Does Nonpayment Influence Infection Rates?


Hello. This is Paul Auwaerter, with Medscape Infectious Diseases, speaking from the Division of Infectious Diseases at Johns Hopkins University School of Medicine. I would like to talk about some interesting recent studies about so-called "never" events. Back in October 2008, the Centers for Medicare & Medicaid Services (CMS) dictated a new approach towards trying to prevent hospital-acquired infections in certain scenarios, with a focus on ventilator-associated infections, urinary catheter-associated infections, complicated urinary tract infections, and vascular catheter-associated infections. Some of us grumbled that it is an unrealistic goal to have "never" events, but many of us applauded an effort to see whether such a heavy-handed approach might reduce the rates of these acquired infections.

Some of the first higher-quality data have now come out from these approaches, and I thought I would comment on 3 studies. The first was published in the spring of 2012 by Lee and colleagues. This was a survey of infection-control practitioners, 81% of whom believed that the CMS policies increased the attention to these catheter-related infections and perhaps they could talk to their hospital administrators about it. This might be similar to a survey of children in a candy store asking if they want more candy. However, it is of some interest that only 15% of those surveyed said that they received any additional resources from the facility's administration for achieving these so-called "pay for performance" goals, which in this case is really just loss aversion (meaning nonpayment). You would not get payment for these particular infections.

This past week in the New England Journal of Medicine, the same group surveyed 398 hospitals to determine whether infection rates had declined compared with historical data from 2006 until implementation in 2008. The upshot was that there was no change in rates of bloodstream infection, urinary tract infection, or ventilator-associated pneumonia. The reason for this finding could be that the investigators were just looking at billing data, and many of these diagnoses may not have been coded. It could also be low motivation as a result of the loss aversion or nonpayment for these infections, or perhaps sufficient practice improvement had already taken place. Although these are laudable goals, in busy medical practices, especially on inpatient services, even the nonpayment strategy is not enough to increase efforts to prevent these infections.

As happened with community-associated pneumonia and the 4-hour rule to administer antibiotics, many of us were concerned about unintended consequences of these nonpayment policies. Morgan and colleagues in Clinical Infectious Diseases examined whether these policies might lead to overtreatment of urinary tract infections or other consequences. Their hypothesis was that clinicians would screen for asymptomatic bacteruria on admission and there would be increased antibiotic use with the initial admission to perhaps characterize the urinary tract infection as a pre-existing condition. They did not find this but in fact found a degree of undertreatment.

The skeptic in me would say that what is really happening is that, based on either billing data or other factors, complicated urinary tract infections are not being described and diagnosed, and this may also be happening with ventilator-associated pneumonia because that diagnosis is already difficult. There may be a tendency to diagnose other problems. Although not the focus of the study, these may be some of the consequences.

So, although these policies remain in force, it is worth stepping back to see whether it is worth continuing these efforts, or perhaps we have achieved rates as low as we can, although many people do not agree with that. None of us believe that these "never" events will never occur, and a "never" event should be something more akin to supervising compounding pharmacies to make certain that they are not mass-producing contaminated medications.

Be that as it may, these are all-important issues that speak to trying to change and prompt behavior to achieve worthy goals. The jury is still out; these data did not appear to be enough, but we will see whether this evidence leads to a reflection of these policies or even further changes to help achieve these goals. Thanks so much for listening.