Risk factors for urinary tract infection.
A case study was conducted among college women using a student health service at a university in southern Michigan to identify and evaluate behavioral factors which may be determinants of urinary tract infection. Women who presented to the student health service with symptoms suggestive of urinary tract infection and found to have pyuria on urinalysis were offered the opportunity to participate in a clinical trial comparing antibiotic regimens in the treatment of their urinary infection. Women with identified structural abnormalities of the urinary tract were not admitted to the study. On their 1st visit, women enrolled in the antibiotic study were asked to complete a questionnaire for the epidemiologic study described here. Midstream specimens were obtained during the 1st visit prior to therapy and were cultured quantitatively using standard methods. A patient with symptoms of acute urinary tract infection was included as a case if the urine culture was positive or probable. 2 different control groups were used to assess risk factors for urinary tract infection. For the 1st control group, women presenting to the student health service with coryza and/or sore throat, an oral temperature of less than 38.9 degrees Centigrade, and a clinical diagnosis of uncomplicated viral upper respiratory infection were asked to complete the same questionnaire as cases. For the 2nd control group, women attending the gynecology clinic at the same student health service for routine scheduled pelvic examinations and/or contraceptive services were asked to complete the same questionnaire. Cases were entered into the study from March through December 1982. The students used as controls participated from March 1982 through March 1983. The subjects completed a standard self-administered multiple-choice questionnaire, asking primarily about activities during the 3 weeks prior to completion of the questionnaire. 47 women with presumptive urinary tract infection participated in the clinical trial. 43 cases of confirmed urinary tract infection were identified, 38 with a positive and 5 with a probable urine culture result. In all, 149 upper respiratory infection controls and 227 gynecology controls participated. The history of previous urinary tract infection was significantly greater in cases than in either of the control groups. 4 variables were included in the final logistic model: coital frequency within 3 weeks, use of the diaphragm within 3 weeks, history of previous urinary tract infection, and age. Coital frequency during the previous 3 weeks was associated strongly with illness, with generally higher risk at higher frequencies. A significant association with the diaphragm was observed in comparison with both control groups. The findings failed to show an association with many of the factors commonly believed to be important such as type of clothing worn and volume of fluids consumed.