Objective: To determine factors associated with success or failure of oral antibiotic treatment for nursing home-acquired pneumonia (NHAP).
Design: Retrospective study of outcomes for all identifiable NHAP cases in 1991.
Setting: The Nursing Home Services Program of St. Paul Ramsey Medical Center and 31 metropolitan St. Paul, Minnesota, community nursing homes.
Participants: Nursing home (NH) cohort: 124 patients (mean age 85.2 years) with a new respiratory symptom and new infiltrate on portable chest X-ray for whom oral antibiotics were prescribed. Hospital cohort: 74 NH patients (mean age 84.3 years) admitted to hospital with new X-ray infiltrate and pneumonia diagnosis. Supportive care status patients were excluded. Forty-three physician/nurse practitioner (MD/NP) teams were represented.
Measurements: Nursing home cohort: Outcomes of hospitalization within 14 days or 30-day mortality. A discriminant model was applied to predict outcome and discriminant rule performance was analyzed. Hospital cohort: 30-day mortality.
Results: Of 198 episodes of NH pneumonia, 63% were treated in the facility; 30.6% (38) failed NH treatment. Thirty-day mortality was 13%. There was no examination by the MD or NP for 59% of NH-treated episodes. The hospital cohort had a higher mean pulse (P < .05) but a similar frequency of feeding dependence. Hospital cohort mortality was 17.6%. The NH treatment failure group had significantly higher proportions of pulse > 90/min, temperature > 100.5 degrees F, respirations > 30/min, feeding dependence, and mechanically altered diets. A discriminant model using these factors was significant (P = .002). The NH treatment failure rate was 11% for no factors present, 23% for two or fewer factors, and 59.5% for three or more (likelihood ratio 3.1). Thirty-two percent of the hospital cohort had zero or one factor present and were alive at 30 days.
Conclusion: The majority of NHAP episodes were treated successfully with oral antibiotics, but 31% failed treatment in the NH. Patients with a mechanically altered diet or requiring feeding assistance by staff had significantly higher failure rates. Feeding dependence and need for a mechanically altered diet as well as abnormal vital signs are associated with oral antibiotic treatment failure. These factors should be considered in treatment decisions for NHAP.