Data Validation Must Catch Up With Disclosure in Public Reporting, Expert Says

Public reporting of infection data is one of the significant developments that are causing an upheaval in our landscape along with active surveillance, environmental decontamination, and bundles of best practices, according to W. Kemper Alston, MD, MPH, of the Infectious Diseases Unit at Fletcher Allen Health Care in Burlington, Vt., who writes in a commentary in the September issue of Infection Control and Hospital Epidemiology.

 

As Alston notes, It was not too long ago that outcome data pertaining to hospital acquired infections (HAIs) were deeply protected secrets, rarely escaping a hospitals firewall and often not even shared within an institution. Incomplete pictures of what was happening in hospitals might be discretely pinned to a poster board once per year or shared confidentially with the National Nosocomial Infection Surveillance system as a way of contributing to a national benchmark, but in almost all cases, the data were strictly for internal use by infection control personnel to inform their own decisions about how to assign priorities.

Alston says that public reporting of infection data has changed that, and that the risks and benefits inherent in this approach are just beginning to be clarified. He says that two potential pitfalls of public reporting are inadequate risk stratification and validation. If a state lacks enough comparable healthcare institutions, he says an appropriate risk stratification cannot be achieved.

As to a lack of data validation, Alston comments, As HAI outcomes are released to the public, high rates will be bad for a hospitals business and reputation, independent of attempts to reduce reimbursement for care. In the past, infection control personnel were not discouraged from finding more infections, because the data were only for their own use. Now hospital administrators will bring significant pressure to bear on infection prevention programs, and rightfully so. High infection rates will provoke demands for explanations and action plans. A tremendous disincentive to report will arise, which could lead to measurement errors. This may not take the form of conscious underreporting but, rather, an unconscious, subtle laxity of surveillance or decisions not to report difficult to define events. Paradoxically, such a phenomenon could result in falling rates nationally and be cited as evidence of greater patient safety. As traditional paper medical records are replaced by enormously complex electronic records, it will be increasingly impractical for outside auditors unfamiliar with hospital systems to perform validation studies unless electronic data mining systems can be consistently applied across institutions.

Alston says that he hopes institutional transparency will drive reductions in HAIs, however, he adds, Meaningful comparisons demand standardized approaches to risk stratification, data validation, and surveillance methodology. Until those components catch up with the rush for disclosure, hospitals with excellent prevention programs will run the risk of being the worst.

Reference: Alston WK. Pitfalls of Public Reporting. Infect Control Hosp Epidemiol. 2010;31:985986