In our recently published paper (Donker et al., PLoS ONE, 2019), we propose a surveillance system for antimicrobial resistance (AMR) and other hospital-associated pathogens, which is based on hospitals reporting the number of AMR positive cases they admit who have previously been discharged from other hospitals. By having hospitals report on each other’s AMR rates, this system promotes case finding in all hospitals. Let’s explain why we think this is necessary, and how the system works.
Reported incidences of AMR are increasingly used as measures of hospital performance. The idea behind this is that hospitals with above average AMR rates are driven to lower these rates, by increasing their infection prevention and control measures, or risk suffering reputational damage or financial penalties.
However, these comparisons rely on each hospital reporting their own AMR rates. Hospitals thus risk suffering reputational damage based on the very numbers they provided. Hospitals may therefore be driven to, for instance, use case definitions as strictly as possible, or lower their screening efforts. In other words, try to report as few cases as legitimately possible.
This ‘gaming’ of the reporting system, can make comparing AMR rates unreliable, because it depends on each individual hospital’s effort to find cases, which may vary. Even worse, it may result in punishment of hospitals that actually try to prevent cases by finding as many as they can.
The surveillance scheme we propose in our paper is less sensitive to this form of gaming, by separating the task of finding and reporting cases from the task of lowering the number of cases (through infection prevention and control).
This is done by suggesting each hospital screens patients who have recently been discharged from another hospital when they are admitted, and reports the number of AMR cases found in these patients. In this way, across the whole hospital network, we can measure AMR rates among patients discharged from each hospital.
Importantly, because the hospital doing the actual patient screening is not the same as the hospital in which we are measuring the AMR rate, a hospital can improve its ‘ranking’ relative to other hospitals in two ways. 1) Detect as many cases from the other hospitals as possible, or 2) improve infection control and prevention, thus lowering transmission and ultimately lowering the number of cases they discharge who are then admitted by other hospitals.
It may seem cynical to have hospitals report on each other. However, if the system currently in place, based on self-reporting, is unable to provide a reliable comparison between hospitals, some of these hospitals will be punished for the wrong reasons. Our proposed surveillance system can also be used in combination with the existing system; it uses a smaller population size but promotes case finding. It should also eventually improve patient safety, as hospitals are driven to find as many cases as possible, as well as reduce transmission.
You can read more about this work here: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0219994
Funding: The research was funded by the National Institute for Health Research Health Protection Research Unit (NIHR HPRU) in Healthcare Associated Infections and Antimicrobial Resistance at University of Oxford in partnership with PHE [grant number HPRU-2012-10041].
