Nine Decision Points | Case 6 episode 3

 The teaching episode. Nine specific moments across this case where a different decision could have changed the outcome — from the PCP's office to the ED triage to the radiologist's read. Each one a small choice. Each one part of a chain. The episode every program director, residency, and quality committee will want to send to their teams.  

Transcript

SPEAKER_01 1:34

This podcast is for educational and informational purposes only. It is not medical advice, legal advice, or professional advice of any kind. And it does not create an attorney client, physician patient, or other professional relationship. Any discussion of cases, outcomes, standards of care, statutes, or regulations is general in nature, may be incomplete, and may vary by jurisdiction and over time. Listeners should consult their own attorney, malpractice carrier, risk manager, or qualified clinician for advice on any specific matter. Any case discussions are based on public records, de-identified information, or educational reconstruction. The views expressed are those of the speakers and do not necessarily reflect the views of any employer, hospital, health system, insurer, or affiliated organization.

SPEAKER_02 2:41

In aviation safety, the question after an adverse event is never simply who made the mistake. It is where in the chain of events could a different decision, a different system, or a different process have altered the trajectory. That is the framework for this episode, and that model was one of my inspirations for developing the charted defense. Over the past two episodes, we walked through the clinical timeline and the legal proceedings in this case. A 70-year-old man with an evolving posterior circulation stroke moved from a primary care office to a hospital emergency department, where he remained for over two hours before a physician evaluation. The hospital settled. The trial against the PCP ended in directed verdict when the causation element was not established on the record. Today, we step back from the chronology and the courtroom and look at the decision points. Nine of them across four layers: the PCP encounter, the hospital triage system, the scientific standard of care, and the courtroom itself. Each one offers something a clinician can carry into practice. And I want to note at the outset: the courtroom is not a clinical layer, but it is not immune to human factors error. The same analytical framework that helps us understand clinical decision making applies to the legal process too. Understanding how things can go wrong in that system is part of being prepared for it. Let me start with the primary care encounter, because this is where most of us live professionally managing acute presentations in settings that are not built for emergencies. And I want to frame this with empathy for the clinical situation before we get into the teaching points. He is dizzy, has a headache, and his gait is unsteady. He has a history of hypertension, coronary artery disease, remote craniotomy, and a seizure disorder. He tells you he took a medication from South America that you do not recognize. He also just had a stressful phone call. That is a broad differential: hypertensive encephalopathy, intracranial pathology, medication reaction, stroke, and several other possibilities. I have seen this presentation, most of us have. The challenge is not that stroke should be obvious. The challenge is that posterior circulation strokes present with exactly the kind of nonspecific symptoms that overlap with a half dozen benign conditions: dizziness, headache, ataxia. Those could be vestibular, metabolic, hypertensive, or vascular. The early lateralizing deficits that make anterior strokes recognizable may not be present. So I am not standing here saying this PCP should have diagnosed a posterior circulation stroke from the information available to him. What I am saying is that there are four decision points in this encounter that carry teaching value regardless of what the correct diagnosis turned out to be. And all four of them apply to any clinician managing an acute presentation where the differential includes a time-sensitive diagnosis.

SPEAKER_01 6:13

The first decision point is the neurological exam. The PCP documented gate, unsteady, and a standard systems exam. He testified that he performed a neurological screening and found no focal signs. He screened for weakness, facial droop, slurred speech, confusion, blurred vision, and one-sided weakness. That screening was not documented in the chart.

SPEAKER_02 6:43

This is the documentation teaching point, and it is the most important one in the episode. If you evaluate a patient with headache, dizziness, and unsteady gait, and you decide it is not a stroke, write down exactly why. Pronator drift negative, cranial nerves intact, speech clear, gait a taxic but symmetric, no lateralizing findings. That documented negative neurological screen, um, even three or four lines, uh, creates a defensible record, regardless of the outcome. Without it, the legal analysis proceeds from the silence in the chart. The PCP testified he did the screening. He may well have done it, but the chart does not show it. And in a legal proceeding, the chart is the primary evidence. I want to acknowledge why documentation gaps happen, because this is not about carelessness. In a busy office with a walk-in patient with a blood pressure of 200 over 120, with a medication you do not recognize. So the cognitive load is high. You are thinking about the differential, the treatment, the disposition. Documentation of a negative screening can feel redundant when you have already decided the screening was reassuring. But the chart is the only witness that doesn't change its story. And that is a lesson that protects the physician who does it.

SPEAKER_01 8:19

The PCP classified the situation as urgent, but not life-threatening. He testified that the patient's headache had improved, the blood pressure had dropped to 180 over 100. The patient was ambulatory, and in his clinical judgment, an ambulance was not warranted.

SPEAKER_02 8:38

The teaching point here is about what transport mode activates on the receiving end. EMS pre-notification is not just faster transport, it activates the entire stroke architecture before the patient arrives. Stroke screening by paramedics en route, hospital pre-notification, stroke team mobilization. The AHA slash ASA pre-hospital chain is the system designed to support the 60-minute door-to-needle target. When a patient arrives by taxi unaccompanied, carrying a handwritten note, they enter through the standard registration pathway. The stroke team does not know they exist until someone in the ED identifies the problem. That is a systems design difference, not a judgment on the PCP's reasoning about the patient's acuity. The PCP did not believe this was a stroke. In his clinical framework, the transport decision was consistent with his assessment. But the teaching point stands. When the differential includes a time-sensitive diagnosis, the transport mode shapes the clinical timeline that follows.

SPEAKER_01 9:49

The third decision point is the prescription note. The PCP provided a handwritten referral note listing two medications he had administered, nebivolol, bistolic, and the combination amylodopine, olmasartan, azor, along with the patient's history and his home seizure medication, phenytoin, dilentine, the note read History, seizure disorder, hypertension. 70-year-old patient has severe hypertension, blood pressure, 200 over 120, given by Stolic, 5 mg, an Azore, 5 over 40 at 5 p.m. Not taking BP pills, and Dillantine 500BID, sent to ER.

SPEAKER_02 10:40

That note conveyed blood pressure and medications. It did not convey the clinical question. It did not include the patient's presenting symptoms, the headache, the dizziness, the unsteady gait. It did not include the discharge blood pressure of 180 over 100. The prescription note was the only clinical communication between the PCP encounter and the emergency department. When you cannot be on the phone with the receiving team, ES it be um and in most office to ED transfers, you cannot, uh, the note is the communication link. And the information in that note shapes how the receiving team prioritizes the patient. BP 200 over 120 gave bestolic and Azure, sent to ER is a medication list. Presenting with headache, dizziness, and unsteady gait. Rule out stroke versus hypertensive encephalopathy, BP 200 over 120, now 180 over 100 after oral antihypertensives, please evaluate urgently is a clinical handoff. The difference matters.

SPEAKER_01 11:47

The fourth decision point is the medication choice. Nebivolol and amlotopine slash ulmasartan. Both oral, both chronic outpatient medications with delayed peak effects. Nebivolol peaks at 1.5 to 4 hours. Amlodopine peaks at 6 to 12 hours.

SPEAKER_02 12:11

The PCP treated the diagnosis he had made: hypertensive urgency. In that framework, oral antihypertensives are a standard tool. I want to be clear about that. He was not treating a stroke. He was treating what he believed was a dangerously elevated blood pressure. The teaching point is about the pharmacokinetic profile in the context of diagnostic uncertainty. Once these medications are given, they continue lowering blood pressure for hours. They are not titratable. They cannot be reversed. The EAN slash ESO prehospital stroke consensus states that pre-hospital treatment of high blood pressure in patients suspected of acute stroke is not recommended. The AHA slash ASA lists only titratable IV agents for blood pressure management in the setting of acute stroke treatment. The reason is that the ischemic penumbra, if the brain tissue at risk, pull up depends on perfusion pressure to survive. If the working diagnosis turns out to be something other than hypertensive urgency, the pharmacokinetic effects of these medications cannot be undone. In this case, the medications reached peak effect at approximately 8.09 p.m. when the blood pressure was recorded at 119 over 94 during what was later confirmed as an active ischemic stroke. The neurology team immediately implemented permissive hypertension to 200 over 100, the opposite physiological direction. That clinical contrast illustrates why the guideline framework reserves acute blood pressure management to titratable IV agents in settings where the differential includes ischemic stroke. I want to frame all four of these as practical considerations for any clinician managing an acute presentation where the differential includes a time-sensitive diagnosis. They are not criticisms of this PCP's clinical judgment, which was based on the presentation he observed and the assessment he made. They are decision points where a different choice, documenting the screening, calling 911, including symptoms in the referral note considering the pharmacokinetic timeline, could have altered the trajectory that followed. The second layer is the hospital triage system. And I want to approach this as a systems design question, not an assignment of individual blame, because that is where the teaching value is. Three things happened in the emergency department between 6.19 p.m. and 8.03 p.m., a span of 104 minutes.

SPEAKER_01 14:56

First, the initial triage. The sorter nurse registered the patient at 6.19 p.m. The chart reflects that no chief complaint was entered, no vital signs were taken, and no assessment note was handed to the registration clerk. At 6 53 p.m., the nurse transported the patient to a staging area for urgent triage, still with no clinical data in the EMR. The system showed awaiting triage with no information.

SPEAKER_02 15:30

This hospital used a manual triage handoff. The sorter nurse was expected to write the patient's complaints on paper and hand it to the registration clerk to initiate the chart. When that handoff did not occur, the downstream clinical team had no information to act on. The ED physicians were looking at the EMR. The EMR showed nothing. They had no awareness that the patient was there, much less what he presented with. Under the ESI triage framework, a patient presenting with symptoms suggestive of acute stroke would be classified as ESI2, high risk with the expectation of rapid evaluation. The AHA slash ASA benchmark set door-to-physician contact at 10 minutes or less. The recorded door-to-physician contact in this case was 124 minutes. That gap was not caused by a single person's error. It was produced by a system that depended on a manual handoff. One person writing on paper and handing it to another person. And when that step did not happen, there was no backup mechanism.

SPEAKER_01 16:39

Second, the undocumented encounter. At 7 p.m., the sorter nurse found the patient on the floor. At 7.05 p.m., an emergency room physician encountered the patient. The chart contains a single line entered by the sorter nurse. Patient was examined by the attendee. No clinical findings. No vital signs. No separate physician note.

SPEAKER_02 17:07

A physician was physically present at the side of a patient who had just been found on the floor. A patient who had arrived carrying a note stating blood pressure 200 over 120. And the record from that encounter contains no clinical findings. We do not know what that physician observed. They were not deposed. They did not create a note. The teaching point is that any patient encounter, but even an informal one, even a brief one, benefits from documentation of findings. A chart entry that says a physician examined a patient without any recorded observations does not create a usable record for anyone downstream. If that physician had noted alert, ambulatory, no acute distress, or conversely, a taxic facial asymmetry noted, stroke team notified, the clinical picture at 7.05 p.m. would be part of the record. Instead, it is a gap.

SPEAKER_01 18:11

Third, the unanswered triage call. At 7 38 p.m., a triage nurse called the patient's name. The chart reads, patient called to triage. No response. The nurse did not physically walk to the waiting area to check on the patient.

SPEAKER_02 18:30

At that point, by the patient's own testimony, he was unable to move his legs, arms, or fingers. His son later described finding him kind of blacked out and fell over. A patient who was referred from a PCP with a blood pressure of 200 over 120 did not respond when his name was called. And the chart documents no response without a physical follow-up. This is the systems design question I want to leave with you. Should triage protocols include a physical check step when a referred patient does not respond to their name? Not because one nurse should have done something differently, but because the process itself should have a built-in step that catches this situation. The ED attending later testified his first contact with the patient was at 8.04 p.m. He testified he was not aware of the hospital's stroke triage protocols. When asked if the hospital had written protocols for patients presenting with stroke symptoms, he said, not that I am aware of. That answer raises a separate institutional question. How stroke alert processes are communicated to the clinical team. If the attending physician in the emergency department is not aware of the stroke protocols, the protocols cannot function as designed, regardless of how well they are written. The third layer is the TPA eligibility question. And I present this not as a relitigated argument, but as a teaching tool about how guideline language, regulatory decisions, and clinical practice can diverge.

SPEAKER_00 20:26

Whether that history disqualified him from TPA was the question neither side in the litigation engaged with fully.

SPEAKER_02 20:33

Here is the guideline evolution compressed. The 2013 AHA slash ASA guideline, the standard in effect on the date of care, Altul, listed history of previous intracranial hemorrhage as an absolute exclusion from TPA. No time frame, no etiology distinction. A hemorrhage from 30 years ago was treated identically to one from 30 days ago. A traumatic hemorrhage was treated identically to a spontaneous one. That language was inherited directly from the 1995 NINDS trial eligibility form and copied forward through every subsequent guideline without modification. In February 2015, it was six months before this patient's stroke, the FDA revised the Active ASE prescribing label and removed history of intracranial hemorrhage from the contraindication section entirely. The post-2015 label retained only current intracranial hemorrhage and subarachnoid hemorrhage. The AHA slash ASA did not follow the FDA. In 2016, the panel acknowledged, and I want to use their exact words, the literature offers only a handful of cases, and the lack of any data on this issue is possibly the reason the revised FDA label removed any history of ICH as a contraindication. Despite that concession, they retained the exclusion as class three, harm, or potentially harmful, on precautionary grounds. The 2018 and 2019 updates maintained the same position. Class III, harm level CEO. Expert opinion. And here is the detail that matters for a standard of care arguments. Prior intracranial hemorrhage remained the only neurological contraindication in the exclusion table without a time-bound threshold. Prior ischemic stroke had a three-month window, severe head trauma had a three-month window, intracranial surgery had a three-month window, prior intracranial hemorrhage had none. The guideline treated a 30-year-old traumatic hemorrhage identically to one from last month.

SPEAKER_00 22:50

For the legal context, a clinician in August 2015 was working under two authoritative but non-identical documents. Standard of care arguments in malpractice cases typically reference the guideline rather than the FDA label. But the divergence between the two creates analytical complexity that neither side in this case engaged with thoroughly.

SPEAKER_02 23:28

The plaintiff's expert cited only a 2009 advisory rather than the 2013 guideline, which contained the stronger class one recommendation for the 4.5-hour window. The defense experts focused on the time window and did not need to defend the prior ICH position in depth. Neither side presented the guideline evolution to the jury in a way that engaged the scientific complexity. And this may have been intentional, at least on the plaintiff's side, if they understood that a history of intracranial hemorrhage was an absolute contraindication under the 2013 guideline, then the causation argument through TPA becomes extremely difficult to make. The delay in presentation did not affect whether this patient received TPA under the applicable guideline, he was never a candidate. If he had presented earlier or been triaged more efficiently, the outcome regarding TPA administration would not have changed. That does not mean the plaintiff had nothing left to argue. If they had focused on the general delay in diagnosis, the delay in stroke recognition, the delay in initial screening and treatment, that would have been a different argument requiring a different strategy from both sides. But even a shift in strategy doesn't guarantee that causation can be proved. This is a framework we have discussed in earlier episodes. A plaintiff must establish four elements in a malpractice case: duty, breach, causation, and damages. For the sake of this discussion, let us assume the plaintiff met the threshold to show a breach of the standard of care. The next hurdle is showing that the breach caused the injury, in this case, the patient's disability. The plaintiff's expert focused his causation theory on TPA, but the patient was not a candidate for TPA under the applicable guideline. There would have been no change in outcome regarding TPA administration in any scenario that involved earlier presentation. There is no reasonable causation link through that pathway. If the plaintiff's experts had focused instead on the general delay in diagnosis and the delay in initial screening and treatment, there could have been a stronger argument for causation. But even then, it is not guaranteed. For teaching purposes, the 2026 AHA slash ASA guideline, published earlier this year, moved prior intracranial hemorrhage from an absolute contraindication to a relative one. The decision to offer thrombolysis in patients with a remote history of intracranial hemorrhage is now individualized based on the etiology time since the event and the patient's risk-benefit profile. Patients with ICH or due to modifiable conditions like hypertension or coagulopathy may have greater treatment benefit than risk. That does not change the analysis for this case. The 2013 guideline was the standard in effect, but it illustrates something every clinician should understand. Guideline language is not always supported by the evidence base it claims. The blanket prior AECH exclusion was precautionary, not data-driven. The AHAA's own panel said so in 2016. It took another decade for the classification to change. Knowing where the evidence is strong and where it is precautionary, that distinction determines how a clinical decision is evaluated in litigation.

SPEAKER_00 27:03

The final layer is the courtroom. The trial against the PCP ended in a directed verdict. The case was dismissed because the causation element was not established on the evidentiary record to explain what that means. In this state's medical malpractice law, the plaintiff must prove through expert testimony that the defendant departed from the accepted standard of care and that the departure was a proximate cause, a substantial factor in bringing about the injury. Both elements are required. During the plaintiff's experts' direct examination, two departure questions were asked and answered affirmatively. That the PCP departed by not performing a comprehensive neurological exam, and that he departed by not calling an ambulance. And then the questioning stopped. The follow-up. Doctor, do you have an opinion as to whether those departures were a substantial factor in causing the patient's injuries? Was never asked. The judge identified this gap at transcript page 334. The defense moved for a directed verdict under CPLR 4401, arguing that the plaintiff had not presented sufficient evidence for a reasonable jury to find in their favor. The court agreed the prima facci case was incomplete. Without the causation element on the record, the case could not go to the jury.

SPEAKER_02 28:28

This connects to what I was discussing a moment ago, a teaching point that goes beyond this case. The clinical questions, yeah, whether the PCP should have performed a more comprehensive exam, whether he should have called an ambulance, whether those decisions contributed to the outcome were never weighed by a jury. The departure opinions were on the record. The causation bridge was not. The directed verdict was procedurally correct on the evidentiary record as it stood. That is not a commentary on the underlying medicine. It is a statement about the legal process, about how the system has its own evidentiary requirements that operate alongside, but independently of clinical reasoning. And this is where the human factors framework applies to the courtroom. The same principles we use to analyze clinical decision chains, cognitive load, handoff failures, missed steps in a sequence, apply to the legal process. Expert testimony has a required structure. Each element must be established in a specific order. If a step is missed, the entire sequence breaks, just as a missed triage step broke the clinical sequence in the emergency department. The courtroom is a system, and systems are vulnerable to the same categories of error regardless of the domain. For physicians who may serve as expert witnesses or who may one day be defendants understanding how expert testimony must be structured, and how cross-examination tests credibility on issues that may seem peripheral to the medicine, is part of being prepared for the medico-legal process. The expert in this case had clinical reasoning that could be evaluated on its own terms. The cross-examination exposed vulnerabilities, the diabetes error, the TPA experience question, the referral pathway, the omitted craniotomy history, that were separate from the validity of the underlying medical opinions. But in a courtroom, credibility and substance are evaluated together. That is how the system works. I want to close with five concrete things you can take away from this case and use during your next shift. These are not theoretical, they come directly from the decision points in this case. One, document the negative neurological screen. If you evaluate a patient with headache, dizziness, and unsteady gait, and you decide it is not a stroke, write down exactly why. Pronator, drift negative, cranial nerves intact, speech clear, gait a taxic but symmetric. No lateralizing findings. Three or four lines, the chart is the only witness that does not change its story. Two, call 911 for suspected acute neurological events. EMS Pre-Notification is not just faster transport, it activates the entire stroke architecture before the patient arrives. Private transport delivers a patient to a registration desk. An ambulance delivers a patient to a stroke team. 3. If you write a referral note, write the clinical question. BP 200 over 120 gave bisteric and Azor sent to ER is a medication list. Presenting with headache, dizziness, an unsteady gait. Rule out stroke versus hypertensive encephalopathy. Please evaluate urgently is a clinical handoff. The note is the communication link when you cannot be on the phone. Four, if a patient does not respond to their name, physically check. No response from a patient referred with a blood pressure of 200 over 120 is not a charting event. It is a clinical event. Walk to the waiting area. Five, know your TPA exclusion criteria and know where the evidence is thin. The prior intracranial hemorrhage contraindication was maintained by guideline consensus for over two decades on what the AHA's own panel conceded was only a handful of cases. You do not need to memorize every exclusion table. But you should know which exclusions are evidence-backed and which are precautionary because that distinction determines how your decision is evaluated if the case goes to court. Beyond individual actions, there are institutional recommendations that come from this case. First, consider direct EMR entry at first patient contact. Triage models that rely on manual paper handoffs introduce a system design vulnerability. If the sorter nurse writes complaints on paper and hands it to a clerk, and that step does not happen, the downstream clinical team has no data. Direct EMR entry even at the sorter nurse level ensures clinical information is available regardless of whether one individual handoff is executed. Second, establish documentation expectations for any patient encounter. A chart entry that notes a physician examined a patient without any clinical findings does not create a usable record. Even brief, informal encounters benefit from documented findings. Alert, ambulatory, no acute distress takes 10 seconds and gives the next provider something to work with. Third, build a no-response escalation step into triage protocols. If a referred patient does not respond when called, the protocol should include a physical check step, particularly when the referral indicates an urgent presentation. This is a systems fix, not an individual one. Fourth, ensure ED physicians know the hospital's stroke alert process. In this case, the ED attending testified he was not aware of whether the hospital had written stroke protocols. If the attending doesn't know the alert process exists, the process cannot function. This is an orientation and communication question. How are alert protocols communicated to the clinical team and how is that communication verified? This case illustrates how clinical decisions, system design, and legal procedure each operate on their own logic. And how the outcome depends on all three. The PCP encounter had elements that the defense argued were appropriate. The PCP testified he screened for stroke and found no focal signs. He treated the diagnosis he made, he sent the patient to the emergency department. The absence of documentation left that screening unverifiable in the legal proceeding. But the absence of documentation does not mean the screening did not occur. The hospital's triage timeline deviated significantly from AHA slash ASA benchmarks, 124 minutes from door-to-physician contact, against a benchmark of 10 minutes or less. The settlement reflected the hospital's assessment of those claims. The trial concluded on a procedural basis. The departure questions were answered. The causation element was not established on the record. The directed verdict was procedurally correct. It was not a jury's assessment of the underlying clinical facts. I want to close with a recognition that every person involved in this case was working within the constraints they faced. The PCP managing an acute walk-in in an office setting, the nurses in a busy emergency department with a waiting room that was two-thirds full, the expert whose clinical reasoning existed alongside testimony vulnerabilities, the legal teams navigating a case with genuine scientific complexity. The teaching point is not blame. It is that at nine decision points across the clinical and legal timeline, a different action system design or procedure could have altered the trajectory. That is the value of reviewing these cases. Not to judge the people who were there, but to identify the decision points so that we can apply them going forward and document the screening. Call 911 when the differential includes a stroke. Write the clinical question in the referral note. Physically check when a patient does not respond. Know your guidelines and know where the evidence behind them is strong and where it is thin. That is what on the record exists to do. Thank you for listening. I will see you on the next shift.

SPEAKER_01 37:14

This podcast is for educational and informational purposes only. It is not medical advice, legal advice, or professional advice of any kind, and it does not create an attorney-client, physician-patient, or other professional relationship. Any discussion of cases, outcomes, standards of care, statutes, or regulations is general in nature, may be incomplete, and may vary by jurisdiction and over time. Listeners should consult their own attorney, malpractice carrier, risk manager, or qualified clinician for advice on any specific matter. Any case discussions are based on public records, de-identified information, or educational reconstruction. The views expressed are those of the speakers and do not necessarily reflect the views of any employer, hospital, health system, insurer, or affiliated organization.