Antibiotic Audit Letters Sent to Primary Care Docs Tied to Clinical, Economic Value

A mail-based audit and feedback (A&F) program to improve antibiotic prescribing among primary care physicians was associated with both clinical and financial returns, according to an economic evaluation of data from a Canadian randomized trial.

Among 4,879 primary care physicians in Ontario, every $1 spent mailing an antibiotic A&F letter to physicians led to $8.82 in savings related to inappropriate antibiotic prescriptions, said Kednapa Thavorn, PhD, of the Ottawa Hospital Research Institute, and colleagues.

The intervention’s estimated per-physician cost was $5.50, while average per-physician spending on antibiotic prescriptions among those in the intervention group fell significantly by $55.00 (P<0.001) -- generating a total monetary savings of $43.03 per physician, the researchers reported in JAMA Network Open.

“Most savings were driven by reductions in adverse events linked to unnecessary antibiotic use, highlighting both the economic and patient safety benefits of the A&F program,” Thavorn and colleagues wrote. “These findings support the integration of antibiotic A&F interventions into routine primary care as a scalable strategy to advance antimicrobial stewardship, improve prescribing practices, and enhance the sustainability of healthcare delivery.”

At least a quarter of antibiotics prescribed in primary care may be unnecessary, they noted, prompting a host of antimicrobial stewardship initiatives that emphasize evidence-based prescribing guidelines. While research has consistently shown such programs’ clinical value, data supporting their cost-effectiveness are scarce, they added.

Return on investment (ROI) would likely rise if the intervention included more physicians, Thavorn told MedPage Today. “Scaling up such an intervention to the broader primary care physician community is even cheaper.” If all 40,000 Canadian primary care physicians were enrolled in the A&F program, the ROI would likely rise to $16.82 for every dollar spent.

“Hopefully, this will provide a very convincing message to any healthcare system that may be tempted to try this type of intervention that it’s not just simple, but it’s also very cost-effective,” Thavorn said.

Of the 4,879 physicians, 3,909 received a mailed letter comparing their antibiotic prescribing rate with their peers and 970 received no letter. Mean time since medical school graduation was 25 years, and 55.6% of physicians were male.

Prescription data were drawn from patients ages 65 years or older. The researchers measured physicians’ antibiotic prescriptions in the 6 months after the A&F intervention, from January through June 2022.

The intervention’s costs included administrative time to code and customize prescription-pattern letters, as well as printing and mailing costs. To assess potential benefits, the researchers quantified the monetary value of prescriptions, adverse events, and undertreatment harms. Adverse-event costs included ambulatory visits and hospitalizations from drug-related adverse events, while undertreatment harms included hospitalizations from infections.

The study’s model assessed the likelihood of antibiotic-related adverse events among patients prescribed antibiotics, as well as the likelihood of harms from undertreatment among patients who weren’t prescribed antibiotics.

Over the program’s 6-month assessment period, Canadian government spending on antibiotic prescriptions was $520 per physician among the group that received the A&F letter and $586 per physician among the group that didn’t, a significant $55 difference (P<0.001). On average, the intervention led to three fewer prescriptions per 1,000 patients, and to significant cost savings for prescriptions of nitrofurantoin, amoxicillin-clavulanic acid, azithromycin, cefuroxime, clarithromycin, trimethoprim, cloxacillin sodium, and linezolid.

Adverse event-related healthcare utilization costs were $960 lower per physician on average among physicians in the intervention group, though the difference wasn’t significant (P=0.40). Undertreatment-related costs were also $404 lower per physician on average in the intervention group, another nonsignificant difference (P=0.94).

Study limitations included an inability to determine each prescription’s appropriateness. Incremental cost-estimate confidence intervals around antibiotics, adverse events, and harms from undertreatment were wide. In addition, the trial’s inclusion only of people ages 65 and older may reduce the findings’ generalizability to younger populations.