Treatment Adequacy for HIV-related Pneumocystis Pneumonia: Quality Measures for Inpatient Care

• Published in: International journal for quality in health care Vol. 9; no. 5; pp. 349 - 359
• Main Authors: MATHEWS, WM. CHRISTOPHER, KANOUSE, DAVID E.
• Format: Journal Article
• Language: English
• Published: England Oxford University Press 01-10-1997; PERGAMON
• Subjects: AIDS; AIDS-Related Opportunistic Infections - epidemiology; AIDS-Related Opportunistic Infections - therapy; AIDS/HIV; Analysis of Variance; California - epidemiology; Comorbidity; Female; Hospital Mortality; Hospitals - standards; Hospitals - statistics & numerical data; Humans; Logistic Models; Logistic regression; Male; Odds Ratio; Outcome Assessment (Health Care); Pneumocystis pneumonia; Pneumonia, Pneumocystis - epidemiology; Pneumonia, Pneumocystis - therapy; Process of care; Quality Indicators, Health Care; Quality of care; Reproducibility of Results; Retrospective Studies; Risk adjustment; Severity of Illness Index; Survival Rate
• ISSN: 1353-4505; 1464-3677
• DOI: 10.1093/intqhc/9.5.349

To develop and evaluate severity-adjusted indicators of treatment timeliness and adequacy for inpatient care of first episode of HIV-related pneumocystis pneumonia, a retrospective cohort study (n=414) using medical record review was conducted in six California medical centers (1 January 1983–30 June 1987). Measures included patient baseline characteristics and complexity, process-of-care indicators (delay in treatment initiation and proportion of adequate treatment delivered), and overall survival of hospitalization and survival without respiratory failure. Logistic regression models of severity were developed among optimally treated patients and cross-validated. Exposure to medication with pneumocystis activity within 30 days prior to admission was protective. After controlling for pre-admission medication and severity, the average proportion of adequate pneumocystis medication delivered during the first 7 and 30 days were significant predictors of outcome in all models. Delay in treatment initiation, while not a statistically significant predictor, was associated with baseline severity. Summary measures of treatment adequacy show promise as process-of-care indicators. © 1997 Elsevier Science Ltd.