A new study from the Dartmouth Atlas Project challenges the effectiveness of Medicare's risk-adjustment efforts, or the formulas commonly used to assess how sick patients are.
Medicare uses risk adjustment in its payment methodology to base reimbursement on the underlying health status of a hospital's patients in an effort to protect hospitals with the sickest patients from losing money.
For this study, published in the journal BMJ, Dartmouth researchers analyzed Medicare claims for services provided in 2007 among 306 hospital referral regions. They analyzed three different formulas that are commonly used to assess how sick patients are. Each formula is based on the number and nature of diagnoses as well patient age, race and sex. Researchers also analyzed the death rates of patient populations in each of the 306 regions.
They found the mean number of physician visits per patient during the last six months of life varied from 10 to 59 and was not correlated with age, sex and race-adjusted mortality. Rather, the researchers found the rate of visits was strongly correlated with the number of diagnoses observed in the claims data.
The authors concluded that "the more one looks, the more one finds." This means that the sicker regions do not necessarily have higher patient visit rates. It also suggests that using diagnoses data to adjust for risk produces problems such as (a) bias in research and evaluation, (b) biased performance measures and (c) biased payment to third-party payors, according to the study.
Bias in research and evaluation. If spending or utilization per capita and rates of diagnosis are highly correlated, studies seeking to evaluate the relationship between patient visits and mortality while controlling for illnesses will be skewed.
Biased performance measures. "Adjusting performance measures using several different diagnoses makes providers who frequently make diagnoses look better than those who manage their patients more conservatively," the study authors concluded.
Biased payment to third-party payors. Payments to third-party payors that are adjusted based on the frequency of diagnoses recorded in a administrative database could result in higher per-capita payments in regions that have more physicians, hospital beds and visits per capita — regardless of the underlying disease burden among patients.
What We Do and Don't Know About CMS' Bundled Payments Program
CMS Releases Proposed 2014 Payment Plan for Medicare Part D, Advantage
Medicare uses risk adjustment in its payment methodology to base reimbursement on the underlying health status of a hospital's patients in an effort to protect hospitals with the sickest patients from losing money.
For this study, published in the journal BMJ, Dartmouth researchers analyzed Medicare claims for services provided in 2007 among 306 hospital referral regions. They analyzed three different formulas that are commonly used to assess how sick patients are. Each formula is based on the number and nature of diagnoses as well patient age, race and sex. Researchers also analyzed the death rates of patient populations in each of the 306 regions.
They found the mean number of physician visits per patient during the last six months of life varied from 10 to 59 and was not correlated with age, sex and race-adjusted mortality. Rather, the researchers found the rate of visits was strongly correlated with the number of diagnoses observed in the claims data.
The authors concluded that "the more one looks, the more one finds." This means that the sicker regions do not necessarily have higher patient visit rates. It also suggests that using diagnoses data to adjust for risk produces problems such as (a) bias in research and evaluation, (b) biased performance measures and (c) biased payment to third-party payors, according to the study.
Bias in research and evaluation. If spending or utilization per capita and rates of diagnosis are highly correlated, studies seeking to evaluate the relationship between patient visits and mortality while controlling for illnesses will be skewed.
Biased performance measures. "Adjusting performance measures using several different diagnoses makes providers who frequently make diagnoses look better than those who manage their patients more conservatively," the study authors concluded.
Biased payment to third-party payors. Payments to third-party payors that are adjusted based on the frequency of diagnoses recorded in a administrative database could result in higher per-capita payments in regions that have more physicians, hospital beds and visits per capita — regardless of the underlying disease burden among patients.
More Articles on Medicare and Studies:
Medicare Advantage Payment Cuts Cause Insurers' Stocks to TumbleWhat We Do and Don't Know About CMS' Bundled Payments Program
CMS Releases Proposed 2014 Payment Plan for Medicare Part D, Advantage