NOSOCOMIAL INFECTIONS IN A NEUROSURGICAL INTENSIVE-CARE UNIT

The occurrence of nosocomial infections in 1017 consecutive patients s een in a neurosurgical intensive care unit (ICU), over a period of 18 months is reported. The frequency of infections is low, which may poss ibly be due partly to the short stay in the ICU. Close interdisciplina ry cooperation is stressed as an important factor in limiting infectio ns. Background. The aim of this study was to analyse the nosocomial in fections in a neurosurgical intensive care unit over a period of 18 mo nths, emphasizing localization and cause of infection, in order to ada pt treatment and to take preventive measures. From 15% to 27% of patie nts treated in ICUs acquire nosocomial infections. In Germany this mea ns 500000-800000 patients a year, and the annual costs related to noso comial infections are estimated at 1.7 billion Deutschmarks. Patients and methods. In all, 1017 consecutive patients were evaluated. The pat ients were divided into two groups, depending on the duration of treat ment in the ICU: Patients who remained for less than 48 h (1017 patien ts) Patients who were treated for a period exceeding 48 h (314 patient s) The evaluation was performed retrospectively from the medical docum entation. Criteria for registration are those of the Centers for Disea se Control (Atlanta 1988). When more than one infection was diagnosed, each was considered as a new infection, regardless of the bacteria in volved. Among the 314 patients who were in the ICU for more than 48 h a total of 114 nosocomial infections were recorded. The frequency of i nfection referred to all patients treated during that time (n = 1017) was 11.2%, while the frequency among those who were treated for longer than 48 h was 36.3%. Most infections (38.6%) affected the respiratory tract, followed by infections of the urinary tract. Of the bacteria d etermined 56.7% were gram-negative. In this group E. coli was the most frequently found (29.8%). In the group of gram-positive bacteria, S. aureus was diagnosed in 56.3% of cases. Twelve (16%) of the infected p atients died and lethality referred to all patients was 8.6%. Discussi on. Compared with other studies, this study revealed a low the infecti on rate, at 11.2%. This can be explained partly by the short stay in t his ICU (mean 3.7 days) and partly by the retrospective method of regi stration and the particular medical characteristics of neurosurgical p atients. The well-known general risk factors for infection, such as ag e, mechanical ventilation, continuous catheterization of the bladder, and long duration of stay, are also found in neurosurgical ICUs. It is quite difficult to determine to what extent nosocomial infections pro long the treatment necessitated by the primary neurosurgical disease. We were not able to extrapolate the influence of immunosuppressant tre atment on the appearance of nosocomial infections, as almost all patie nts in this study were receiving steroids. This study underlines the n ecessity of interdisciplinary cooperation between neurosurgeons, anaes thesiologists, microbiologists and nurses in neurosurgical ICUs, where most patients staying longer than 48 h are immunosuppressed and venti lated and thereby particularly at risk of nosocomial infections.