Discussion

Variability in prescribing practices reflects an opportunity to optimize antibiotic prescribing. ^{Reference Gerber, Prasad and Russell Localio6} Although further investigation into the indication for therapy is needed, evaluating duration of therapy and selection of antibiotics can help to identify specific areas for detailed review. For example, anti-MRSA antibiotic use contributed to 50% or greater of the antimicrobial DDD per jail (Fig. 1), with 10 days of therapy as the common mode for many jails (Supplementary Fig. 1). For uncomplicated skin and soft-tissue infections, a common indication for oral anti-MRSA treatment, shorter courses of therapy (5–7 days) are recommended. ^{Reference Stevens, Bisno and Chambers7} Therefore, there may be an opportunity to educate clinicians on the selection of agents within this setting.

Contrary to jails, the federal prison system has developed strategies aimed at improving antibiotic prescribing, including implementation of antimicrobial stewardship programs (ASPs). ^{Reference Long, LaPlant and McCormick8} Although ASPs may exist at jails or state prisons, we could find no publicly available data about the process or outcomes. Better understanding antibiotic prescribing practices in the carceral healthcare setting comes with an additional urgency owing to the significant health equity and public health concerns associated with the incarcerated population. The carceral system is the primary health provider to a large portion of the United States, with disproportionate incarceration of poor, minoritized, and marginalized communities. ^{Reference Kajeepeta, Mauro, Keyes, El-Sayed, Rutherford and Prins9} Carceral facilities in the United States have high rates of multidrug-resistant infections, and prudent antibiotic use in these settings is highly relevant. ^{Reference Malcolm10}

This study had several limitations. Jail healthcare delivery models vary by county in Massachusetts (ie, direct oversight vs medical vendor oversight), and this was not considered in our analysis. Because our data unit was deidentified prescriptions, we could not ascertain nor control for how pre-existing conditions affected antibiotic utilization nor evaluate the appropriateness of antibiotic prescriptions.

Despite these limitations, our research is novel as the first publication evaluating variability in prescribing practices among jails within the same state system. Having established this baseline of antibiotic use, our next steps will include analyzing disaggregated data to identify underlying trends and appropriateness of prescriptions. Our goal is to work with key stakeholders in jail and prison healthcare to promote integration of antimicrobial stewardship practices, decrease exposure to unnecessary antibiotics, and decrease the development of resistant bacteria.