Background: Pneumocystis pneumonia (PcP) is a life-threatening condition for patients with HIV and other immunocompromised patients. Effective treatment requires timely diagnosis, but little is known about the frequency of diagnostic delays.
Methods: We conducted a retrospective cohort study using the Truven Health Analytics Commercial Claims and Encounters Database from 2011-2016. We identified index case visits for PcP as the point at which a patient was first diagnosed with PcP. We analyzed visits prior to the index diagnosis for symptoms related to PcP; these symptoms included cough, fever, dyspnea, pneumonia or other respiratory infections. We performed a change-point analysis to identify the time window before the index PcP diagnosis where diagnostic opportunities began to appear. We used a simulation model to estimate the likelihood of potential delays representing actual diagnostic opportunities.
Results: In our cohort, we identified 7,656 case patients with PcP, 5,187 in patients with HIV and 2,469 in patients without HIV. We observed a dramatic spike in hospitalizations, ED visits and outpatient visits for symptoms associated with PcP in the period just before the index PcP Diagnosis (see figure). In the time prior to the index PcP diagnosis, we identified 15,963 visits, and 4,674 patients, with at least one symptom of PcP. Of these potential opportunities, we estimated that 5,396 visits (CI 5,256-5,542), in 2,868 patients (CI 2,813-2,925), represented likely diagnostic delays. Around 56% of diagnostic opportunities occurred within 14 days of the PcP diagnosis, and 72% occurred within 3 weeks. Diagnostic delays were most likely to occur in outpatient settings: 4,429 outpatient opportunities (CI 4,546-4,679) compared to 674 (CI 706-738) inpatient or ED opportunities.
Patients without HIV were more likely to experience a diagnostic delay: 77.4% (CI 74.7-80.2) of non-HIV patients, compared to 18.5% (CI 17.7-19.2) of HIV patients, experienced a delay. Diagnostic delays tended to be longer in patients without HIV, 16.1 days (15.7-16.4), compared to patients with HIV, 14.9 days (CI 14.8-15.1).
Conclusion: Many opportunities to diagnoses PcP are missed or delayed, especially in outpatient settings. Patients without HIV are at greater risk of experiencing a diagnostic delay, and may experience delays of longer duration