By: Sandeep Wadhwa
The talk about Accountable Care is turning into action with the announcement of the Pioneer ACO awardees from CMS. In addition, the Medicare Shared Savings Program solicitation has been released with initial responses now being accepted. As delivery systems and providers consider whether they want to respond, I’d like to suggest some key areas to consider.
- Quantify the potentially preventable events for the population under consideration. Preventable events represent cost and quality opportunities in the delivery system. Reducing preventable events can result in real savings and improved patient care. There are five major categories of potentially preventable events: admissions, readmissions, complications, emergency room visits, and ancillary services. If your system has low preventable event rates, there may be limited opportunity for additional savings. Furthermore, matching up the areas of opportunities with your ability to institute change in those areas will help assess the probability of success. If cardiovascular related complications and readmissions are the dominant source of high rates of preventable events and the clinical leadership and program is in disarray; better to know that upfront.
- Rigorously risk adjust. The allocation of new resources and services should, in part, be driven by the risk profile of the population. Setting additional medical home/care management fees, for example, for primary care should be a function of the risk profile of the primary care provider’s panel. In addition, the rates of preventable events differ by the chronic disease burden and severity of illness of the population served. Understanding the mix of healthy, moderately ill, and very ill patients is another dimension in assessing the opportunity for savings and will help focus where interventions should be channeled.
- Think through savings attribution up front. If emergency room visits decline who receives the credit? Primary care or the hospital? How does the responsible entity attribute savings among multiple provider types? We’d suggest that the comparison of actual and risk-adjusted expected rates of potentially preventable events can be used to quantify savings and allocate those savings among individual providers. Any method of sharing savings must balance rewarding historical good performance with rewarding subsequent improved performance.
Done correctly, potentially preventable events and risk adjustment can be an integral part of an effective shared savings program.
All very true – though the devil is in the details.
Clinical risk-adjustment is wonderful for managing illness; however does one use an ICD-9/ICD-10 database or a clinically abstracted database to categorize illness? I wish I could say that ICD-10 databases could be used, given their ease of access; however due to limitations in ICD-10 (e.g. there’s only one code for native coronary artery disease, not multiple codes for left main, one, two, or three vessel disease), the differences in coding-language and physician-language (e.g. “urosepsis” is not “sepsis due to UTI”; “heart failure with preserved systolic function” is not “diastolic heart failure”) the reality that physicians do not document their conditions using ICD-10 lanugage, and the inability of coding professionals to clinically interpret the record to “translate” the physicians intention (e.g. coding from radiology reports or abnormal laboratory values), these databases do not necessarily reflect the patient’s medical conditions. Perhaps, at some point, we can overlay lab values to the risk-adjustment, enhancing the reliability of these databases.
Furthermore, our health system (and society) does not reward wellness. Other than professional satisfaction, what benefit does a primary care physician receive from an time-consuming intervention that promotes sobriety in an alcoholic, saving the healthcare system thousands of dollars in preventable illness or untold pain and suffering (and money) from an avoidable traffic accident? How do we eliminate teenage pregnancy, a sure-fire poverty generator?
I applaud all the work that addresses these issues.
I agree we’re starting down the road to ACO’s with this announcement, but I’m not sure how well the regulations meet the suggestions that you are making. The suggestions you make are reasonable and interesting, but they seem to ignore the idea that what people are really going to do with ACO legislation is find the simplest way to extract the most amount of money out of the regulation. There will be some exceptions, but this is how it works with most government programs.
Pingback: Accountable Care Organizations Becoming Action Thanks to Pioneer ACO Awardees | EMR and EHR
Excellent points and guidance above. The front end of the quality assurance movement albeit aided by the liability carriers coined the term: ‘potentially compensable events’, which led to the development and refinement of the I/SD intensity of illness/severity of disease ‘generic screening criteria’ set for auditing and CME purposes. One more time. we’re there again with the ‘preventable events’ noted above.
Unlike the previous commenter, I see no conflict nor constraint in the rule that prohibits any and all of the above being purposely deploy to risk adjust and assist in the management of an ACO population once attributed or otherwise identified.
This is true at the Federal level whether via the MSSP program, etc., as well as the many creative iterations we are likely to see in the commercial market as health plans team up with ACOs to manage their defined populations.
Nice work!
Pingback: New Metrics for Health Care Reform | 3M Health Information Systems