The oft-repeated mantra in medicine is the history and, to a lesser extent, physical examination, hold the vast majority of the clues necessary for appropriate diagnosis and treatment. Such wisdom, handed down by sages trained in the pre-penicillin era, is reinforced throughout medical training.
While such platitudes may have an element of truth, unfortunately, the patient may in fact be the least-qualified person to provide said history.
These authors compared two data sources, the Medicare claims database and a self-reported Health and Retirement Study, looking for the diagnosis of acute myocardial infarction. Of the 45,335 patients verified in both data sources, 3.1% self-reported having an acute MI during the preceding 2.5 years. However, only 32.3% of those could be verified using Medicare claims data; using acute coronary syndrome as a broader definition of AMI verified the self-reported history in only 48.7%. Conversely, of the 1.4% of patients for whom Medicare claims data indicated an acute MI, only 67.8% self-reported the event. 90.5%, at least, did state they had heart problems.
So, your undifferentiated elderly patient may or may not have had an acute MI, regardless of what they actually report. These results are mildly surprising, considering it is reasonable to expect the general public to have sufficient health literacy to understand a major diagnosis like “heart attack”. Then again, anyone working in the Emergency Department knows the profound challenges of extracting reliable information from the undifferentiated patient.
“Comparison of Self-Reported and Medicare Claims-Identified Acute Myocardial Infarction”
http://www.ncbi.nlm.nih.gov/pubmed/25747935
Thanks to @bloodman for the article!
I think the subtext of the old saw is that the history of the present illness holds the key to a patient's ills, but for their PAST medical history, you'd better read from a sheet.
I think MI might not have been the best diagnosis to pursue here. Given the pitfalls and wide practice variation regarding how elevated troponin levels are used to diagnose MI and how that information is relayed to the patient, I'm not surprised these numbers were all over the place based on that alone. I can't tell you how many times I've seen a physician tell a patient "Well, you have pneumonia, and it looks like you also had a heart attack…" when someone presents septic with a troponin-I of 0.08 ng/mL (ref <=0.04 ng/mL).
Perhaps stroke would have been a better diagnosis to investigate since the criteria are a bit more concrete (though clearly not 100% straightforward).
Probably true – on both counts!
There are a lot of different definitions of "heart attack", and, yes, there can be imprecise coding along a spectrum of chest pain, ACS, and STEMI. However, I think it's pretty reasonable to use this example of limited health literacy/erroneous recall. Stroke is a good suggestion, however, I've seen plenty of instances of patients confusing "stroke" with AMS NOS or "mini-stroke". Plus, anyone who gets tPA is labeled a stroke with "full recovery" if no one goes looking closely to rule-out stroke mimic.