Estimation of increased hospital stay due to nosocomial infections in surgical patients: comparison of matched groups.

An investigation, using a prospective cohort study, was performed to estimate the prolongation of hospital stay caused by nosocomial infections in surgical patients. An evaluation of the one-to-one matching method, as a model for similar studies was also undertaken. Between 1992 and 1994, 225 of 1482 surgical patients (15%) developed infection. Of these, 223 evaluable patients were compared with 1256 uninfected cases, in an unmatched analysis. In a further analysis, 151 infected cases were matched one-to-one with uninfected controls, and other factors such as age, length of preoperative stay, presence of malignancy or diabetes and presence of foley catheters or drains, were evaluated. Patients with infection were hospitalized for 10.6 days longer than the matched controls. In the unmatched analysis, infected patients appear to remain in hospital for a further 17 days. The difference in the prolongation of stay between matched and unmatched groups indicates the importance of matching as an essential model for such studies. In cases where matching cannot be performed, the median can be a good alternative measure to the arithmetic mean.