Almost 100 hospitals reported suspicious data on dangerous infections to Centers for Medicare & Medicaid Services officials, but the agency did not follow up or examine any of the cases in-depth, according to a report by the Health and Human Services inspector general's office.
Most hospitals report how many infections strike patients during treatment, meaning the infections are likely contracted inside the facility. Each year, Medicare is supposed to review up to 200 cases in which hospitals report suspicious infection-tracking results.
The IG said Medicare should have done an in-depth review of 96 hospitals that submitted "aberrant data patterns" in 2013 and 2014. Such patterns could include a rapid change in results, improbably low infection rates or assertions that infections nearly always struck before patients arrived at the hospital.
The IG's report, released Thursday, was designed to address concerns over whether hospitals are "gaming" a system in which it falls to the hospitals to report patient-infection rates and, in turn, the facilities can see a bonus or a penalty worth millions of dollars.
The bonuses and penalties are part of Medicare's Hospital Inpatient Quality Reporting program, which is meant to reward hospitals for low infection rates and give consumers access to the information at the agency's Hospital Compare website.
The report zeroes in on a persistent concern about deadly infections that patients develop as a result of being in the hospital. A recent report in the journal BMJ identified medical errors as the third-leading cause of death in the country. Hospital infections particularly threaten senior citizens with weakened immune systems.
Rigorous review of hospital-reported data is important to protect patients, says Lisa McGiffert, director of the Consumers Union's Safe Patient Project.
"There's a certain amount of blind faith that the hospitals are going to tell the truth," McGiffert says. "It's a bit much to expect that if they have a bad record, they're going to fess up to it."
Yet there are no uniform standards for reviewing the data that hospitals report to Medicare, says Dr. Peter Pronovost, senior vice president for patient safety and quality at Johns Hopkins Medicine.
"There are greater requirements for what a company says about a washing machine's performance than there is for a hospital on quality of care, and this needs to change," Pronovost said. "We require auditing of financial data, but we don't require auditing of [health care] quality data, and what that implies is that dollars are more important than deaths."
In 2015, Medicare and the Centers for Disease Control and Prevention issued a joint statement cautioning against efforts to manipulate the infection data. The report said CDC officials heard "anecdotal" reports of hospitals declining to test apparently infected patients so there would be no infection to report. They also warned against overtesting, which helps hospitals assert that patients came into the hospital with a pre-existing infection, thus avoiding a penalty.
In double-checking hospital-reported data from 2013 and 2014, Medicare reviewed the results from 400 randomly selected hospitals, about 10 percent of the nation's more than 4,000 hospitals. Officials also examined the data from 49 "targeted" hospitals that had previously underreported infections or had a low score on a prior year's review.
All told, only six hospitals failed the review, which included a look at patients' medical records and tissue sample analyses. Those hospitals were subject to a 0.6 percent reduction in their Medicare payments. Medicare did not specify which six hospitals failed the data review, but it did identify dozens of hospitals that received a pay reduction based on their reports on the quality of care.
The new IG report recommends that Medicare "make better use of analytics to ensure the integrity of hospital-reported quality data."
A response letter from Centers for Medicare & Medicaid Services Administrator Seema Verma says Medicare concurs with the finding and will "continue to evaluate the use of better analytics ... as feasible, based on [Medicare's] operational capabilities."
Questions about truth in reporting hospital infections have percolated for years, as reports have trickled out from states that double-check data.
In Colorado, one-third of the central-line infections that state reviewers found in 2012 were not reported to the state by hospitals, as required. Central lines are inserted into a patient's vein to deliver nutrients, fluids or medicine. Two years later, though, reviewers found that only 2 percent of central-line infections were not reported.
In Connecticut, a 2010 analysis of three months of cases found that hospitals reported about half, or 23 of the 48 central-line infections that made patients sick. Reviewers took a second look in 2012 and found improved reporting, with about a quarter of the cases unreported, according to the state public health department.
New York state officials have a rigorous data-checking system that they described in a report on 2015 infection rates. In 2014, they targeted hospitals that were reporting low rates of infections and urged self-audits that found underreporting rates of nearly 11 percent.
Not all states double-check the data, though, which Pronovost says underscores the problem with data tracking the quality of health care. He said common oversight standards, like the accounting standards that apply to publicly traded corporations, would make sense in health care, given that patients make life-or-death decisions based on quality ratings assigned to hospitals.
"You'd think, given the stakes, you'd have more confidence that the data is reliable," he said.