[Nosocomial infections in surgical patients. Problems of measuring and comparing results]

What is striking when studying the frequency of nosocomial infection (NI) is the variability of the study data. Different frequency indicators and infection criteria are used for estimates and these make it difficult to compare works. The aim of this work is to estimate the frequency of hospital infection by using different indicators to compare the results.

A market study was carried out including patients admitted to four surgical units over the period of one year. The following indicators were used: proportion of patients infected, cumulative number of cases of infection and density of number of cases. The infections were detected through active search and included those acquired in Intensive Care Units and those diagnosed after patients had been discharged from hospital.

A total of 14.5% of the patients suffered NI and 5% of the infections were diagnosed after discharge from hospital. In 38.5% of the cases of infection a microbiological study was not requested. The General Surgery Unit had the highest figures for the three indicators. Nevertheless, the magnitude of the differences between services was modified in line with the indicator used.

The real percentage of patients with NI is higher than the values given by the usual monitoring systems. Given the trend witnessed over recent years whereby the length of hospital stays is being reduced and early discharge programmes promoted with the aim of increasing efficiency, densities for the number of cases should be estimated and these should include the NI cases diagnosed after hospital discharge in order to make valid comparisons between different institutions and periods of time.