A single-center, person-month-based analysis of the risk of developing Pneumocystis pneumonia (PCP) in immunosuppressed non-HIV patients: Preventive effects of trimethoprim-sulfamethoxazole

We performed a retrospective study to evaluate the risk factors for acquiring Pneumocystis pneumonia (PCP) by pharmacologically immunosuppressed HIV-negative patients.Patients who received corticosteroids, immunosuppressive agents, anticancer agents, and radiotherapy with or without trimethoprim-sulfamethoxazole (TMP-SMX) at Himeji Medical Center between 2010 and 2021 were evaluated. Drugs and doses of the treatments for each patient were divided by month into person-month units. Each person-month datum includes information on the administered drug (or radiotherapy), average doses, and whether the patient had PCP during the corresponding month. ROC curves with person-month data were generated for each treatment, and AUCs >0.7 were identified as possessing positive classification utility. The risks for PCP according to gender, age (grouped by median) and each treatment were examined by univariate analysis, followed by multivariate analysis to identify independent factors.Of a total of 17,733 patients (214,676 person-months), 32 developed PCP. The cut-off values by ROC analysis were 13.7 mg/day for corticosteroid (prednisolone equivalent), 0.92 mg/day (6.45 mg/week) for methotrexate (MTX), and 34.3 mg/day for TMP-SMX. The cut-off values for other treatments could not be estimated. The above three drugs and male sex were significant variables in univariate analysis and were all confirmed as independent factors by multivariate analysis.The results suggest that a monthly average dose of ≥13.7 mg/day of prednisolone, ≥0.92 mg/day of MTX and male sex are significant independent risk factors for PCP, and that prophylaxis with ≥34.3 mg/day of TMP-SMX is to be recommended.