Intelligence Unit Special Reports Special Events Subscribe Sponsored Departments Follow Us

Twitter Facebook LinkedIn RSS

Risk-adjusted model underpays, ignores functional status

Medicare study

Risk-adjusted model underpays, ignores functional status

Medicare’s capitation model does not take into account a beneficiary’s functional status and underpays for multiple comorbidities. This lowers reimbursements paid to physicians and health insurers involved in Medicare managed care plans who treat frail patients with multiple chronic conditions, according to a study that appeared in the October American Journal of Managed Care.

“Medicare Capitation Model, Functional Status, and Multiple Comorbidities: Model Accuracy” found that the Centers for Medicare & Medicaid Services Hierarchical Condition Categories (CMS-HCC) risk-adjusted model for Medicare payments underpredicted payments for patients with hypertension, lung disease, chronic heart failure (CHF), and dementia. (See Figure 8 on p. 8 and Figure 9 on p. 9 of the PDF.)

“The difference between the actual costs and predicted payments was partially explained by beneficiary functional status and less-than-optimal adjustment for these chronic conditions,” the researchers state.

CMS-HCC is a risk-adjustment model that “relies on demographic and diagnostic information available from administrative data to predict resource use,” according to the study. Rong Yi, PhD, senior scientist and principal of analytic services at Verisk HealthCare, Inc., in Boston, says CMS tried to discourage vague coding and gaming of the system when it created HCC, which uses a subset of ICD-9-CM codes that focuses on chronic conditions or acute complications of chronic conditions.

“While this is a very good policy decision, we have found that nonchronic conditions often have cost implications for future years,” says Yi.

Although HCC includes lung disease/cancer, stroke/arthritis, and diabetes/coronary artery disease (CAD), there is still a discrepancy between the actual and predicted cost ratios, according to the study. (See Figure 10 on p. 10 of the PDF.)

The underpredicting suggested in the study could play a large role in reimbursements to physicians and insurers because two-thirds of noninstitutionalized Medicare be-neficiaries over the age of 65 have two or more chronic conditions, according to the study.

Katia Noyes, PhD, MPH, associate professor in the department of community and preventive medicine at the University of Rochester (NY) and the lead author of the study, says plans and practices with a large proportion of frail elderly could lose money because of the model.

The HCC model does not take into account functional impairment. This could affect Medicare’s Special Needs Plans (SNP), which provide coordinated care to dual eligibles with certain chronic illnesses who reside in institutions such as nursing homes.

Yi says non–claims based information such as functional status, socioeconomics, culture, linguistics, and geography affect how patients interact with the healthcare system. “Including non–claims based data has been discussed for quite a number of years in the risk-adjustment and predictive modeling field, functional status being one of them. Clinically, functional status significantly affects how a patient seeks care and follows the doctor’s treatment requirements,” she says.

Yi says collecting these kinds of data elements will take a large effort. “Unless there is a systemwide effort to start enforcing the collection of such data elements, we can debate the underpayment relating to not having functional status or other factors like this forever. I personally don’t think one can do much about it at all,” she says.

The model also doesn’t take into account dementia, osteoporosis, and other chronic conditions that could affect a patient’s health status and care.

Researchers looked at how 11 target comorbidities may affect functional impairment, including arthritis, osteoporosis, diabetes, CAD, CHF, and dementia.

Certain combinations of those ailments could affect patient performance in activities of daily living (ADL), which would affect medical costs. ADLs include bathing, dressing, and eating. The researchers say the more ADL deficiencies a patient has, the greater the difference between the HCC model’s predicted cost and the actual expenses.

In the study, researchers looked at 46,790 communitydwelling Medicare beneficiaries between 1992 and 2000. Nearly three-quarters of beneficiaries in the study had two or more target comorbidities.

“Patients with CHF and dementia reported the highest level of deficiency across all ADL categories: 14.38% relied on others’ help with eating (feeding), and more than 50% used help or assisted devices for bathing. Other groups with a high ADL deficiency level included patients with stroke combined with hypertension or arthritis, CHF and osteoporosis, and CAD and diabetes,” according to the study.

The findings mean that the CMS-HCC model “significantly underpredicts expenses for patients with hypertension, lung disease, CHF, and dementia after adjusting for patents’ disability level,” wrote the researchers. (See Figure 11 on p. 11 of the PDF.)

Ross Winkelman, managing director at Wakely Consulting Group in Denver, who helps managed care organizations develop Medicare Advantage filings and bids, says his company has seen a similar understatement of the HCC risk-adjustment models for subgroups. Another issue is the limited number of comorbidities recognized in the HCC model and delays that increase risk scores. Managed care organizations or provider groups that accept a percentage of risk-adjusted benchmarks and serve a disproportionate number of sicker, more frail individuals will likely be underpaid, and those with healthy populations will be overpaid.

“Risk adjustment dampens these effects but does not completely eliminate them,” says Winkelman.

Kirk L. Shanks, MAS, actuarial analyst at Wakely Consulting Group in Clearwater, FL, points to another concern: The Medicare HCC model is based on the previous year’s claims data. For example, if an individual is healthy in 2007 and becomes ill in 2008, the provider group or managed care organization would receive lower payments because of the patient’s low-risk score based on 2007 data.

If the patient leaves the plan in 2008, the group or organization would not receive increased payments. However, if the patient is in the same plan in 2009, and the HCC increases the individual’s risk score, the group or organization would recoup the previous year’s costs with 2009’s higher payments.

Shanks says another example is a new 65-year-old Medicare beneficiary who is ill but has no encounter data with CMS. “The plan would be paid a default rate, which is relatively low, in the first year. The plan would not get the benefit of the increased risk score and associated plan payment until the next year,” he says.

Effect on health plans

These discrepancies affect not only physicians, but also health plans involved in Medicare managed care programs.

“Our results demonstrate that unless a special disability adjustment is introduced for patients with comorbidities, entering into risk arrangements with Medicare for services provided to people with multiple comorbid conditions may be more risky for health plans serving this population than anticipated,” wrote the researchers.

SNPs that are not qualified for the frailty adjustment are “financially at a disadvantage in providing care to the very frail and disabled,” the researchers wrote. These kinds of disincentives could push managed care plans to not enroll those individuals.

“[Private insurers] probably should start looking at collecting more information on functional status just as Medicare should,” says Noyes, adding that the payment model’s deficiencies could be a catalyst for primary care doctors leaving for specialty care.

In specialty care, doctors are responsible for only one condition rather than the multiple comorbidities in primary care.

The payment model could potentially benefit specialty care, which is more costly than primary care. “I think that’s one reason why there is an outflow of providers from primary care,” says Noyes. “[Primary care is] where most of the elderly get their care.”

Noyes says she is hopeful that Medicare will review its payment system and adjust for comorbidities and functional status. However, she does not expect changes, because tweaking the system is unlikely to save money, although it could improve quality and outcomes. “My experience tells me that very few things in this current system of healthcare saves money. You may improve quality of care, but everything comes at a cost,” she says.