Clinicians seem to be overusing "low-value" medical services that provide little or no benefit to Medicare beneficiaries, according to a report published online May 12 in JAMA Internal Medicine.
In the study, Harvard researchers developed 26 claims-based measures drawn from evidence-based lists of services providing minimal clinical benefit. Categories included low-value cancer screening, diagnostic and preventive testing, imaging, and surgical procedures. In all, the study tracked more than 1.3 million Medicare beneficiaries’ claims from 2009, analyzing the proportion of them receiving such services, mean-per-beneficiary service use, and the proportion of total spending devoted to the services.
Among the study’s 26 designated measures of low-value services: cervical cancer screening for women 65 years and older, computed tomography scanning of the sinuses for uncomplicated acute rhinosinusitis, head imaging for uncomplicated headache, preoperative cardiac stress testing, and back imaging for patients with nonspecific low back pain. Researchers tweaked measures by age, symptoms, and site of care, creating different levels of sensitivity for each measure (i.e., a more sensitive, less specific baseline definition, along with a less sensitive, more specific definition with additional restrictions).
Nationwide, between 25% and 42% of beneficiaries received low-value services, accounting for as much as $8.4 billion, or 2.7% of annual spending for services covered by Medicare Parts A and B, the researchers estimated (JAMA Intern. Med. 2014 [doi: 10.1001/jamainternmed.2014.1541]).
"Now that we have this measurement tool, we can use it to try to better understand when overuse is happening and what can be done to reduce it," Aaron Schwartz, the study’s lead author, who is in the MD-PhD program at Harvard Medical School, Boston, said in an interview.
Researchers sought to "cast a broad net" in creating a low-value medical services index that could be tracked over time and used to evaluate various reform efforts, Mr. Schwartz said. "For the vast majority of services, whether it’s ‘low value’ or ‘high value’ really depends on the patient and the clinical setting," he said. Analyzing the use of different services within various geographic areas, researchers found that even geographic areas spending less on low-value Medicare services "still spent a lot on them," he said.
Payment reforms, such as global – or bundled – payment models, could afford greater discretion to clinicians in identifying low-value services and in finding incentives to eliminate them, according to Mr. Schwartz. "Our team generally favors proposals that retain provider discretion at point of care," he said.
The study’s results are "consistent with extensive overuse in the system ... even with just 26 measures," Mr. Schwartz noted. But many claims-based measures of overuse may not be accurate enough to support targeted coverage or payment policies without resulting in unintended consequences, he cautioned.
Arkansas Medicaid Director Dr. William Golden said the Harvard study’s findings are not surprising, with physicians’ overuse of "low-value" services extending beyond Medicare. But new payment strategies and financial incentives, such as those employed by his state Medicaid program, seem to be helping to change such practice patterns and manage total cost of care, he said.
"Low-value activities are embedded throughout the routine orders of health care," Dr. Golden said in an interview. "Since Arkansas Medicaid has made health providers accountable for total cost of care and included shared savings and cost sharing, the health care community has invested the time and energy to retrain clinical reflexive behavior. Delineation of effective strategies coupled with financial incentives has promise to be a promising strategy in our state."
Mr. Schwartz reported having no conflicts of interest.