Pneumonia in HIV-infected patients: a case-control survey of factors involved in risk and prevention

Abstract

OBJECTIVE: To assess the factors that increase or decrease the risk of pneumonia with particular attention to immunization with pneumococcal and influenza vaccines in a group of HIV-infected persons. DESIGN: A retrospective, case-control study based on information entered into a standard database and the medical record. SETTING: Patients attending a referral clinic specializing in AIDS/HIV care at a public hospital. PATIENTS: Among over 2000 subjects entered into a database in 8 years, 127 incidents of pneumonia were identified from the record. These cases were matched with 127 CD4 cell count matched, concurrent controls. INTERVENTIONS: None. MAIN OUTCOME MEASURE: The principal hypothesis was that chart review would find a decreased frequency of pneumococcal immunization in the pneumonia cases compared with matched controls. RESULTS: Pneumococcal immunization was associated with a reduction of the risk of pneumonia by nearly 70%. The effect was seen even when immunization was given with a CD4 cell count of less than 100/mm3. Injection drug users and African-Americans had a twofold increased risk of pneumonia. CONCLUSION: The study provides data to support the current recommendation for pneumococcal immunization of all HIV-infected persons. Although this conclusion could lead to renewed enthusiasm for increasing pneumococcal immunization rates in HIV-infected persons, it must be recognized that the study is observational and ascertainment bias cannot be excluded.