Pneumocystis pneumonia in HIV-positive patients in Spain: epidemiology and environmental risk factors

Abstract

INTRODUCTION: Specific environmental factors may play a role in the development of Pneumocystis pneumonia (PCP) in HIV-positive patients. The aim of this study was to estimate the PCP incidence and mortality in hospitalized HIV-positive patients in Spain during the combination antiretroviral therapy (cART) era (1997 to 2011), as well as to analyze the climatological factors and air pollution levels in relation to hospital admissions and deaths. METHODS: We carried out a retrospective study. Data were collected from the National Hospital Discharge Database and the State Meteorological Agency of Spain. A case-crossover analysis was applied to identify environmental risk factors related to hospitalizations and deaths. For each patient, climatic factors and pollution levels were assigned based on readings from the nearest meteorological station to his or her postal code. RESULTS: There were 13,139 new PCP diagnoses and 1754 deaths in hospitalized HIV-positive patients from 1997 to 2011. The PCP incidence (events per 1000 person-years) dropped from 11.6 in 1997 to 2000, to 5.4 in 2004 to 2011 (p<0.001). The mortality (events per 10,000 person-years) also decreased from 14.3 in 1997 to 2000, to 7.5 in 2004 to 2011 (p<0.001). Most hospital admissions and deaths occurred in the winter season and the fewest occurred in the summer, overlapping respectively with the lowest and highest temperatures of the year in Spain. Moreover, lower temperatures prior to PCP admission, as well as higher concentrations of NO2 and particulate matter up to 10 m in size (PM10) at the time of admission were associated with higher likelihoods of hospital admission due to PCP when two weeks, one month, 1.5 months or two months were used as controls (p<0.01). Furthermore, higher concentrations of ozone at one month (p=0.007), 1.5 months (p<0.001) and two months (p=0.006) prior to admission were associated with higher likelihoods of hospital admission with PCP. For PCP-related deaths, lower temperatures prior to admission and higher concentrations of atmospheric PM10 at the time of admission were related to higher likelihood of death when two weeks, one month and 1.5 months were used as controls (p<0.05). CONCLUSIONS: PCP was a significant health problem in the cART era (1997 to 2011), and PCP epidemiology was adversely influenced by colder climatological factors and higher ambient air pollution levels.