This article and accompanying editorial from the Annals of Internal Medicine addresses several of our themes:
1. pay for performance
2. disparities
3. health services research methods
A healthy meal of food for thought.
Jha AK, Orav EJ, Epstein AM. The Effect of Financial Incentives on Hospitals That Serve Poor Patients. Annals of Internal Medicine. 2010 September 7, 2010;153(5):299-306.
The Effect of Financial Incentives on Hospitals That Serve Poor Patients — Ann Intern Med
Werner RM. Does Pay-for-Performance Steal From the Poor and Give to the Rich? Annals of Internal Medicine. 2010 September 7, 2010;153(5):340-1.
http://www.annals.org.ezpminer.urmc.rochester.edu/content/153/5/342.extract
jd
This article certainly suggests that there is some promise in the use of a pay for performance reimbursement scheme. From a policy making stand point, I’m sure this article was well received. I wonder, however, how the doctors and other health care providers who work in the hospitals that opted to participate in the P4P program view this payment approach. We have already seen several examples on this blog of physicians bemoaning the attempts of policy makers to improve quality and reduce costs. Does this program infringe on their autonomy? Do they feel they are able to adequately treat their patients? I suspect that any financial and quality improvement benefits gained from this plan could quickly become overshadowed if health care clinicians were to vocally oppose its widespread adoption.
ReplyDeleteI suppose clinicians’ perceptions of P4P would be another interesting research question to pursue, but I think this issue also points to a broader challenge for us as future health service researchers: how does one bridge the gap between research and “real-world” implementation? As someone who is not too far removed from being a practicing physical therapist, I would say there is at times a lack of collaboration and appreciation between the research community and the provider arm of our health care system. This was most apparent to me as I started to inform my former colleagues that I was going to be entering this program. I was met with more than a few eye-rolls and incredulous looks, suggesting that they viewed health services research as something totally separate from the actual delivery of care rather than a crucial supporting component. Likewise, this blog has provided several pertinent examples of a policy which seems good on paper having unintended negative consequences for providers, further perpetuating the conflict between these groups.
I’m sure there is no easy answer to this challenge. In fact, I think this issue is one of the major reasons the university is building an entirely new building outside of my window right now. But at the very least, this barrier is certainly something worth keeping in mind right from the start of our HSR careers.
If I were in the position to implement a P4P quality improvement incentive scheme for a region, this article would be very useful as part of the policy decision making process. This leads me to believe that there is some level at which quality can only improve in very small increments and at that high level (think Mayo Clinic) financial rewards may have a smaller effect than for these DSHs. Based on this article, I would target DSHs for financial rewards and would try to maintain quality in high performing hospitals using strategies that target organizational culture and prestige. From the findings they cited about publicly reported performance data, it seems that carrots more than "hidden" sticks works better, at least at the level of low performing hospitals with a mostly poor patient profile.
ReplyDeleteThe other point that would interest me while structuring the P4P program, is that condition matters. I think it might be helpful to compare processes and alert systems in high quality hospitals to lower quality ones and see what key performance indicators are critical to success. And, from there, look into what these indicators entail --- is it more money, more/greater knowledge workers, or a different system for operations? Then see if targeting the aspects that funnel into the indicators help improve quality. A DSH cannot improve if they do not have the basic tools necessary for improving quality in that condition.
But at the same time, these condition specific quality improvement efforts do seem siloed and unable to take advantage of organization wide processes and resources. If a technology or a process or a set of knowledge workers are crucial for the improvement efforts in one condition, it may be worth looking if that resource can be utilized across the organization for all or many conditions. So, while knowing what is necessary for each condition will go a long way toward improving the quality of that one condition; that information should be pooled to make the best use of resources in addressing as many conditions as possible. Or it may simply be that the DSHs get a different patient profile and either they do not perform enough of certain services and therefore have not built (and will never build) the repertoire of skill to excel in that area.
Finally, I thought it was interesting that no for-profit hospital participated in the program. Isn't the profit-motive/capitalism supposed to inspire/attract the most ingenious among us, those with a pro-active, motivated, entrepreneurial spirit?
This study provides promising results for pay for implementing performance measures to increase quality for some conditions in hospitals that serve more poor patients. And in fact, hospitals serving poor patients showed greater quality improvements than hospitals with less poor patients. I thought it was interesting in light of the 9/1 blog post titled “Why pay for performance does not work and may impair patient care.” Aside from the logical conclusions regarding motivation theory that were drawn from the study in question, perhaps the economic status of patients played a role in the adverse effects of the pay for performance measures. It would definitely be useful to extend the work of both of these studies to determine some reasoning for the discrepancies in findings.
ReplyDeleteViji, in response to your great observation that no for-profit hospitals participated in this study, I think the underlying factor behind this result is that there was not only the potential for a reimbursement bonus but also a penalty for poor performance. It seems clear that for-profit hospitals made the calculation that the potential rewards of innovation and improvement of our payment system, in addition to the possibility of a 1-2% financial bonus in every year of the study, were not worth the risk of being penalized with a 1-2% payment in only the last year of the study.
ReplyDeleteApparently the warm, fuzzy feelings you get from being on the cutting edge of reform and quality improvement are not worth as much as I thought they would be!
As I read through the article, I was impressed with the meticulous structure, the study methodology, the quantitative findings, the graphs etc. However, as the conclusion drew closer, a sense of incompleteness settled in. The ‘Discussion’ section quite elaborately discussed the limitations of the study, but a detailed qualitative analysis of the study findings was conspicuously missing.
ReplyDeleteIt is heartening to read that the study findings re-establish that a P4P approach does work. I’m not sure if it would be beyond the scope of the study, but a qualitative analysis of the following would add value to the study findings:
- The baseline gaps that existed in the delivery of quality care in the P4P hospitals
- How the financial incentives were used to address these gaps
- Why did the incentives work only for AMI & Pneumonia
- Why did the incentives not work for CHF
- How was the financial health of the hospital before the study commenced?
As a greenhorn in the field of research, I do not know if a study would produce convincing findings given the significant limitations already known. Like we learn in the stats class, the results flow from what we model, from the sample we select. Would it then be worthwhile to pursue a study in the face of so many limitations? In this case, the result though endearing to me leaves many questions unanswered.
Before posting my views on the blog, I browsed through the reader comments section on the net (with the hope of not being the only one who felt this way about the study), and to my relief the solitary comment that figured, panned the study as well. The comment from Ishak Mansi, provides a whole new dimension to the shortcomings of the study, and is worth a glance. Hope to read some more comments there in the days to come!