Aim: To evaluate the bacterial contamination of the water from the micro¯ora of pressure ulcers in para-and tetraplegic patients, when they were exercising in water at 368C for half an hour. Material: Twelve spinal cord lesioned (SCL) patients with ulcers participated, and six of the 12 SCL patients with healed ulcers constituted a control group. Methods: The evaluation was performed both with and without the ulcer covered with a moisture reactive occlusive dressing, DuoDERM 1 . Bacterial samples were taken from the bath-water before and after the exercise programmes and additionally specimens were obtained from the ulcer, the patients skin and urine, and from the skin of the physiotherapist who exercised the patient. A similar procedure was carried out in the controls. Results: The bacteriologic analyses showed no signi®cant dierence in the contamination of the bath-water after exercising with or without DuoDERM 1 covering the ulcers. In half of the patients DuoDERM 1 loosened. After all exercise programmes with or without DuoDERM 1 dressing the water was contaminated with facultative aerobic intestinal bacteria, ie E. faecalis and Enterobacteriaceae (E. coli, Klebsiella species, Proteus species, Enterobacter species). In nearly one-third of the exercise sessions the bath-water was contaminated with P. aeruginosa before starting, and after the exercise programmes one fourth of the ulcers were colonized with these pathogens. Conclusion: The water specimen showed the bacteria from the intestine to be much more prominent than the bacteria coming from the ulcers. Thus the pressure ulcers were of minor importance for the bath-water and ought not to prohibit patients from the potential bene®ts of water exercise, but chlorination of the water in the training pool seems appropriate.

Aim: To investigate the positive and negative predictive values for bacteriuria of a rapid chemical dipstick procedure for leukocytes and nitrite, and a microscopic examination for leukocytes and bacteria with a urine culture being the reference in patients with spinal cord lesion (SCL). Methods: A prospective study. Signi®cant bacteriuria was de®ned as 510 5 CFU/mL. The microscopic examination for leukocytes was divided in four dierent cut-o limits for positive results. Material: A total of 256 consecutive early morning urine samples were collected from 143 SCL patients admitted to our in-patient facility. Results: One hundred and twenty-eight urine cultures revealed signi®cant bacteriuria. Eightyseven were infected by only one microorganism, 41 samples contained signi®cant growth of more than one species. A total of 186 microorganisms in signi®cant growth were identi®ed. Conclusion: Three or more leukocytes should be considered as a positive result. The dipstick and microscopy tests are equally valuable, considering the single tests as well as the combinations evaluated.