Trends in Hospitalizations for Pneumocystis Pneumonia among Patients with Rheumatoid Arthritis in the United States: 1996 to 2007

Pneumocystis (jiroveci) pneumonia (PCP) is an uncommon but frequently fatal occurrence among patients with rheumatoid arthritis (RA). In the era before aggressive immunosuppressive therapy and biologicals, the frequency of PCP among patients with RA in France was estimated to be 1.3 cases/10,000 patients/year (1). We previously reported that rates of PCP among hospitalized patients with RA in California from 1983 to 1994 were similarly low and that there were no distinguishable trends in rates over time; however, in-hospital mortality of patients with RA who had PCP was 30.8% (2). After the introduction of infliximab and etanercept for the treatment of RA, there were reports of PCP developing in patients with RA who had been treated with these medications (3). The U.S. Food and Drug Administration subsequently issued a “black box” warning regarding the risk of PCP associated with use of tumor necrosis factor (TNF) inhibitors. A postmarketing surveillance study in Japan in 2003 – 2004 reported 22 cases of PCP during a 6-month follow-up of 5000 patients with RA who had received infliximab (4). Similar findings were reported in a separate postmarketing study of patients with RA treated with etanercept in Japan (5). In light of these findings, we investigated recent trends in the occurrence of PCP among patients with RA in the U.S. We performed two separate analyses using two different population-based hospitalization databases: a national-level database to estimate cases of PCP in the country, and a state-level database which allowed us to compute rates of PCP. Data from the Nationwide Inpatient Sample (NIS) provided national estimates of the number of cases in which hospitalized patients with RA were diagnosed with PCP, while data from the California Office of Statewide Health Planning and Development (OSHPD) provided a statewide estimate of the rates of PCP among patients with RA. We analyzed data from NIS for the 12-year period from 1996 to 2007. NIS is the largest inpatient database in the U.S. and, through 2007, contained data for up to 8 million inpatient stays per year from 1044 hospitals in 40 states, representing 20% of all community hospitals (6). We abstracted data on all hospitalizations of patients aged ≥ 18 years with a discharge diagnosis of RA by International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code (714.0, 714.1, 714.2). To restrict our analysis to PCP associated with RA, we excluded patients with a concomitant diagnosis of HIV/AIDS, cancer, or organ transplant recipient. Cases of PCP were identified by ICD-9-CM code (136.3). NIS trends files from 1996 to 2002 and core files from 2003 to 2007 were used, and corresponding discharge sample weights were applied to provide trends in national estimates. The California OSHPD collects discharge abstracts of all acute care hospitalizations from non-federal hospitals in the state (7). There are approximately 3.6 million hospitalizations annually. Rates of PCP in hospitalized patients with RA were computed using the total number of cases of PCP among patients with RA as the numerator (excluding those with concomitant HIV/AIDS, cancer, or organ transplant, as above), and the estimated number of persons with RA in California as the denominator for each year (8). Estimates for the denominator were derived by applying age- and sex-specific prevalences of RA (based on data from the Third National Health and Nutrition Examination Survey (9)) to yearly estimates of the population of California. Annual crude rates were standardized to the age and sex distribution of the 2000 RA population in California. SUDAAN 10 (Research Triangle Institute, NC) and SAS 9.2 (SAS Institute, Cary, NC) were used for the analyses. Based on NIS, there were 788 cases of PCP among hospitalized patients with RA in the U.S. from 1996 to 2007. Patients were mostly older (73.5% were between ages 50 to 80 years) and women (67.6%). There were no discernable trends in the number of PCP cases over the study period (Figure 1a). Using the California data, we identified 84 cases (range 2 to 12 cases/year) of PCP among patients with RA from 1998 to 2007. Patients were primarily older (65.5% were aged ≥ 50 years) and women (75.0%). Adjusted rates of PCP among patients with RA in California ranged from 0.6 – 4.0 cases/100,000 patients/year, with no apparent trend over time (Figure 1b). Figure 1 (a) National estimates of Pneumocystis pneumonia (PCP) cases among hospitalized patients with rheumatoid arthritis (RA) in the United States from the Nationwide Inpatient Sample. Error bars represent standard deviations. (b) Age- and sex-adjusted rates ... We did not identify any trends in cases of PCP among hospitalized patients with RA in the U.S, and replicated this result in an analysis of rates of PCP among patients with RA in California. These findings indicate that any change in the occurrence of PCP associated with the use of immunosuppressants and biologicals is not detectable over the background occurrence. Trends may have been obscured if PCP was undiagnosed, if patients with RA who developed PCP were treated as outpatients, or if chemoprophylaxis was routinely used, although there is no evidence that any of these factors changed substantially over the time period examined. Alternatively, use of these medications may not yet be prevalent enough, or this complication may not be frequent enough, to result in increases in rates of PCP at the population level. Our findings are reassuring in that they do not suggest any detectable increase in the frequency of PCP among patients with RA in the U.S. through 2007.