Risk factors for nosocomial pneumonia in the elderly.

PURPOSE: Elderly patients have a disproportionate incidence of nosocomial pneumonia (NP) and a higher mortality rate, yet few studies have focused on this high-risk population. We undertook a study to examine risk factors for NP in elderly inpatients and to describe how these patients differ from younger patients with NP. METHODS: In a public teaching hospital, all cases of NP in patients aged 65+ were ascertained by prospective surveillance during a 2-year period (n = 59). These elderly cases were compared with 59 cases of NP in patients aged 25 to 50 to describe differences in risk factors and outcomes. Elderly cases were then matched to elderly control subjects who were admitted to the same hospital service but did not develop NP. Data were collected on known risk factors and on the potential risk factors of poor nutrition, neuromuscular disease, and dementia. Significant differences in risk factors were analyzed using univariate and multivariate comparisons of cases and controls. RESULTS: Elderly patients had twice the incidence of NP (RR = 2.1) as younger patients. Onset of infection was earlier for young than for older cases (6 versus 11 days, p less than or equal to 0.02), but mortality following NP was equal for the two age groups (42% versus 44%). No significant differences in risk factors were found for old and young cases, although older cases tended to have higher rates of poor nutrition, neuromuscular disease, and aspiration preceding their pneumonias. Comparison of elderly cases and elderly controls revealed significantly increased frequencies of poor nutrition, neuromuscular disease, pharyngeal colonization, aspiration, depressed level of alertness, intubation, intensive care unit admission, nasogastric tube use, and antacid use among cases. Cases were more severely ill on admission and had more pre-existing risk factors (2.8 versus 1.3, p less than or equal to 0.001) and more in-hospital risk factors (4.7 versus 1.6, p less than or equal to 0.001). Logistic regression analysis revealed low albumin, diagnosis of neuromuscular disease, and tracheal intubation to be strong independent predictors of risk for NP among elderly inpatients. CONCLUSIONS: We conclude that the specific risk factors of poor nutrition, neuromuscular disease, and tracheal intubation may prove useful to target future clinical interventions to prevent NP in the elderly.

Duke Scholars

Author Gregory P. Samsa Biostatistics & Bioinformatics, Division of Biostatistics