Caution on mandatory public reporting

To the Editor—Gonzalo Bearman’s otherwise excellent commentary suggests that mandatory public reporting of risk-adjusted patient infections offers important value, citing 2 references.¹ Having run one state’s mandatory program as a research opportunity, I feel compelled to raise a yellow card on that score. Mandatory public HAI reporting to accelerate safety improvement is a noble experiment, but it remains of unproven value and cost-effectiveness. One of Bearman’s 2 references, a systematic review with meta-analysis,² derives its statistical significance from cardiac surgery mortality reporting but includes only 1 study regarding HAI, a study that finds no impact of public reporting on hospital infection rates. His other reference³ finds that hospitals in states newly enacting HAI reporting mandates soon demonstrated greater reduction in CLABSI rates but later no greater reduction than what was seen in states without mandates. Given the cliché that “data unites, theories divide,” there are 3 possible interpretations. First, legislative mandate could motivate change. For example, Marsteller et al³ notes that at baseline hospitals in states with new or impending legislation started with higher CLABSI rates than hospitals in states without a mandate and were more likely to then adopt well-known prevention strategies. Second, legislative mandates do not impact performance. Several studies fail to find statistically significant association, so a single positive signal could be the result of random chance variation or bias. Third, some could appear to be doing better than others simply due to widely ranging rates of underreporting. Standardized methods that are practical, sustainable, and internationally credible for ongoing assurance of reliable quality exist that can be used for annual validation to confirm hospitals meet predefined sensitivity and specificity requirements in their data reporting, but the vast majority of American state HAI programs have performed no credible on-going validation.⁴

Together with colleagues across the 10 academic domains needed to address a sequence of research questions leading to understand what works, for whom, in what settings (Fig. 1), we used one state’s mandatory HAI program to seek answers.⁵ Essentially, all participating hospitals continued to exceed our annual validation requirements for high-quality reporting, and all maintained low HAI rates, which were not affected by adding reporting requirements. Risk-stratified rates were more meaningful and accurate indicators of performance than risk-adjusted ratios.^4,6 And as others have reported about public reporting websites, the general public showed little evidence of using such websites to actually influence their care decisions.⁷ Today, as before, “More research is needed to better understand what health care consumers need on the WWW to support their decision making involving HAIs.”⁸

Figure 1. Improving our understanding of what works, for whom, in what settings. From Figure 1 in Birnbaum D, The US experience with mandatory public reporting. Clin Govern Int J 2012;17:113–123.