From 2018 to 2024, a total of 217 Massachusetts LTC facilities submitted at least 1 month of AS data, representing 12 518 102 resident-days. The median number of reporting facilities per month was 53 (interquartile range [IQR], 46–65), and participation increased from 18.8% to 24.2% of all LTC facilities statewide (Figure 1). Among participating facilities, the median bed size was 118 (IQR, 81–137); 96.1% provided long-term general nursing care, 75.4% reported hospice or palliative care services, and 2.2% offered ventilator services (Table 2). Short-stay residents comprised 22.4% of the total census among facilities reporting both short- and long-term populations each month.

Over the study period, the overall AS rate increased by 7%, from 7.22 to 7.70 starts per 1000 resident-days (Figure 2). The most frequently prescribed antibiotic classes were penicillin/beta-lactamase inhibitors, first- and third-generation cephalosporins, fluoroquinolones, and tetracycline, which together accounted for 62.7% of all AS. Fluoroquinolone use declined by 36% from 1.68 to 1.08 starts per 1000 resident-days, while beta-lactam use increased by 26%, from 2.64 to 3.33 starts per 1000 resident-days (Figures 3 and 4).

Interrupted time-series analysis showed an immediate significant increase (step change) in beta-lactam start rates following the October 2022 implementation of LTC office hours webinars and the expanded benchmarking reports (P < .001). No specific intervention was significantly associated with the observed decline in fluoroquinolone prescribing. Segmented regression plots illustrating level and slope changes for beta-lactam and fluoroquinolone AS rates are shown in Figures 3 and 4, with corresponding parameter estimates presented in Supplementary Tables 2 and 3.

In a subgroup analysis, among the 217 participating facilities, 86 (39.6%) achieved honor roll recognition based on reporting frequency: 58 (26.7%) achieved gold-tier status, 14 (6.5%) silver, and 14 (6.5%) bronze. Honor roll facilities had a median total AS rate of 6.5 (IQR, 4.1–8.8) per 1000 resident-days compared with 5.7 (IQR, 3.7–8.1) among nonhonoree facilities (P = .005). Median beta-lactam AS rates were 2.8 (IQR, 1.8–3.7) versus 2.3 (IQR, 1.4–3.3) per 1000 resident-days (P < .001), and median fluoroquinolone AS rates were 1.0 (IQR, 0.6–1.7) versus 0.9 (IQR, 0.5–1.5) per 1000 resident-days (P = .032) for honoree and nonhonoree facilities, respectively.

This statewide benchmarking initiative offers important insights into how antimicrobial stewardship can be operationalized across LTC facilities through sustained public health–academic collaboration and multiple, complementary intervention strategies. Beyond numerical trends, its success highlights the value of embedding stewardship within a continuous and data-driven feedback loop that integrates reporting, benchmarking, education, and recognition. Prior LTC stewardship interventions often relied on 1-time educational sessions or bundled strategies without mechanisms for sustainability and therefore could not be readily replicated in routine practice [12–16, 28]. In contrast, the Massachusetts program has been sustained for over 6 years by embedding stewardship activities within the state public health infrastructure. Unlike earlier national efforts, such as the Agency for Healthcare Research and Quality (AHRQ) Safety Program, which delivered a successful but time-limited educational intervention with benchmarking and implementation support [23]. That initiative operated as a centrally coordinated project without ongoing alignment with local health systems or public health priorities. The Massachusetts experience demonstrates how stewardship activities can be sustained through continued, iterative engagement.

In this program, facility-submitted AS data were transformed into individualized benchmarking reports and supported by real-time public health engagement through educational webinars and recurring LTC office hours, which functioned as peer-learning forums. These components were implemented sequentially and expanded over time, with overlap across program periods, precluding attribution of observed changes to any single intervention. Rather, the program reflects a longitudinal, evolving approach in which layered strategies collectively supported stewardship engagement and practice change.

Although the overall AS rate increased over the study period, this finding should be interpreted cautiously and in context. Reporting was voluntary and incomplete, and participation varied over time, raising the possibility that observed changes in AS rates reflect shifts in the composition of participating facilities rather than direct effects of specific interventions. Increases in reporting during later program periods coincided with expanded educational activities and LTC office hours. Despite these limitations, we observed meaningful changes in prescribing patterns, including a substantial decline in fluoroquinolone use and a concurrent increase in beta-lactam prescribing. This pattern is consistent with the intended stewardship outcomes reflecting safer, guideline-concordant prescribing practices. The increase in beta-lactam use may also reflect improved provider confidence, supported by program resources that encouraged appropriate antibiotic selection and discouraged reliance on high-risk agents, such as fluoroquinolones, which are associated with clinically significant adverse effects. Similar class-level trends have been reported in other LTC stewardship studies [12, 28].

The program’s ability to promote engagement outside of formal regulatory enforcement is notable. Although antimicrobial stewardship requirements for nursing homes have been mandated by CMS since 2017 and are assessed through state survey processes, the AS Reporting Program operated separately from regulatory inspection and enforcement activities [29]. Participation was voluntary and not tied to licensure, reimbursement, or survey outcomes. Within this context, the honor roll system provided a visible, low-cost recognition framework that leveraged peer comparison rather than penalties. This aligns with findings from the AHRQ Safety Program for improving antibiotic use, in which LTC facilities were more motivated by shared learning and recognition than top-down mandates [28]. Reporting participation temporarily declined during the Coronavirus disease 19 public health emergency, consistent with CDC reports of stewardship program disruptions in LTC settings [30]. However, the program infrastructure and stewardship activities remained active, and participation rebounded as facilities recovered operational capacity, highlighting the importance of LTC-specific public health infrastructure in supporting stewardship continuity and recovery.

Facilities achieving honor roll recognition reported higher AS rates, which may reflect more complete and consistent reporting over time. Honor roll facilities may therefore represent a more engaged subset of participants, and differences in prescribing patterns between honoree and nonhonoree facilities may reflect differential responsiveness to stewardship activities rather than uniform effects across all facilities. As such, improvements observed at the program level may be driven in part by highly engaged facilities. Differences were most pronounced for beta-lactam prescribing, while differences in fluoroquinolone use were small in absolute terms (increased by 0.1 per 1000 resident-days). Collectively, these class-specific and engagement-related patterns suggest that sustained participation may amplify stewardship impact, even if overall AS rates have not uniformly declined.

Despite these successes, several limitations should be acknowledged. First, the voluntary nature of participation may introduce selection bias, as facilities with stronger leadership engagement or existing stewardship culture may have been more likely to participate, potentially limiting generalizability. Second, antibiotic appropriateness was not assessed due to the absence of resident-level clinical data. Although reductions in high-risk agents are generally desirable, increases in beta-lactam use should be interpreted cautiously without diagnostic context or outcome data. This limitation has since been addressed through updates to the survey tool that capture treatment indication and days of therapy, enabling more precise stewardship evaluations in future analyses. Third, some data submissions were incomplete or inconsistently reported. Although thresholds were applied to exclude implausible entries, future efforts should incorporate automated data pipelines via electronic health records. As an ecological design, causal inference is limited. While ITS methods accounted for baseline trends, unmeasured confounders and concurrent events may have influenced results. Finally, the honor roll’s influence on behavior versus its reflection of preexisting motivation warrants further exploration.

In summary, the Massachusetts Antibiotic Start Reporting Program demonstrates that antimicrobial stewardship in LTC is feasible through sustained public health–academic collaboration and a multicomponent approach. Its success derives from the combined impact of benchmarking, feedback, education, and recognition implemented over time, rather than from any single discrete intervention. Observed changes occurred in the context of overlapping strategies, and antibiotic prescribing appropriateness could not be directly assessed in this analysis. Nonetheless, this experience offers practical lessons for engaging LTC facilities at scale and underscores the critical role of public health infrastructure in supporting stewardship implementation and sustainment in resource-limited settings.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.