He nodded. “Good.” “Why?” “Because I think we’ve been treating attention like it has to end in intervention.” She looked back at the board. WHERE DOES THE REPRESENTATION STOP? Then at the directory on the screen. Reviewed. Still valid. CHAPTER SEVENTEEN The Handoff At 8:14 Monday morning, the room was empty. Not temporarily. Empty. The monitor was off. No coffee. No reports. No receipts. No policies. No one waiting for Lena to ask why. Only the board remained. BETTER FOR WHOM? BETTER AT WHAT? WHERE DOES THE REPRESENTATION STOP? Jonah entered behind her. “That’s unsettling.” “What?” “The room.” Lena sat. Jonah looked around. “Nothing is wrong.” “No.” “No one brought us anything.” “No.” “No mystery email?” “Not yet.” He checked his phone. “Nothing.” Lena opened her notebook. Jonah looked at her. “What do we do?” “Work.” “On what?” She looked at him. “The report.” His expression changed. “Oh.” At 8:21, Samantha arrived. She carried nothing. “Morning.” Jonah looked at her hands. “What?” “Nothing.” “That wasn’t nothing.” “No.” She sat. “Where’s everyone else?” “Working,” Lena said. Samantha nodded. Then: “Oh.” Carla arrived at 8:27. No folder. Daniel joined at 8:31. No spreadsheet. Marcus appeared on the monitor at 8:34. Megan came in last. She looked around. “Why does this feel like somebody died?” Carla pointed at her. “Thank you.” Lena looked around the table. “We’re done.” Silence. Megan sat. “With what?” “The review.” Another silence. Carla recovered first. “Done done?” “Yes.” “No follow-up committee?” “No.” “Working group?” “No.” “Phase two?” “No.” Carla leaned back. “I don’t trust this.” Daniel looked at Lena. “What does done mean?” Jonah smiled. Lena looked at him. “Don’t.” “I didn’t say anything.” “You were going to.” “Yes.” Samantha opened her laptop. “What goes in the report?” Lena looked at her. “What happened.” “All of it?” “No.” Samantha nodded. “Right.” Then she started typing. Lena noticed. She said nothing. The first draft appeared on the screen. Investigation initiated following recurring medication- access exceptions associated with duplicate patient encounters. Megan read it. “Associated with?” Samantha nodded. “Yes.” “Not caused by?” “No. We never established that all of them were.” Marcus said, “Some had different mechanisms.” Megan nodded. “Keep it.” Lena watched. No intervention required. The next section described the obsolete translation rule. Created during an upgrade. Reasonable at the time. Intended to remain until a vendor patch was validated. Patch later validated. Rule remained. Its continued presence contributed to wrong-encounter-first behavior. Daniel read the paragraph. “Bounded testing?” “Yes,” Samantha said. “Not comprehensive?” “No.” “Good.” Carla looked at him. “You’re grading again.” “I’m checking the claim.” Carla nodded. “Carry on.” Jonah looked at Lena. She had noticed too. They continued. The cabinet contact list. Operational language versus official terminology. The directory. Administrative ownership versus operational contact. Historical workload estimates. Direct observation. Definition changes. Trend discontinuities. The old alert pilot. Samantha paused. “The staffing policy?” Carla shook her head. “Outside scope.” Samantha deleted it. Megan looked at Lena. Lena hadn’t moved. At 8:58, Megan read the draft from the beginning. “The cash drawer isn’t here.” Jonah looked at her. “Should it be?” Megan thought. “No.” “Why?” “It changed how we approached the discrepancy, but it wasn’t part of the medication-access review.” Daniel nodded. “Then it doesn’t support a finding here.” Megan looked at Samantha. “Leave it out.” Samantha did. Jonah glanced toward the board. Then back at the group. He did not reach for the marker. At 9:11, Samantha reached the findings section. The heading read: ROOT CAUSES Carla pointed. “No.” Samantha looked at it. Then laughed. “Right.” She changed it herself. FINDINGS Marcus nodded. “Better.” Megan asked, “Because they don’t all share one cause?” “And some aren’t causes at all,” Samantha said. “Some are conditions,” Daniel added. “Some are mismatches,” Marcus said. Carla looked around. “Good. We’re all irritating now.” There were six findings. Obsolete interface rule. Ambiguous terminology. Historical measure mismatch. Administrative and operational ownership conflation. Incomplete maintenance trigger. Aggregate measures obscuring changes in failure composition. Daniel looked at the list. “Should we rank them?” Carla asked: “By what?” He stopped. She smiled. “Exactly.” Daniel looked at Lena. “Was that yours?” “No,” Carla said. “It’s mine now.” Vasquez arrived at 9:26. She looked at the screen. “What did I miss?” “We’re done,” Megan said. “With what?” Everyone laughed. Vasquez looked confused. Carla waved her in. “Sit down. Apparently we’re being released into the wild.” Lena looked at her. “No one said that.” “Yet.” Vasquez read the report. She stopped at: No patient harm was identified during the review period. Then the next sentence: Absence of identified harm should not be interpreted as evidence that the observed representational failures were without operational consequence. She nodded. “Good distinction.” Daniel said, “We argued about that one.” “We did not,” Samantha said. “We discussed it.” Carla looked at Vasquez. “You see what’s happened?” “Yes.” “I’m surrounded.” The recommendations came next. There were eleven. Carla stared at the screen. “No.” Samantha laughed. “I knew you’d say that.” “Then why are there eleven?” “Because I wanted to see which ones survived.” Jonah smiled. Carla noticed. “What?” “Nothing.” They went through them. Maintain operational terminology beside official terminology. Preserve definition-change markers. Document temporary-rule review conditions. Separate administrative ownership from operational contact where relevant. Continue bounded monitoring of the removed interface rule. Standardize representation requirements across hospital operational systems. Megan stopped. “No.” Everyone looked at her. “That last one.” Samantha nodded. “Why?” “We looked at one workflow.” Daniel added, “Several connected systems, but still one bounded review.” Marcus nodded. “We don’t have evidence for an enterprise standard.” Carla smiled. “Delete it.” Samantha did. Lena watched. Three more recommendations disappeared. One proposed a new review committee. Carla deleted it before Samantha finished reading. “No.” Daniel laughed. “You didn’t even let her finish.” “I knew where it was going.” Another proposed mandatory periodic review of every temporary operational rule. Vasquez shook her head. “That could create more work than the rules themselves.” Deleted. Another proposed standardizing all operational directory structures. Marcus stopped them. “Different directories serve different purposes.” Deleted. The remaining recommendations attached directly to observed findings. Nothing more. Megan looked at the shortened list. “That’s better.” Daniel asked: “Better at what?” She looked at him. “Supporting what we actually found without pretending we audited the hospital.” Daniel nodded. Jonah looked at Lena. Still nothing. At 10:02, Daniel asked: “Who gets the report?” Carla answered. “The board.” Vasquez said, “Clinical leadership.” Marcus said, “Information Services.” Megan said, “Nursing.” Samantha looked at the report. “Same findings.” Daniel nodded. “Different detail.” Marcus said, “IT needs the technical appendix.” “The board doesn’t,” Carla said. Vasquez added, “Clinical leadership needs operational consequence.” “Finance needs the measurement methodology,” Daniel said. Megan shrugged. “Nursing needs to know what changed and who to call.” Samantha nodded. “One report. Different entry points.” Jonah looked at Lena. She met his eyes. Neither said anything. Carla frowned. “We are not writing four reports.” “No,” Samantha said. “Good.” “Core report. Appendices.” Daniel nodded. “Executive summary for the board.” “Technical mapping for Informatics.” “Measurement notes for Finance.” “Operational changes for Nursing and Pharmacy.” Marcus said, “Same evidence underneath.” Samantha nodded. “Exactly.” Lena closed her notebook. Jonah noticed. Nobody else did. At 10:24, Samantha opened the executive summary. One sentence read: The review identified systemic failures in medication- access representation. Vasquez frowned. “Systemic?” Samantha looked at it. “Too broad.” Daniel nodded. “We didn’t establish hospital-wide prevalence.” Marcus said, “And ‘failures’ treats unlike findings as equivalent.” Samantha deleted the sentence. Typed: The review identified recurring representational mismatches across several components of the medication- access workflow. Megan read it. “That says what we actually found.” “Keep it,” Carla said. Jonah leaned toward Lena. Quietly: “You haven’t said anything in twenty minutes.” “Twenty-three.” He smiled. At 10:41, they reached the conclusion. Blank. Megan looked at Samantha. “You forgot it.” “No.” “You left it blank?” “Yes.” “Why?” “Every version I wrote claimed more than the evidence supported.” Lena looked up. Samantha saw her. “Don’t.” Lena said nothing. Samantha smiled. Daniel looked at the blank section. “We still need to tell the board what the review established.” Carla nodded. “That’s the question.” Jonah looked at her. “What can the report actually support?” Carla pointed at him. “That.” Nobody moved toward the board. They answered from their seats. Marcus went first. “The obsolete translation rule was no longer required for its original function.” Samantha added: “Removing it improved observed encounter-selection behavior during the review period.” Megan: “Without the original missing-observation problem coming back.” Daniel: “Historical workload estimates can’t be directly compared with current observed workload.” Carla: “Administrative ownership and operational contact are different things and need to remain distinct where the workflow depends on them.” Vasquez: “No patient harm was identified.” Samantha looked around. “And we don’t have evidence to generalize beyond the reviewed workflow.” Silence. Carla looked at Lena. “Well?” Lena looked around the table. “Write it.” Samantha did. The final paragraph read: Within the reviewed medication-access workflow, several recurring operational discrepancies were traced to differences between what information was recorded, how it was presented, and what users reasonably understood that presentation to mean. Targeted corrections reduced the observed discrepancies during the review period. Historical comparisons remain limited by changes in measurement and definition. No patient harm was identified. Findings should not be generalized beyond the reviewed workflow without additional evidence. Samantha read it once. Then again. “Anyone?” Marcus shook his head. Daniel said, “Supported.” Vasquez nodded. “Accurate.” Megan said, “Readable.” Carla looked at the recommendations. “Bounded.” Jonah looked at Lena. She nodded. “Done.” At 11:03, Samantha saved the report. She paused. Then laughed. “What?” “The filename.” Carla looked suspicious. “What did you call it?” “Medication_Access_Review_FINAL.” Marcus groaned from the monitor. Samantha deleted it. Typed: Medication_Access_Review_2026-08 Marcus nodded. “Thank you.” Samantha hovered over Send. Then stopped. Megan looked at her. “What?” “Nothing.” “That wasn’t nothing.” Samantha smiled. “I was checking the recipients.” Carla nodded. “Good.” She checked them. Then sent it. Nothing happened. No alarm. No congratulatory message. No immediate response from the board. The room remained a room. Megan looked around. “That’s it?” Carla shrugged. “Apparently.” At 11:14, Marcus disconnected. “Call if the board has technical questions.” Samantha nodded. “We will.” He disappeared. Vasquez stood. “If Clinical asks about the patient-safety language?” Megan answered. “We have the underlying cases.” Vasquez nodded. “Good.” She left. Daniel closed his laptop. “Finance will probably ask about the historical comparison.” Samantha said: “The methodology appendix explains it.” “And if they still ask?” Carla looked at him. “Then we answer.” Daniel smiled. He left. Megan stood. “What about Nursing?” Samantha looked at her. “You are Nursing.” Megan stopped. “Oh.” Jonah laughed. Megan pointed at him. “Don’t.” She picked up her coffee. “I’ll take the operational summary back.” Carla nodded. “And if they find something wrong?” Megan shrugged. “We check it.” She looked at Lena. “Right?” Lena smiled slightly. “Right.” Megan left. Samantha packed her laptop. “I’ll track board questions.” Carla looked at her. “Why you?” “I sent the report.” “Fair.” “And I’ll route technical questions to Marcus, clinical questions to Vasquez, measurement questions to Daniel.” Carla nodded. “And questions that don’t fit?” Samantha thought. “We figure out what they’re actually asking.” Jonah looked at Lena. She had heard it. Samantha left. Carla remained. She looked at the empty chairs. Then at Lena and Jonah. “What happens when the board has another question?” Lena looked at her. “You answer it.” Carla waited. “We do?” “Yes.” “What if we don’t know?” “Find out.” “What if we disagree?” Jonah smiled. “Good.” Carla looked at him. “Good?” “Yes.” “Why?” He glanced at Lena. Then back at Carla. “Because now somebody will ask why.” Carla considered that. Then nodded. “Fair.” She picked up her folder. Stopped. “There isn’t another review scheduled.” “No,” Lena said. “Should there be?” “Do you need one?” Carla looked through the glass. Samantha was talking to Megan at the nurses’ station. Daniel had stopped farther down the corridor and was showing Vasquez something on his laptop. On the monitor, Marcus’s disconnected window had disappeared. Carla looked back. “No.” Lena nodded. “Then don’t schedule one.” Carla smiled. “That’s going to make somebody uncomfortable.” “Probably.” She left. The room was quiet. Jonah looked at Lena. “So.” “So.” “They don’t need us.” “No.” He said it again, differently. “They don’t need us.” Lena closed her notebook. “No.” Jonah looked at the board. BETTER FOR WHOM? BETTER AT WHAT? WHERE DOES THE REPRESENTATION STOP? He walked toward it. Picked up the eraser. Then stopped. “Leave it?” he asked. Lena looked at the questions. “No.” He erased the first. Then the second. Then the third. The board was blank. Jonah stepped back. “That feels wrong.” “Why?” “We’ve had a question up there for weeks.” “Yes.” “And now we don’t.” Lena picked up her bag. “They do.” Jonah looked through the glass. Megan had joined Samantha. Daniel was gone. Carla had stopped beside them. Three people talking. One pointed toward the report on Samantha’s screen. Another shook her head. Megan said something. Carla laughed. Then Samantha turned the laptop around. They kept talking. No one looked toward the room. Jonah smiled. “When did that happen?” “What?” “That.” Lena looked through the glass. “They stopped waiting for us.” Jonah nodded. “I noticed.” “When?” “I’m not sure.” “Good.” He looked at her. “Why good?” “Because it probably wasn’t one moment.” They left the room. At the nurses’ station, Samantha looked up. “The board already replied.” Carla sighed. “Of course they did.” “What do they want?” Jonah asked. Samantha read the message. “They want clarification on why the historical workload trend can’t be used to quantify improvement.” Daniel was already gone. Samantha looked at Carla. Carla looked at Megan. Megan looked at the report. Then Samantha said: “Different measurement methods.” Carla nodded. “And definition changes.” Megan added: “Tell them the operational improvement is supported by the bounded post-change observations, not the three-year trend.” Samantha typed. Jonah waited. Nobody looked at him. Nobody looked at Lena. Samantha read the reply aloud. Carla changed one word. Megan nodded. Samantha sent it. “Done.” Lena kept walking. Jonah followed. Neither said anything until they reached the elevators. He pressed the button. Behind them, the conversation at the nurses’ station continued. A disagreement now. Carla thought the board needed the definition markers attached. Samantha thought the methodology note was enough. Megan wanted both. Jonah looked back. “Should we—” “No.” He stopped. Listened for another second. Then smiled. “Right.” The elevator arrived. They stepped inside. Before the doors closed, Jonah looked back once more. “They’re going to come to a different conclusion eventually.” “Yes.” “Maybe one we wouldn’t.” “Yes.” He looked at Lena. “And that’s good?” The doors started closing. Lena considered the question. “They’ll have to support it.” Jonah smiled. The doors closed. Upstairs, Samantha’s reply reached the board. Down the hall, Marcus received a question from Megan about an interface field and sent back the documentation instead of an answer. Daniel reopened the historical graph and added the definition markers to the appendix. Vasquez disagreed with one phrase in the clinical summary. Carla agreed with her. Megan didn’t. They changed it twice. Then changed it back. No one called Lena. No one called Jonah. Nothing was finished. It didn’t need to be.