The Case
by Brady Pregerson, MD
A 55-year-old male with a history of hypertension and GERD was brought by ambulance to the emergency department for a syncopal episode that occurred while he was at his doctor’s office. The day prior he had been to the same physician for evaluation of a two-week history of dry cough, subjective fever, and dyspnea on exertion. At that time, he was diagnosed with pneumonia based on his exam and chest x-ray and was started on ciprofloxin and azithromycin. He returned because he felt worse, had intermittent sweating, and had fainted twice earlier in the day. The primary care physician evaluated him briefly then sent him to the emergency department by ambulance.
The emergency physician obtained a similar history and also noted that the patient had complained of intermittent chest tightness for the prior two days. Triage vital signs showed temperature 96.8, BP 109/69, pulse 140, RR 24, and oxygen saturation 90% on face mask. The physical exam documented by the emergency physician was notable for tachycardia and rales about half way up both lungs. The rest of the exam was documented as being normal.
An EKG showed atrial fibrillation with a rate of 143 and non-specific ST and T wave changes but no ST elevation. A portable chest x-ray showed bilateral infiltrates and cardiomegaly. Lab tests revealed a WBC count of 21.3, a BNP of 977, and a troponin of 0.8. BUN and creatinine were elevated at 35 and 1.9. The remainder of the CBC and chemistry panel were normal.
The patient was initially treated with IV fluids, aspirin, and diltiazem. His heart rate improved to the 90s. When the CXR and BNP showed findings consistent with CHF, he was also given IV Lasix. He was admitted to the on-call internist to a telemetry bed with diagnoses of new onset atrial fibrillation, CHF, and possible pneumonia. No further testing or treatment was initiated in the emergency department.
Analysis
by William Sullivan, DO, JD
First, we have to acknowledge that this patient was quite ill. If we assume that the patient had pneumonia, based on his PORT Score, he was at least a Level IV Risk Class with a 9% risk of mortality when he hit the door. There is no mention of an ABG being done, but with the patient’s hypoxemia and hyperventilation, it is likely that he would have had abnormal ABG results, pushing him into a Level V Risk Class with a 28% risk of mortality. The patient failed outpatient treatment (although perhaps not on the optimum antibiotic regimen), fainted twice, and was hypoxic, tachypneic, tachycardic, and in renal insufficiency when he arrived. He met sepsis criteria, but we don’t have enough information to calculate a SOFA score.
The differential diagnosis in this case is wide. Cough, rales in both lungs, bilateral infiltrates on x-ray, hypoxemia, and WBC count of 21.3 certainly point toward pneumonia with potential underlying sepsis as a diagnosis. However, cardiomegaly, elevated BNP, and mildly elevated troponin just as easily support a diagnosis of CHF. His renal insufficiency didn’t meet criteria for prerenal azotemia, but his vital signs improved with an IV fluid bolus and Cardizem. Was his renal insufficiency intrinsic or due to a fluid deficit? Then there’s the atrial fibrillation. One thing that always helped me figure out the etiologies for new onset atrial fibrillation was to remember that atrial fibrillation is a symptom of another medical problem. The patient’s hypoxia and/or CHF could have precipitated his atrial fibrillation. Other potential causes of atrial fibrillation including MI, thyrotoxicosis, drugs, valvular abnormalities, and pulmonary embolism could also be considered.
Initial treatment with fluids for potential pneumonia was appropriate and seemed to improve the patient’s vital signs, although the patient then received diuretics once he was diagnosed with suspected CHF.
There are a lot of pieces missing in this puzzle, though.
Was the patient’s troponin addressed? Could an ischemic event have precipitated these symptoms?
Why wasn’t the patient started on anticoagulants for his atrial fibrillation? His CHA2DS2-VASc score was probably 0, making the patient a low risk for thromboembolism and possibly negating the need for anticoagulation, or warranting aspirin, which the patient received. Ultimately, though, that decision would probably be deferred to a cardiologist.
And why did the patient pass out? Hypoxia? Arrhythmia? Vasovagal event? Should the patient have received antibiotics in the emergency department? We can’t tell whether the patient met severe sepsis criteria based on the information given. Best practices may suggest that patients with pneumonia or sepsis receive early antibiotics; however, early antibiotics may not affect outcome. Some studies show a mortality benefit with early antibiotic administration [Ferrer, Crit Care Med Aug 2014], while others do not [Sterling, Crit Care Med Sept 2016].
This patient undoubtedly required further treatment and further evaluation for his symptoms. The case likely involved a missed diagnosis, which would raise the issue of whether management in the emergency department was sufficient given the patient’s symptoms.
Case presentations such as these tend to create a bias since we assume that there was a bad outcome at the outset of the presentation. We then look backward to determine whether any steps could have been taken to prevent the known or suspected bad outcome. In this case, the patient’s vital signs stabilized with his initial treatment. Anticoagulants and antibiotics could have been started in the emergency department but were not. Would earlier administration of these medications have prevented an adverse outcome? Possibly, but before making that determination we would also have to ask why the medications were withheld. Recall that a “reasonableness” standard is situation-specific. A formal or bedside echocardiogram would have been helpful in guiding treatment. I don’t think we have reached the point that failure to order such testing in the emergency department should be considered unreasonable. An ICU admission may have allowed closer monitoring of the patient’s condition, but was closer monitoring necessary after the patient improved with treatment?
The standard by which we are judged is “reasonable” or “average” care—not “outstanding” or “perfect” care. I personally would have initiated antibiotics and contacted a cardiologist before sending the patient to the telemetry floor, but I also don’t think failing to do so constitutes “unreasonable” care. If this case resulted in a bad patient outcome, the treating physician’s greatest challenge will be explaining why he didn’t do more in the emergency department.
3 Comments
Hello,
I am a medical coder for a couple of emergency physician practices and see this scenario on a frequent basis. Approaching this case from a coder’s perspective, I would have pended the chart querying for critical care based on the diltiazem and Lasix and work up; and captured the a fib, CHF, syncope, and hypertension. Pneumonia would not be codeable as it is suspect in an outpatient setting per coding guidelines. Some of my colleagues would link the CHF to the hypertension but I would not as the CHF is not documented as caused by the hypertension or the a fib.
I find it interesting that the CHF can cause the a fib, I’ve come across other sites that say the a fib can cause the CHF. In regards to the anti-coagulation, wouldn’t that be determined while the patient is in-house?
As you note in your article, the lack of documentation can be detrimental to our ability to code correctly as well as affect continuity of care.
.The initial presentation of upper respiratory tract infection symptoms followed by Heart failure (proved by clinical & radiological evidence ) is suggestive of Myocarditis .Ideally the patient should have undergone bed side echo followed & should have been taken over by cardiology for further managment.
Pt was sick enough to flag the physician of a complicated pneumonia at the very least. Lactate, ABG’s, antibiotics, a shot of LMWH, POCUS to check “squeeze” plus ICU admission would have put a bow on top of the management.