Bundled payment policies may not lead to ‘cherry-picking’ patients
Click Here to Manage Email Alerts
Researchers did not find an association between initiation of CMS mandatory bundled payment policies and decreased access to hip and knee replacements for Medicare patients, according to published results.
To evaluate changes in patient case mix from pre-policy to post-policy, Casey Jo Humbyrd, MD, and colleagues compared Medicare beneficiaries who received hip and knee replacement in bundled metropolitan statistical areas with patients who received hip and knee replacement in non-bundled metropolitan statistical areas from 2015 to 2016. Researchers collected information on race, dual eligibility for Medicare and Medicaid, tobacco use, obesity, presence of diabetes with or without complications, and Charlson comorbidity index (CCI) values among the groups.
To determine whether there were changes in access to hip and knee replacement, researchers also compared Medicare beneficiaries in bundled metropolitan statistical areas who underwent hip and knee replacements to those who underwent hip hemiarthroplasty to evaluate pre-policy to post-policy changes in patients case mix.
Pre-policy, post-policy comparisons
Results showed both unmatched and matched samples had significant differences in the unadjusted baseline characteristics between the bundled metropolitan statistical area and non-bundled metropolitan statistical area cohorts. Prior to matching, researchers found white race, comorbidities and cases of diabetes with complications were more common pre-policy among patients in bundled metropolitan statistical areas compared with patients in non-bundled metropolitan statistical areas. Post-policy, results showed younger age, white race and cases with more comorbidities were more common in patients who underwent hip and knee replacement in bundled metropolitan statistical areas.
In the overall group of patients in bundled metropolitan statistical areas, patients who underwent hip and knee replacement pre-policy were less likely to use tobacco and were more likely to have diabetes without complications compared with post-policy, according to results.
After matching, researchers noted patients in bundled metropolitan statistical areas were less likely to be white and were less likely to be obese but more were likely to have a history of tobacco use pre-policy compared with patients in non-bundled metropolitan statistical areas. However, results showed no significant post-policy changes in the characteristics of patients undergoing hip and knee replacement.
Hip, knee replacement vs hemiarthroplasty
Compared with patients who underwent hip and knee replacement in bundled metropolitan statistical areas post-policy, researchers found significantly higher CCI values among patients who underwent hemiarthroplasty. However, researchers noted the difference was small. Results showed diabetes mellitus without complications was 2.4% more likely among patients who underwent hemiarthroplasty vs. hip and knee replacement post-policy.
The authors noted it may be necessary to examine the concerns of bundled payments limiting access to patients at the provider level instead of the metropolitan statistical area level “if variations by provider are ‘averaged out’ when examining the geographic area unit” in longer-term studies.
“Although we did not find behavior changes in aggregate, it is possible that behavior changes are present in a small subset of providers that we have not identified,” the authors wrote. “Subgroup analysis should also measure access for higher-risk individuals (for example, patients with end-stage renal disease) and follow the patterns in access to [hip and knee replacement] HKR since the implementation of the Affordable Care Act and the introduction of new payment models.” – by Casey Tingle
Disclosures: The authors report no relevant financial disclosures.
Perspective
Back to TopJames D. Slover, MD
Casey Jo Humbyrd, MD, and colleagues examined the impact of mandatory participation in a CMS bundled payment program known as the Comprehensive Care for Joint Replacement in their manuscript entitled, “patient selection after mandatory bundled payments for hip and knee replacement.” Previous CMS bundled payment programs have achieved the goal of improved quality and increased cost savings for total joint replacement services through incentivizing and rewarding increased efficiency, care coordination and cost savings efforts by providers. One concern has been that these programs could incentivize so called “lemon dropping” or “cherry picking,” whereby providers would be incentivized to avoid providing total joint replacement to higher risk patients that may have increased risks of complications or require more resources throughout the bundled episode, or be more inclined to offer total joint replacement to low risk patients that are likely to require few resources for their care. “Lemon dropping” can threaten access to these services to patients that could substantially benefit from them, and “cherry picking” can undermine the program by incentivizing increased surgery rates in some low-risk populations.
The authors found no difference in the patient characteristics in the year before and the first year after implementation of Comprehensive Care for Joint Replacement (CJR) in this study. This is encouraging, but not definitive, evidence that the program has not incentivized these practices. The parameters examined in the study including obesity, diabetes mellitus, tobacco use and the Charlson Comorbidity Index provide some general description of patient populations, but do not provide data about individual patients or providers and how the surgical indications may have changed for them. In addition, other factors such as changes in coding practices, whereby there is increased capture of patient comorbidities have occurred at many institutions, may have impacted results. Furthermore, as providers see and understand the results of the CJR program after participation, behavior could change going forward. Lastly, the structure of CJR causes reimbursement to go down in future years through reduction in target process as the whole system becomes more efficient, and this may impact future behavior. Overall, this first look is promising, but continued further monitoring and evaluation of the behavior and experience of individual patients and providers as CJR moves forward is needed.
NYU Langone Health
New York
Read more about
Click Here to Manage Email Alerts