Current issues of ACP Journal Club are published in Annals of Internal Medicine


Gentamicin plus ampicillin and gentamicin plus trimethoprim-sulfamethoxazole were equally effective in patients with non-ampicillin-resistant acute pyelonephritis

ACP J Club. 1991 July-Aug;115:7. doi:10.7326/ACPJC-1991-115-1-007

Source Citation

Johnson JR, Lyons MF II, Pearce W, et al. Therapy for women hospitalized with acute pyelonephritis: a randomized trial of ampicillin versus trimethoprim-sulfamethoxazole for 14 days. J Infect Dis. 1991 Feb;163:325-30.



To compare gentamicin plus ampicillin with gentamicin plus trimethoprim-sulfamethoxazole (TMP-SMZ) for the treatment of acute, uncomplicated pyelonephritis in women.


Randomized trial with blinding of laboratory personnel.


A military hospital and a city-county hospital in Washington state.


All women ≥ 18 years of age hospitalized for intravenous (IV) therapy for suspected acute uncomplicated pyelonephritis were randomized if they had costovertebral pain or tenderness, microscopic pyuria or bacteriuria, and a positive bacterial culture. The ampicillin and TMP-SMZ groups comprised 30 and 40 patients, respectively, after patients with ampicillin-resistant pathogens were excluded.


All patients were initially given ≥ 72 hours of IV therapy: gentamicin (every 8 h) and either ampicillin (1 g every 6 h) or TMP-SMZ (160 mg/800 mg, every 12 h). Subsequent therapy was given orally: ampicillin, 500 mg 4 times per day; or TMP-SMZ, 160/800 mg 2 times per day, for a total (IV and oral) of 14 days.

Main outcome measure

Recurrent bacteriuria confirmed by quantitative cultures from midstream-voided urine specimens.

Main results

Of all patients participating in the study, 29 (34%) were infected with an organism resistant to ampicillin, but only 1 patient (1%) had a pathogen resistant to TMP-SMZ (P < 0.001 for difference in the number of patients that required change in therapy because of resistance). 14 patients (32%) in the ampicillin group with ampicillin-resistant organisms were excluded from further follow-up. 90% of patients in each treatment group were evaluated 7 days after completing therapy. Recurrent bacteriuria was identified in 2 patients (7%) on ampicillin and 1 patient (3%) on TMP-SMZ, in all cases caused by re-infection with a new organism (P = 0.39). In total, recurrent bacteriuria occurred in 11% and 8% of patients on ampicillin and TMP-SMZ, respectively, who returned for post-therapy visits (CI for the difference in response 0% to 18%). Adverse reactions were equally common in the 2 groups. (33% of patients in HMP + GM group vs 32% of patients in trimethoprim-sulfamethoxazole + gentamicin group, P = 0.94)


Urine cultures from patients with acute, uncomplicated pyelonephritis were much more likely to grow organisms resistant to ampicillin than to TMP-SMZ. When patients with ampicillin-resistant organisms were excluded, ampicillin and TMP-SMZ were equally effective when given with gentamicin, and, after therapy for 14 days with either regimen, episodes of recurrent bacteriuria reflected reinfection rather than relapse.

Source of funding: Burroughs-Wellcome, Inc.

Address for article reprint: Dr. W.E. Stamm, Infectious Diseases ZA-89, Harborview Medical Center, 325 9th Avenue, Seattle, WA 98104, USA


This well-designed and well-executed investigation supports the idea that ampicillin is an inadequate empiric therapy for acute pyelonephritis. Escherichia coli resistance to ampicillin is widespread, and this drug is no longer recommended for empiric therapy of acute urinary tract infections. The study validates the current practice standards for the clinical management of women hospitalized for acute, apparently uncomplicated, pyelonephritis (1, 2): Initially prescribe empiric therapy with an antimicrobial agent that is reliably effective against gram-negative bacilli (e.g., an aminoglycoside, or a quinolone such as ciprofloxacin); later, when in vitro susceptibility tests become available, change to the drug that is most effective, least expensive, least toxic, and easiest to take, which usually will be trimethoprim-sulfamethoxazole.

This investigation, as the authors note, again raises the question of how long such patients need to be treated. They refer to evidence that short courses of aminoglycosides, with high and persistent renal parenchymal levels, may be curative. Other information supports adherence to the current practice standards of 14 days of therapy for acute pyelonephritis. Further investigations may establish, however, that oral therapy is not needed if intravenous aminoglycosides have been given.

Stephen R. Jones, MD
Good Samaritan Hospital and Medical CenterPortland, Oregon, USA


1. Sobel JD, Kaye D. Urinary tract infections. In: Mandell GL, Douglas RG, Bennett JE, eds. Principles and Practice of Infectious Diseases. 3d ed. New York: Churchill Livingstone Inc.; 1990:596.

2. Ward T, Jones SR. Genitourinary tract infections. In: Reese RE, Betts RF, eds. A Practical Approach to Infectious Diseases. 3d ed. Boston: Little, Brown & Co.; 1991:370-2.