Effect of a Structured Microbiology Laboratory Report on Antimicrobial Prescribing for Asymptomatic Bacteriuria in Elderly Females (2010)
Description
This project sought to test the hypothesis that ‘withholding’ of antimicrobial sensitivity reporting in conjunction with a selective evidence-based informative comment will result in:
Flinders Medical Centre and South Australia Pathology
Aim
to test the hypothesis that the ‘unjustified’ antimicrobial treatment of asymptomatic bacteriuria in elderly females could be reduced by utilising a modified microbiology report in which full microbial speciation was not performed, antimicrobial sensitivity data was not reported and a short educational comment regarding the likely (or lack of) clinical significance of the result was included in the pathology report.
This aim was achieved by this project.
Outcomes
Urine samples received in the routine diagnostic microbiology laboratory from females aged 70 yrs and over with detected bacteriuria (>105 cfu/ml) were considered for randomisation to either:
full reporting of isolate at the species level with an ‘appropriate’ choice of oral and intravenous antibiotics reported. As a minimum, sensitivity to ampicillan, cephazolin, timethoprim and gentamicin was reported, with additional antibiotics if first-line drug resistance was present
OR
general reporting of the microbiology result without provision of any microbial sensitivity data, but inclusion of the following comment: ‘Asymptomatic bacteriuria in elderly women does not usually require antibiotic treatment. Sensitivities are available upon request.’
Women were excluded from the trial if:
there was an indication from clinical notes to suggest their bacteriuria was symptomatic or that antimicrobials might be indicated
they were receiving antimicrobials for another clinical indication which had activity against the urinary isolate.
Findings
Rates of antimicrobial prescribing for urinary tract infection (UTI) were 202 of 272 (74%) in the reported group, and 205 of 293 (70%) in the sensitivities masked group.
The subsequent analysis was performed in 470 subjects after exclusion of 95 subjects for reasons not available to the laboratory at the time of randomisation.
The intervention of masking antimicrobial sensitivities and providing educational comment had a small effect on antimicrobial use for UTI.
Sensitivities were only requested in 45% of cases of the group receiving masked reports who were prescribed antibiotics. This finding indicated the practice of not confirming antibiotic sensitivities despite prescribing antibiotics was common.
There was considerable variation in the duration of antimicrobial use.
Antimicrobials were widely used ‘unnecessarily’ in both groups which highlights the importance of targeted educational programs.
A limitation of the study was the considerable number of subjects where antimicrobials were initiated empirically, either for presumptive symptomatic UTI (with no indication of this on the pathology request form) or on the basis of immediate dipstick tests collected on admission.
There was no apparent adverse clinical effect associated with masking antimicrobial sensitivity results.
The mean length of stay for the masked versus reported group were 13.3 days and 14.2 days respectively.
Follow-up urine specimens were collected with similar frequency in both groups.
Educational interventions are beneficial but require significant resources to instigate, may lose effect over time and may not prevent antimicrobial prescribing once a ‘positive’ urine result is obtained.
Antimicrobial use is associated with additional costs, toxicity and the emergence of resistance.
Recommendations
Strategies for reducing unnecessary prescribing for asymptomatic bacteriuria could be directed at reducing the number of specimens initially sent from asymptomatic patients, or providing educational information regarding the significance of the result.
Providing only limited information on the laboratory report, and adding a brief educational comment in an attempt to ‘downplay’ the potential significance of the positive urine culture, could be utilised at essentially no cost and be easily sustained with ongoing resources. Incorporating additional components, such as comprehensive educational programs, will produce the best outcomes.
Educational programs should be targeted to specific groups to empower them to develop confidence in withholding antibiotics for asymptomatic patients.
Educational programs should also target nursing and medical staff to educate them on not obtaining urine for analysis unless the patient is symptomatic.
Automatic stop orders after three or five days of antibiotic use for UTI could be initiated in a hospital setting.
Additional comments such as ‘Recommended duration of antimicrobial therapy for symptomatic, uncomplicated UTI is 5 days for β-lactams and 3 days for trimethoprim/quinolones’ could be also be added as a supplement to laboratory reports.
Areas for Future Consideration
Produce a fact sheet on antibiotic use and bacteriuria for all levels of medical staff, including nurses, to empower them to develop confidence in withholding antibiotics for asymptomatic patients.
