Volume, quality of care, and outcome in pneumonia

Background: The establishment of minimum volume thresholds has been proposed as a means of improving outcomes for patients with various medical and surgical conditions. Objective: To determine whether volume is associated with either quality of care or outcome in the treatment of pneumonia. Design: Retrospective cohort study. Setting: 3243 hospitals participating in the National Pneumonia Quality Improvement Project in 1998 and 1999. Patients: 13 480 patients with pneumonia cared for by 9741 physicians. Measurements: The association between the annual pneumonia caseload of physicians and hospitals and adherence to quality-of-care measures and severity-adjusted in-hospital and 30-day mortality rates. Results: Physician volume was unrelated to the timeliness of administration of antibiotics and the obtainment of blood cultures; however, physicians in the highest-volume quartile had lower rates of screening for and administration of influenza (21%, 19%, 20%, and 12% for quartiles 1 through 4, respectively; P< 0.01) and pneumococcal (16%, 13%, 13%, and 9% for quartiles 1 through 4, respectively; P < 0.01) vaccines. Among hospitals, the percentage of patients who received antibiotics within 4 hours of hospital arrival was inversely related to pneumonia volume (72%, 64%, 60%, and 56% for quartiles 1 through 4, respectively; P < 0.01), while selection of antibiotic, obtainment of blood cultures, and rates of immunization were similar. Physician volume was not associated with in-hospital or 30-day mortality rates. Odds ratios for in-hospital mortality rates rose with increasing hospital volume (1.14 [95% Cl, 0.87 to 1.49], 1.34 [Cl, 1.03 to 1.75], and 1.32 [Cl, 0.97 to 1.80] for quartiles 2 to 4, respectively); however, odds ratios for 30-day mortality rates were similar. Limitations: This study was limited to Medicare beneficiaries 65 years of age and older. Ascertainment of some measures of the quality of care and severity of illness depended on the documentation practices of the physician. Conclusion: Among both physicians and hospitals, higher pneumonia volume is associated with reduced adherence to selected guideline recommendations and no measurable improvement in patient outcomes.