RESUMEN
BACKGROUND: A rehabilitation hospital has the characteristics of longer hospital stay, a variety of medical and paramedical personnels involved in treatment, and more probability of the primary care giver being a nonmedical person. The purpose of this study is to describe the characteristics of nosocomial infection in a rehabilitation hospitals METHODS: Surveillance was done in a rehabilitation hospital affiliated university hospital from June to August 1996, May to June 1997, and July to September 1999. Nosocomial infection was investigated by an infection control practitioner (JCP) according to the Center for Disease Control and Prevention (CDC)'s definitions visiting the rehabilitation wards regularly. RESULTS: The overall NI rates were 3.83 in 1996, 4.42 in 1997 and 1.91 in 1999 per 1,000 patient days, and the urinary tract infection (UTI) rates were 2.82 in 1996, 2.87 in 1997 and 1.07 in 1999 per 1,000 patient days. Both of overall NI rate and UTI rate in 1999 decreased significantly compared with that in 1996 and 1997 (P<0.001). Comparing the ward specific NI rate, the average NI rate of the spinal cord injury ward was the highest (5.67 per 1,000 patient days) and that of the pediatric rehabilitation ward was the lowest (0.83 per 1,000 patient days). UTI constituted of the largest proportion of NIs. The distribution of UTI showed a decreasing tendency from 73.47% in 1996 to 56.00% in 1999, but the distribution of gastrointestinal infection showed an increasing tendency from 4.08% and 5.40% in 1997 and 1996 to 24.00% in 1999. The commonly isolated organisms were E. coli (33.63%), Enterococcus spp. (12.73%) and C. difficile (7.27%). It reflected that the main infection was UTI. CONCLUSION: The main NI in rehabilitation hospital was the UTI. It means that the infection control program of rehabilitation hospital should be focused on the control of UTI. Surveillance in the rehabilitation hospital enabled us to describe the characteristics of nosocomial infection in the rehabilitation hospital. NI rate was decreased through continuous communications and educations with factual datum and involvement of various medical and paramedical personnels. These results show that NI surveillance is crucial for an efficient infection control program.