We determined if the deliberate establishment of asymptomatic bacteriuria with Escherichia coli 83972 in patients with incomplete bladder emptying and recurrent urinary tract infection protects against recurrence.

In phase 1 of the study the patients were randomized to blinded inoculations with E. coli 83972 or saline. Crossover occurred after monitoring for 12 months or after a urinary tract infection. The outcome was the time to the first urinary tract infection in patients with and without E. coli 83972 bacteriuria. In phase 2 patients were subjected to additional blinded inoculations to extend periods with and without E. coli 83972 bacteriuria. The outcome was the number of urinary tract infections during 12 months with and 12 months without E. coli 83972 bacteriuria.

A total of 20 patients completed the study. In phase 1 the time to the first urinary tract infection was longer with than without E. coli 83972 bacteriuria (median 11.3 vs 5.7 months, sign test p = 0.0129). Phase 2 was analyzed after patients had spent a total of 202 months with and 168 months without E. coli 83972 bacteriuria. There were fewer reported urinary tract infection episodes with vs without E. coli 83972 bacteriuria (13 vs 35 episodes, paired t test p = 0.009, CI 0.31–1.89). There was no febrile urinary tract infection episode in either of the study arms and no significant side effects of intravesical bacterial inoculation were reported.

Deliberately induced E. coli 83972 bacteriuria protected patients with incomplete bladder emptying who are prone to urinary tract infection from recurrent urinary tract infection as demonstrated by the delay in time to urinary tract infection and the decrease in number of urinary tract infection episodes.