942
Efficacy of Single-Dose versus Seven-Day Trimethoprim Treatment of Cystitis in Women: A Randomized Double-Blind Study Eva Osterberg, Hans Aberg, Hans O. Hallander, Anders Kallner, and Arne Lundin
From the Departments of Family Medicine. Clinical Chemistry, and the Medicine and Research Center. Karolinska Institute and Huddinge Hospital. and the National Bacteriological Laboratory, Stockholm, Sweden
Single-dose trimethoprim treatment of cystitis was compared with 7-day treatment in a randomized double-blind study in primary health care. Consecutive female patients (613) with symptoms oflower urinary tract infection (UTI) and positive bacteriuria screening tests were enrolled. In 502 cases UTI was confirmed by urine culture. Follow-up was performed twice, after 2-3 and 5-6 weeks. Short-term efficacy could be evaluated in 425 cases and accumulated efficacy in 344. Short-term efficacy was 82% for single-dose and 94% for 7-day treatment (P < .001). Accumulated efficacy was 71% for single-dose and 87% for 7-day therapy (P < .001). Fewer adverse reactions were noted with single-dose therapy (not significant). The cure rate for UTI caused by P-fimbriated Escherichia coli was not different from that of other E. coli infections. Infections with Staphylococcus saprophyticus showed a lower cure rate than E. coli infections with the singledose regimen (P < .05 for short-term efficacy).
Several trials have shown no statistically significant difference in cure rate between 7-day treatment of urinary tract infection (UTI) and 3-day [1, 2] or single-dose [3-6] treatment. Other studies, however, have shown single-dose therapy to be less effective than a conventional regimen [7-9]. Reviews [10, 11] of several single-dose trials have noted severe methodologic errors: The statistical power of most studies is too low «80 %), double-blind studies are few, and definitions of cure and failure are unclear. Pooled results from 14 studies [11] showed that single-dose treatment with trimethoprim-sulfamethoxazole (cotrimoxazole) probably is comparable to 7-day treatment (power 15 years old seeking medical attention because of symptoms indicative of UTI and having bacteriuria determined by chemical screening tests (see below) were enrolled. The same patient could participate in the study more than once if the former infection was eradicated according to culture and clinical picture and ~5 weeks had passed since onset of therapy. Exclusion criteria were: (I) kidney lesions with reduced renal function, urine outflow obstruction, two or more previous episodes of acute pyelonephritis, (2) temperature ~38.ooC, (3) pregnancy, (4) intolerance to trimethoprim, and (5) other unforeseen medical circumstances (e.g., malignant disease) that might make single-dose treatment unsuitable.
Design The study was carried out at four community health centers in Stockholm. Each patient brought a urine specimen I h before seeing a physician. If possible, the urine was voided after incubation for 4 h or overnight. Each patient completed a form concerning ac-
JID 1990;161 (May)
943
Single-Dose Trirnethoprim Treatment of UTI
tual bladder incubation time, symptoms and duration, prior UTIs, and kidney disease. At the health center laboratory, urine nitrite (U nitrite) and the bacterial adenosine triphosphate concentration in urine (U ATP, see below) were estimated and CLED agar plates (Oxoid, London, UK) were inoculated. Laboratory personnel entered the screening results on the form. Results were interpreted as: ATP 50 nmolll or ATP >10 nmol/I plus a positive nitrite test = bacteriuria. When ambiguous results were obtained, the recommendation was to postpone a treatment decision, if possible, until the next day when the culture result would be available. The treatment decision, however, was left to the patient's physician, even if screening was negative. The physician had access to the initial treatment form but completed another more comprehensive form. Patients judged to have cystitis were randomized and in a doubleblind fashion assigned to treatment for I week with either a singledose of 320 mg of trimethoprim and then placebo or 160 mg of trimethoprim twice a day. After 2 and 5 weeks the patient brought a urine sample for analysis and completed a form describing symptoms during and after the course of treatment. At the first follow-up the doctor also interviewed the patient (by telephone, using a standardized form) regarding duration of symptoms, completion of therapy, and adverse drug reactions. At the first visit, the patient was told to call the nurse if symptoms did not disappear in a few days or recurred. If symptoms persisted or recurred, the first or second follow-up was performed early. If screening results or culture indicated UTI, the patient received new treatment and was not studied further. Otherwise she was encouraged to continue and return for follow-up according to the original schedule and results of premature follow-upwere excluded from evaluation.
Laboratory Methods Rapid methods/or urine analysis. Bacterial U ATP was assayed by bioluminescence [15]. U nitrite was demonstrated by Nitur-Test (Boehringer-Mannheim, FRG). Bacteriologic methods. Culture was performed at the health center by inoculating two CLEO agar plates using a 1-S-t1 and a IO-JLI loop, respectively. After overnight incubation at 37°C the plates were transported to the National Bacteriological Laboratory (NBL), Stockholm, for final evaluation. Urine specimens with no growth but an ATP value >10 nmol/l were also cultured anaerobically. All strains of E. coli were frozen at -70°C. The P-fimbriated E. coli were identifiedeither by the P-specific particle agglutination test [16], which was supplied for scientific purposes by Stefan Svensson (NBL), or by commercially available reagents from Kabi Diagnostics (Stockholm). In cases of recurrence, E. coli strains were compared by serotyping performed at the NBL or Institute of Clinical Microbiology, Gothenburg University. Susceptibility to trimethoprim was tested by the disk diffusion technique according to Ericsson et al. [17]. The break point for bacterial resistance to trimethoprim was set at 4 mg/I [18]. Definitions Cystitis was defined as a condition with typical symptoms: dysuria, urgency or frequency, suprapubic discomfort, temperature
Efficacy of Single-Dose versus Seven-Day Trimethoprim Treatment of Cystitis in Women: A Randomized Double-Blind Study Eva Osterberg, Hans Aberg, Hans O. Hallander, Anders Kallner, and Arne Lundin
From the Departments of Family Medicine. Clinical Chemistry, and the Medicine and Research Center. Karolinska Institute and Huddinge Hospital. and the National Bacteriological Laboratory, Stockholm, Sweden
Single-dose trimethoprim treatment of cystitis was compared with 7-day treatment in a randomized double-blind study in primary health care. Consecutive female patients (613) with symptoms oflower urinary tract infection (UTI) and positive bacteriuria screening tests were enrolled. In 502 cases UTI was confirmed by urine culture. Follow-up was performed twice, after 2-3 and 5-6 weeks. Short-term efficacy could be evaluated in 425 cases and accumulated efficacy in 344. Short-term efficacy was 82% for single-dose and 94% for 7-day treatment (P < .001). Accumulated efficacy was 71% for single-dose and 87% for 7-day therapy (P < .001). Fewer adverse reactions were noted with single-dose therapy (not significant). The cure rate for UTI caused by P-fimbriated Escherichia coli was not different from that of other E. coli infections. Infections with Staphylococcus saprophyticus showed a lower cure rate than E. coli infections with the singledose regimen (P < .05 for short-term efficacy).
Several trials have shown no statistically significant difference in cure rate between 7-day treatment of urinary tract infection (UTI) and 3-day [1, 2] or single-dose [3-6] treatment. Other studies, however, have shown single-dose therapy to be less effective than a conventional regimen [7-9]. Reviews [10, 11] of several single-dose trials have noted severe methodologic errors: The statistical power of most studies is too low «80 %), double-blind studies are few, and definitions of cure and failure are unclear. Pooled results from 14 studies [11] showed that single-dose treatment with trimethoprim-sulfamethoxazole (cotrimoxazole) probably is comparable to 7-day treatment (power 15 years old seeking medical attention because of symptoms indicative of UTI and having bacteriuria determined by chemical screening tests (see below) were enrolled. The same patient could participate in the study more than once if the former infection was eradicated according to culture and clinical picture and ~5 weeks had passed since onset of therapy. Exclusion criteria were: (I) kidney lesions with reduced renal function, urine outflow obstruction, two or more previous episodes of acute pyelonephritis, (2) temperature ~38.ooC, (3) pregnancy, (4) intolerance to trimethoprim, and (5) other unforeseen medical circumstances (e.g., malignant disease) that might make single-dose treatment unsuitable.
Design The study was carried out at four community health centers in Stockholm. Each patient brought a urine specimen I h before seeing a physician. If possible, the urine was voided after incubation for 4 h or overnight. Each patient completed a form concerning ac-
JID 1990;161 (May)
943
Single-Dose Trirnethoprim Treatment of UTI
tual bladder incubation time, symptoms and duration, prior UTIs, and kidney disease. At the health center laboratory, urine nitrite (U nitrite) and the bacterial adenosine triphosphate concentration in urine (U ATP, see below) were estimated and CLED agar plates (Oxoid, London, UK) were inoculated. Laboratory personnel entered the screening results on the form. Results were interpreted as: ATP 50 nmolll or ATP >10 nmol/I plus a positive nitrite test = bacteriuria. When ambiguous results were obtained, the recommendation was to postpone a treatment decision, if possible, until the next day when the culture result would be available. The treatment decision, however, was left to the patient's physician, even if screening was negative. The physician had access to the initial treatment form but completed another more comprehensive form. Patients judged to have cystitis were randomized and in a doubleblind fashion assigned to treatment for I week with either a singledose of 320 mg of trimethoprim and then placebo or 160 mg of trimethoprim twice a day. After 2 and 5 weeks the patient brought a urine sample for analysis and completed a form describing symptoms during and after the course of treatment. At the first follow-up the doctor also interviewed the patient (by telephone, using a standardized form) regarding duration of symptoms, completion of therapy, and adverse drug reactions. At the first visit, the patient was told to call the nurse if symptoms did not disappear in a few days or recurred. If symptoms persisted or recurred, the first or second follow-up was performed early. If screening results or culture indicated UTI, the patient received new treatment and was not studied further. Otherwise she was encouraged to continue and return for follow-up according to the original schedule and results of premature follow-upwere excluded from evaluation.
Laboratory Methods Rapid methods/or urine analysis. Bacterial U ATP was assayed by bioluminescence [15]. U nitrite was demonstrated by Nitur-Test (Boehringer-Mannheim, FRG). Bacteriologic methods. Culture was performed at the health center by inoculating two CLEO agar plates using a 1-S-t1 and a IO-JLI loop, respectively. After overnight incubation at 37°C the plates were transported to the National Bacteriological Laboratory (NBL), Stockholm, for final evaluation. Urine specimens with no growth but an ATP value >10 nmol/l were also cultured anaerobically. All strains of E. coli were frozen at -70°C. The P-fimbriated E. coli were identifiedeither by the P-specific particle agglutination test [16], which was supplied for scientific purposes by Stefan Svensson (NBL), or by commercially available reagents from Kabi Diagnostics (Stockholm). In cases of recurrence, E. coli strains were compared by serotyping performed at the NBL or Institute of Clinical Microbiology, Gothenburg University. Susceptibility to trimethoprim was tested by the disk diffusion technique according to Ericsson et al. [17]. The break point for bacterial resistance to trimethoprim was set at 4 mg/I [18]. Definitions Cystitis was defined as a condition with typical symptoms: dysuria, urgency or frequency, suprapubic discomfort, temperature
