A 68-Year-Old Diabetic With 'Indigestion' and a Recent GI Bleed
A 68-year-old woman with type 2 diabetes and hypertension tells triage she has had 4 hours of 'indigestion,' nausea, and a dull ache between her shoulder blades. She is pale and diaphoretic. Vitals: HR 104, BP 168/94, RR 22, SpO2 94% on room air, T 37.1 C, glucose 210 mg/dL. She denies chest pain. A 12-lead ECG shows 1 mm horizontal ST depression in V4-V6 with T-wave inversion. Her chart notes an upper GI bleed two weeks ago with hemoglobin 9.2 g/dL; she takes aspirin 81 mg daily, metformin, and lisinopril. She is on the fence about staying: 'It's probably just my stomach.'
1. Ten minutes after arrival, which interventions do you start now, and which reflexive order would you withhold or modify specifically for this patient? Justify each with the findings.
Read the worked answer
Start now
- Monitored bed, continuous cardiac monitoring, repeat 12-lead in 15 minutes, troponin, CBC, BMP, coags. One millimeter of horizontal ST depression with T-wave inversion is ischemia, not artifact, and ST segments in diabetics drift; a single ECG is a snapshot, not a diagnosis.
- IV access and a monitored bed, not the waiting room. This is a high-risk anginal equivalent: diabetes, diaphoresis, tachycardia at 104, ischemic ECG changes, and a presentation the patient herself has mislabeled. The absence of chest pain is not reassuring in a diabetic.
- Aspirin 162-325 mg chewed. The two-week-old GI bleed with Hgb 9.2 is a reason to add a PPI, notify the provider, and avoid anticoagulation without a plan. It is not a reason to withhold aspirin when ACS is suspected; the mortality benefit is large and the bleed is remote and presumed stable. Document the reasoning.
- Nitroglycerin 0.4 mg SL, repeated every 5 minutes up to 3 doses because SBP is 168 and there is no right-ventricular infarct pattern, no PDE-5 inhibitor use, and no hypotension. Recheck BP before every dose; stop if SBP falls below about 90-100 or the pain worsens.
Withhold or modify
- No routine supplemental oxygen. At 94% on room air she does not need it; hyperoxia in normoxic ACS patients has no benefit and may cause coronary vasoconstriction. Titrate only if she falls below the ordered target.
- Do not empirically load a second antiplatelet before provider input in a patient with recent GI bleeding; that is precisely where the nurse advocates for nuance rather than following a default order set.
- Do not give a GI cocktail or antacid and reassess in an hour. That single decision is the error this case is built around; it delays the ECG and troponin and converts a treatable occlusion into lost myocardium.
2. The provider is tied up. In the next 60 seconds, what bedside findings would change your plan before aspirin and nitroglycerin are given, and why?
Read the worked answer
The competing diagnosis that changes management is acute aortic dissection, and her back pain plus hypertension makes the 60-second check mandatory rather than academic.
Findings that would stop me
- Abrupt, maximal-at-onset tearing or ripping pain (hers is a 4-hour dull ache, which argues against it, but ask specifically).
- Radiation to the back or scapula, or pain that migrates.
- A systolic difference of more than 20 mm Hg between arms, or unequal radial/femoral pulses.
- A new early diastolic murmur of aortic regurgitation.
- Syncope, focal neurologic deficit, or a pulse deficit.
- Refractory hypertension with pain out of proportion to the ECG.
Why it matters in the next 15 minutes
Antiplatelet and anticoagulant therapy in dissection can promote extension and hemorrhage, and the priorities become rate and blood pressure control, immediate CT angiography, and surgical referral. Nitroglycerin itself is not forbidden in dissection, but the decision architecture is completely different.
Why I would not let the check become paralysis
If bilateral pressures are equal, pulses are symmetric, there is no murmur, and the pain is gradual with diaphoresis and nausea, the pretest probability of dissection is low and delayed aspirin in a probable ACS costs lives. The move is a structured 60-second exam, documented, then proceed.
3. After one sublingual tablet she says she feels fine, wants to use the restroom, then wants to go home. What is your plan, and what do you document?
Read the worked answer
Plan
- Symptom relief is not risk stratification. Ischemic pain resolving after nitroglycerin is expected and does not exclude ongoing ischemia; a diabetic with autonomic neuropathy can have silent infarction. Serial troponins and serial ECGs in a monitored bed are non-negotiable.
- She is anemic (Hgb 9.2) with an ischemic ECG and a heart rate of 104 - oxygen delivery is marginal, so even modest ischemia matters more than the numbers suggest.
- Bed rest with bathroom privileges per protocol, continuous monitoring, provider notification, cardiology consult, and a GI-bleed plan (PPI, type and screen) if antiplatelet therapy escalates.
- Do not let her walk unassisted to the bathroom while on nitroglycerin and during an unresolved ischemic evaluation; orthostasis and syncope are predictable.
Documentation
- Exact time of symptom onset, character, radiation, and the patient's own words ('indigestion').
- Time of each ECG, the specific findings, time of troponin draws.
- Each medication with dose, route, time, and the BP before and after nitroglycerin.
- The conversation about the recent GI bleed and the decision to give aspirin anyway, with the provider's name and time.
- If she insists on leaving: her stated reasons, her demonstrated understanding of the risks, notification of the provider, and the against-medical-advice process per policy. A patient who 'feels fine' after one tablet is the patient most likely to be undertreated.