A federal proposal to base 30% of hospital "incentive" payments on patient experience scores is meeting with strong resistance, in part because of concerns that culturally, patients in some parts of the country are just harder to please.
"We have to look at the data, but there's a perception that some parts of the country do better than other parts," says Blair Childs, senior vice president for public affairs for Premier Healthcare Alliance, a performance improvement group with 2,400 hospitals. "It might be less a function of what the hospital does than the attitude of the population. For example, In New York City, they are grumpier than they are in Minneapolis."
For example, patients in New York, New Jersey, and Pennsylvania, one of nine survey regions in the country, are indeed less likely to answer "definitely yes" that they would recommend the hospital where they received care to friends and family than patients in the other eight regions, according to charts from 2010 Press Ganey surveys. Other possible answers are "probably yes," "probably no," and "definitely no."
The Press Ganey survey found that 73% of patients in New England states, from Connecticut and Massachusetts, Maine, New Hampshire, Rhode Island and Vermont, are likely to recommend, versus 64% in New York, New Jersey and Pennsylvania.
Patients in Great Plains states such as Iowa, Kansas, Minnesota and Nebraska are among the likeliest to recommend. They scored 72%, while south central region patients in Arkansas, Louisiana, Oklahoma and Texas score 71%. 19 states along the lower eastern seaboard, in the southeast, and in the west from California to Washington were 68% likely.
Press Ganey surveyed patients with that and other questions used in the HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems) questionnaire. Under the federal proposal for Hospital Inpatient Value-Based Purchasing incentive payments, patient experience scores will be used to determine 30% of reimbursement for discharges at more than 3,000 hospitals.
The Centers for Medicare & Medicaid released a 126-page document Jan. 7 detailing how it will make value-based purchasing incentives. There is a 60-day comment period which ends March 8. The nine-month measurement period to determine the scores begins July 1.
Indeed, New York hospitals are very concerned about the 30%. "We believe that even 20% is far too high," says Vincent Fitts, associate vice president of informatics for the Greater New York Hospital Association.
"Wholly separate from the apparent regional bias in HCAHPS results and the fact that large urban hospitals generally fare worse, the current formula penalizes hospitals that treat a higher-than-average percentage of patients for whom English is a secondary language – even though they tend to express satisfaction with their care," Fitts says.
"We simply don't think HCAHPS scores accurately reflect patient experience in the New York metropolitan area."
The American Hospital Association has also weighed in. "We are still evaluating the proposal, but our first reaction is that 30% seems to be a bit high for the HCAHPS proportion," says Marie Watteau, AHA spokeswoman.
In addition to a so-called "grumpiness" variable, Press Ganey surveys indicate that size is a factor in patient responses. Patients in the smallest hospitals, those with 50 or fewer beds, are the most likely to recommend their care (73%), perhaps out of community loyalty and because they know their caregivers personally, while patients in hospitals with 600 or more second most likely to recommend. Hospitals with 150-299 beds have patients least likely to recommend, 66%.
Based on a separate category addressing the patient's communication with nurses, the Press Ganey survey found that patients in hospitals with 50 or fewer beds are also most likely to be pleased with their care enough to recommend the hospital.
But the percentages who answered positively to this question declined as the number of beds per hospital went up. For hospitals with 51-149 beds, positive response to communication with nurses was only 76% and for hospitals with 600 or more beds, 73%.
Dennis Kaltenberg, chief science officer for Press Ganey, says it's unclear why survey responses vary so much by region and bed size. "It's difficult to say whether it's the way patients answer questions, or it's the actual (quality of) the services being delivered," he says.
Childs says that assigning a weight of 30% to payment based on patient experience scores is controversial, and Premier is still studying the issue before issuing its written comments to CMS.
"I expect this will be a hotly-debated topic," Childs says.
Under the proposed regulations, hospitals will start receiving adjusted payments based on scores received on the following:
• 17 process of care measures. These include whether a patient with an acute myocardial infarction received an aspirin at discharge, or whether patients received prophylactic antibiotics prior to surgery.
• 8 patient experience of care measures asked within the HCAHPS survey, such as how well they perceived communication with their nurses and doctors, the responsiveness of hospital staff and their perception of the adequacy of pain management. HCAHPS surveys are administered randomly to patients between 48 hours and six weeks after discharge, and is not restricted to Medicare beneficiaries.
For discharges in FY 2014, which begins Oct. 1, 2013, CMS proposes to add three 30-day mortality scores, eight hospital acquired condition measures and nine patient safety indicators, inpatient quality indicators or composite measures established by the Agency for Healthcare Research and Quality.
The period of evaluation would begin this July and extend nine months to March 31, 2012. Scores achieved during this period would be relayed to the hospital at least 60 days prior to Oct. 1, 2012. Payment based on these scores would begin for the fiscal year starting Oct. 1, 2012.
CMS declined comment on the controversy. In its 126-page proposal, it says in part, "We believe assigning a 30% weight to the patient experience of care domain is appropriate because the HCAHPS measure is comprised of eight dimensions that address different aspects of patient satisfaction."
The proposal also says that "Measures or measurement domains need not be given equal weight, but over time, scoring methodologies should be more weighted towards outcome, patient experience and functional status measures."
In summary, Childs of Premier notes that "it's a different world for hospitals now, with a different approach for payment and performance is being valued in a new way. Management of hospitals have really got to pay attention here."