As recent readers of this blog know, I'm interested in whether our framework can be useful in real time diagnosis of cases where multiple diagnoses are being considered to explain clinical findings. In this case, an immunocompromised patient (SLE, RA; receiving prednisone and cyclophosphamide) presented with two weeks of fever and chills without other symptoms. So, we can be pretty confident she's infected, and is highly predisposed to opportunistic infections as well as garden variety ones.
Her chest imaging (in the thumbnail) showed a predominant finding of pulmonary micronodules along with ground glass opacieies (GGO) and lymphadenopathy. We can make a differential right here based on a miliary pattern, and it would include mycobacterial diseases, fungal diseases, lymphoma, etc.
What would not be atop the differential diagnosis is PJP pneumonia because this would be an unusual presentation for that. We would expect PJP to present with subacute dyspnea on exertion and other chest symptoms, and with just GGO or interstitial opacities. The radiology in this case is not consistent with PJP, though it could rarely present in this fashion (an multifarous others).
Alas, a bronchoalveolar lavage (BAL) was performed and PJP was found (we are not told what finding on the BAL supported this). Treatment for PJP was initiated but she didn't improve, was intubated, and, without molecular diagnostics available at their hospital, they did transbronchial biopsies and very clearly documented the presence of both PJP (right) and Histoplasma capsulatum (left), the latter being a perfectly satisfactory explanation for the clinical presentation and imaging findings:
Based on our framework, we expect immunocompromised patients to not uncommonly have more than one pathogen because of Reichenbach's Common Cause Principle, which states that if an unusual coincidence occurs between events A&B, one either caused the other, or both share a common cause, C. So whenever we have an underlying disease like immunosuppresion, we should be prepared for multiple downstream infectious (or other) complications stemming from it. Alternatively, it's possible that her entire presentation is explained by histoplasmosis and the PJP represents an asymptomatic incidentaloma, albeit a consequential one that must be treated in the context of immunosuppression (and absence of prophylaxis). In short, we expect HD in immunocompromised patients. No surprises there.
One lesson from this case is that whenever you go testing (fishing) for things with a big net, you have to carefully discern whether what you dredge up is a suitable explanation for the clinical presentation. In this case, garden variety pneumonia (for which she was treated with 2 antibiotic regimens on days 1-2) was never a suitable explanation. BAL was not performed until day 7, and the premature settling upon PJP delayed further the ultimate diagnosis until biopsies on day 12.
The biggest lesson here is not whether multiple diagnoses occur, they commonly do! It is whether you are attending to context and whether or not your findings map well enough onto the clinical presentation and findings to settle upon a diagnosis that is an inadequate explanation for the clinical picture leading to delays in the real underlying diagnosis.
