We reviewed 153 episodes of invasive pneumococcal disease involving 147 hospitalized patients with and without human immunodeficiency virus (HIV) disease to examine and compare epidemiologic and clinical features, capsular serotypes, and antibiotic susceptibility patterns. HIV infection was the most common risk factor for invasive pneumococcal disease. Pneumococcal disease in HIV-infected individuals was characterized by the greater frequency with which pneumonia was the source of bacteremia (90% vs. 63%) (P < .01) and an increased recurrence rate (15% vs. < 1%) (P < .01). The overall mortality rate was 12% and did not vary by HIV serostatus. Capsular-type data were available for 149 episodes; 90% of the types were among those found in the polyvalent pneumococcal vaccine. The four most common capsular types causing invasive disease were 14, 6b, 9v, and 22f; capsular type 9v was significantly more common among HIV-infected patients (P < .01). Penicillin-resistant isolates were identified in 7.2% of all cases, and their presence did not vary by HIV status; 20% of isolates from cerebrospinal fluid were resistant. The majority of the resistant isolates were of capsular type 9v. Given the worldwide increase in both HIV and penicillin-resistant pneumococcal infections, better preventative and therapeutic strategies are greatly needed.

No abstract

We offered standardized gynecologic examinations to consecutive women admitted to an AIDS-designated inpatient medical service; 65 (97%) of 67 women consented to the examination. The median CD4+ T lymphocyte count was 54/mm3. Only 9% of the women were admitted for primary gynecologic or genitourinary diagnoses; however, on evaluation, 83% of these women had gynecologic disease. The overall prevalences of vaginitis, cervical dysplasia, genital condylomata, genital herpes, and pelvic inflammatory disease were 51%, 45%, 23%, 20%, and 5%, respectively. Unexpected findings included adenovirus infection and foscarnet-associated genital ulcerations (two cases each). For predicting disease, gynecologic symptoms had a sensitivity of 76% and a positive predictive value of 95% but a negative predictive value of only 41%. Our results document the high prevalence of comorbid gynecologic disease among women infected with human immunodeficiency virus (HIV). Because of the inability to fully predict disease by symptom history, it is imperative that comprehensive gynecologic evaluation be offered routinely to all HIV-infected women hospitalized for acute medical illnesses.