No pay, no gain: reducing hospital-acquired infections
Reducing hospital-acquired infections has become a huge part of broader infection control strategies the world over. Measures taken in the UK and Europe have had mixed results and, as Grace Lee of the Children’s Hospital in Boston explains to Practical Patient Care, the same can be said of the US ‘no pay’ policy.
It is four years since the US Centers for Medicare and Medicaid Services (CMS) introduced a new policy to improve infection control by ceasing additional payments for hospitalisations that result in complications deemed preventable, which includes hospital acquired conditions (HACs). The effectiveness of the so-called 'no-pay' policy has recently been reviewed in the American Journal of Infection Control, and the study has raised many questions about its effectiveness.
The reviewers set out to examine the perceived impact of the Medicare policy of adjusting payment to hospitals on the levels of hospital-acquired infections. It found that, while intentions were noble, the mechanism has not been as effective as some had hoped, as one of the review team, Grace Lee, associate medical director of Infection Control at the Children's Hospital Boston, explains.
"A perverse situation existed where hospitalisation for HACs got a higher payment, so the policy says that there will no longer be any extra payment," says Lee. "Secondly, the policy wanted to align value with payments to ensure that what is paid for is higher-quality care. This was the first example of the CMS using a stick rather than a carrot.
"It makes sense from a philosophical standpoint, but no extra resource is provided as a result of the policy. It has proven a much needed boost for some hospitals, but for others it has just shifted attention away from more important things."
In a December 2010 national survey of infection preventionists from a random sample of US hospitals, 81% of respondents reported that conditions targeted by the policy had been given increased attention - the much-needed boost to which Lee refers - but at the same time one-third reported spending less time on conditions that fell outside the policy. Furthermore, only 15% reported additional funding being provided to improve infection control.
The policy specifically targets ten categories of HACs, including infections linked to surgical procedures, central venous catheters and urinary catheters. Where infections not documented on admission are observed as a result of such procedures in Medicare patients during their hospital stay, the CMS denies additional payment to the hospital for the extra cost of treatment.
Lee notes that, in the study, respondents reported faster removal of urinary catheters (71%) and venous catheters (50%) as a result of the policy, but that this improvement was sometimes balanced by a shift of resources away from infections not targeted in the policy.
The data gathered by Lee and her colleagues suggests that, for one condition targeted by the policy - catheter-associated bloodstream infections - the rate of infection has not changed since 2008, and for catheter-associated urinary tract infections there has been a small increase in the rate of infection.
Overall, the study concludes that it is unclear whether the 'no pay' policy is resulting in better outcomes for patients.
"Some of the patients getting HACs are very sick at the baseline anyway, so the financial impact of the policy is reduced," observes Lee. "One of the first national evaluations found that there was a declining trend in HACs that began in 2006 and continued into 2011. The 'no pay' policy has not accelerated that rate of decline.
"A lot of effort is being made to control HACs by many state health departments, local agencies and bodies like the Society for Healthcare Epidemiology of America, as well as quality improvement organisations and now accrediting agencies, too. So, there are many contextual pressures that have led to a decline in HACs."
Concerns about the possible effectiveness of the 'no pay' policy were voiced from the very start, in part because a national policy does not take into account the specific operating conditions of individual hospitals and clinics. At the same time, however, unanticipated and unintended consequences are emerging from the policy that could set the tone for the future.
"We questioned at the start if a non-payment policy would have an impact," says Lee. "The penalties - whether perceived or actual - are too low and the metric focuses on billing information, which may not be the most appropriate measure. Our study focused only on HACs, but if we looked at other factors, like pressure ulcers, then maybe the policy would be effective. That should be looked at.
"With the National Healthcare Safety Network, which is part of the CDC (Centers for Disease Control and Prevention), we are seeing a shift away from the billing metric, so it will be interesting to see if there are incremental gains. What is needed is a lot of attention on how the policies are implemented."
The problem here is that there is poor correlation between the number of hospital-acquired infections identified through billing data and the number identified using standardised clinical definitions adopted by the CMS. Accordingly, the CMS asked hospitals to report central line-associated bloodstream infection data using National Healthcare Safety Network definitions on Hospital Compare - the online service that allows patients to compare the quality of care provided by Medicare-certified hospitals - from January 2011, with more to follow.
Another consequence of how the policy has been applied is that is appears to have a detrimental effect on hospitals serving particular socio-economic groups, especially those with a higher incidence of baseline illness. Nevertheless, Lee points out that, although the policy does not account for the circumstance of individual hospitals and has not materially improved the existing trend in the reduction of HACs, there have been positive outcomes.
"The policy could have a disproportionate effect on hospitals that treat a higher proportion of poorer patients," explains Lee. "Our study prompts further evaluation of how the policy affects different institutions. There is no definitive solution to reducing HACs.
"There are, however, some useful things about the policy. One showed up in a prior study we did, which showed that the 'no pay' policy does get the attention of hospital leadership. What is done at the bedside needs the support of hospital leadership, so that can only help."
The unintended consequences identified in the report Lee helped compile include resource shifting, particularly in large hospitals and in some cases where front-line staff were considered to be more receptive to changes in clinical processes. An increase in the level of unnecessary diagnostic testing on admission was also identified, notably at smaller hospitals.
The conclusion is that the organisational context is an important factor in determining how hospitals respond to policies applied nationally.
A platform for the future
In general, the first national, empirical evaluation of the 'no pay' policy is relatively positive, recognising that it is a novel approach to an issue that certainly deserves attention. Financial disincentives are increasingly common tools for improving the safety of healthcare, and should not be dismissed, though there are reasons to re-examine how they are applied.
At the very least, the policy has focused the attention of both hospital administrators and front-line staff on the issue of hospital-acquired infections. Yet the report also notes that it is still unclear whether the policy will result in fewer infections or, indeed, save the CMS any money.
The majority of hospitals reacted positively to the policy by increasing surveillance, education and implementation of infection prevention measures. Nevertheless, in some cases this was at the expense of attention to non-targetted infections and an increase in cultures performed on patients at admission without clinical indication. Perhaps the most key result of this review of the policy is that it is prompting clinicians and policymakers to ask what should be done next. For Lee and her co-authors, one conclusion is that more attention should be given to the design of non-payment policies to ensure that they lead to a direct and significant improvement in patient outcomes.
Their report openly admits its limitations - principally that it can only identify the perceived impact on hospitals rather than providing an empirical measure of its impact on patient outcomes, and that its large sample cannot tell the full story of all hospitals across the US - but its conclusions should be considered important in determining the next stage in refining the policy.
This is particularly important because the policy is being expanded to cover Medicaid through the Affordable Care Act and will impose greater financial penalties on hospitals that perform poorly against its measures. In such a situation, getting the measures and the penalties right is paramount.
"Infection control is not a money-maker," believes Lee. "Being tough on the reimbursement system is not necessarily rewarding these kinds of services. Each hospital has a different environment to work in. Some ideas in the policy may need to be adapted. If they are identified and evaluated then the next iteration of the policy could be better.
"Our report is not damning, though there are people who would take that view. I am more neutral and would like to see the report as the basis for adapting the policy."