U p to 25% of hospitalized patients will have an indwelling urinary catheter inserted (1). While most of these catheters are required for optimal patient care, about one in five is unnecessary (2). Catheter-related urinary tract infections are the leading cause of nosocomial infection. They account for up to 40% of hospital-acquired infections, with an incidence of 3% to 10% per day of indwelling catheterization (1,(3)(4)(5). Bacteremia occurs in 1% to 4% of those who develop nosocomial bacteriuria (6). Most patients also find an indwelling urinary catheter to be uncomfortable and activity restricting (7). Given these potential morbidities, it is remarkable how infrequently the use of a urinary catheter is documented by a physician's order in the medical record (8).The strongest predictor for catheter-associated bacteriuria is the duration of catheterization (3,4). Thus, methods to shorten this period should reduce the risk of nosocomial urinary tract infection, yet physicians are often unaware that their own hospitalized patients have a urinary catheter in place (9). Furthermore, these "forgotten" catheters were about twice as likely to be unnecessary as those remembered by the physician (9). Thus, a system that would remind physicians which of their patients had urinary catheters might shorten the duration of catheterization.One such approach involves using automatic reminders to physicians through computerized medical record systems (10 -15). At our Veterans Affairs (VA) medical center, nearly all patient records and orders are entered into a computer. We hypothesized that a computer-based order for inserting an indwelling urinary catheter, combined with computer-generated reminders to remove the catheter, would improve documentation of urinary catheters, alert physicians that the catheter was in place, and encourage discontinuing catheterization that was no longer required.
METHODSAll inpatients at the VA Puget Sound Health Care System are cared for by teams of resident physicians supervised by University of Washington faculty. Patients on the medicine and cardiology services are predominantly housed on two hospital wards (second and fourth floor), with most of each team's patients on a single ward. During the first 8 weeks of the study, the fourth floor served as the study ward and the second floor served as the control ward; during the second 8 weeks, the second floor became the study ward and the fourth floor became the control.At the beginning of each month, a training session is held to teach incoming residents about the computer system. During this session, one of the investigators (PBC) explained the purpose of the study, including the direct association between duration of catheterization and catheter-associated urinary tract infection, and instructed residents on use of the computerized urinary catheter study order. We also explained the purpose of the study to the ward nurses and encouraged them to remind physicians to place a computerized order for each indwelling urinary catheter. Physicians whos...