CHANGING USE OF INTENSIVE-CARE FOR HIV-INFECTED PATIENTS WITH PNEUMOCYSTIS-CARINII PNEUMONIA

Clinicians' approach to acute respiratory failure from Pneumocystis ca rinii pneumonia (PCP) is hypothesized to have gone through three phase s: aggressive management due to an absence of data on prognosis (1981- 84), withholding of intensive care based on a few small studies showin g high case fatality (1985-87), and an increase in intensive care to a n intermediate level (1988 forward). Unfortunately, studies of surviva l from acute respiratory failure among such patients have been small a nd have been limited to patients in the intensive care unit. To determ ine whether this three-phase scenario has empirical support, we perfor med a retrospective chart review of all patients with human immunodefi ciency virus (HIV) infection and PCP at a university-affiliated munici pal hospital from 1983 to 1990. We identified 180 patients, representi ng 218 episodes of PCP: The previously hypothesized relationship betwe en intensive care and year of diagnosis was confirmed: intubation rate s were 30% before 1985, 0% in 1987, and 12% after 1988 (p = 0.03). Amo ng all patients, the percentage dying in the hospital without intensiv e care had the opposite relationship with year of diagnosis, increasin g from 0% in 1984 to 21% in 1987 and then declining to 0% in 1990 (p = 0.001). Overall mortality from an episode of PCP was 25% and did not change significantly over time. Disease severity also did not change s ignificantly over time. In summary, the significant swings in the use of intensive care for HIV-infected patients with PCP suggest that judg ments about the futility of intensive care were strongly influenced by incorrect perceptions of survival.