Do your Home Health Consumer Assessment of Healthcare Providers and Systems (HHCAHPS) results published on Home Health Compare sometimes not match results HHCAHPS vendor survey results? Ever wonder why?
 
CMS announced April 23 that the reason for such discrepancies is likely attributable to the fact that vendor results are not statistically adjusted for patient mix as is done by the federal Medicare agency. Or, if vendors do adjustments to the data for case or patient mix, the adjustments are made with older coefficients that aren’t applicable to current HHCAHPS data.
 
“CMS maintains that the HHCAHPS Survey results published on the Home Health Compare website are the only ‘official’ HHCAHPS patient survey results,” CMS says.
 
Patient mix adjustment coefficients are estimated each quarter, with each new set of HHCAHPS data that agencies submit and CMS posts quarterly on the HHCAHPS Survey website here: http://tinyurl.com/ogv4dly.
 
CMS then averages the last four quarters of adjusted scores to produce the current quarter’s published scores. Publicly reported HHCAHPS scores are adjusted for differences between an agency’s patient mix and the overall national composition of home health patients on these same characteristics, CMS says.
 
Agencies should check to see which patient characteristics CMS is adjusting each quarter against their particular patient mix. If the agency has more patients with mental health patients, for instance, the difference between the information CMS releases quarterly and the agency’s vendor data might be more pronounced, says Sarah Balmer, vice president of business development at Louisville, Ky.-based data vendor Deyta.
 
Still, agencies wishing to focus on quality assurance performance improvement (QAPI) projects or other outcomes improvement plans will likely find HHCAHPS and outcomes data from a vendor more meaningful as it can be relayed much faster than at a quarterly rate, Balmer says.
 
For instance, negative comments on HHCAHPS surveys are transmitted to an agency by Deyta in real time, she says. Therefore, agencies can be more proactive in improvement quality.
 
Patient characteristics influence scores
 
In 2009 and 2010, CMS and the HHCAHPS Coordination Team researched both the mode (mail, telephone and mail with follow-up telephone call), and patient mix data, or characteristics like age and diagnoses such as schizophrenia, to see how they affected patients’ assessment of home health care, CMS says.
 
They found that the mode had no impact on HHCAHPS results but patient characteristics did. CMS says it decided to make statistical adjustments to the HHCAHPS results using coefficients that indicate the tendency of patients with specific characteristics, like non-English speakers, education level and age, to respond to the survey questions more positively or negatively.
 
For instance, analyses of the data collected between October 2013 and September 2014 and currently reported on Home Health Compare showed patients between 50 and 64 were 2.8% less likely to provide the most positive response (between a 9 or 10 rating for the applicable measure) for the ‘overall rating’ HHCAHPS measure when compared with patients aged 65-74, CMS says. Thus, CMS included an adjustment factor for patients aged 50-64 that is positive 2.8%.
 
Similarly, patients diagnosed with schizophrenia were 4.5% less likely to report the most positive response for the ‘Willingness to recommend’ HHCAHPS measure, so the federal Medicare agency adjusted this factor positively 4.5%, CMS says.