Should performance data rely on patient reports?
Some docs advocate using surveys, not records
When it comes to measuring health care performance, even simple indicators can present major headaches. For example, a parent may take a child to a clinic or school for needed vaccines. Meanwhile, records from the pediatrician show the child isn’t properly immunized.
Some physicians and outcomes experts are advocating a low-tech, less expensive way of solving data collection problems: Ask the patients themselves. Surveys also allow patients to answer questions about care that go beyond mere calculations: Did you receive adequate education about your condition? How much is your chronic illness interfering with your daily activities? Did the treatment alleviate your symptoms?
Including the patient’s experience
"There’s a patient experience of care that needs to be incorporated into the performance system overall," says Ted von Glahn, a consultant with the Foundation for Accountability (FACCT) in Portland, OR, a coalition of consumer and purchaser groups that develops condition-specific performance measures. FACCT is testing a survey, titled FACCT/ONE, to answer questions about how patients live with chronic illness. The first surveys focus on diabetes, asthma, and coronary artery disease.
In an article in the Archives of Family Medicine, Arch G. Mainous III, PhD, of the department of family medicine at the Medical University of South Carolina in Charleston, argues that surveys could replace some Health Plan Employer Data and Information Set (HEDIS) data collection. They would provide standardized outcomes data while avoiding the accuracy problems of administrative data and medical record reviews, he states.
Moreover, the surveys could ask patients why they didn’t receive needed services, such as mammograms, states Mainous and his co-author, Jeffrey Talbert, PhD, of the Martin School for Public Policy at the University of Kentucky in Lexington.
That dramatic shift in data collection isn’t likely. von Glahn and others cite a need for multiple sources of information on performance measurement. But surveys are an important component, he says. "The key issue is asking the question we heretofore haven’t asked: What happens to people who are sick?’" says von Glahn.
An easy way to collect data
Patient surveys themselves are nothing new. It’s the only way to find out about patient satisfaction and health status. And it’s a simple way to gather other information.
The National Committee for Quality Assurance (NCQA) in Washington, DC, already uses surveys to measure flu shots in older adults, a quality indicator that is required by HEDIS. After all, the shots are available in grocery stores, pharmacies, churches, and synagogues. Likewise, several pediatric measures under consideration are likely to be gathered through surveys, as is the Health of Seniors measure. (See related article, p. 9. For more information about pediatric measures, see Patient Satisfaction & Outcomes Management, December 1998, pp. 125-138.)
The Pacific Business Group on Health (PBGH) in San Francisco collects a more extensive range of health care performance information from patient surveys for its comparisons of medical groups. For example, patients report whether they received cholesterol screening. The business coalition is also measuring health status changes and satisfaction.
"We consider this just one method of collecting data," says Anne Castles, MA, MPH, senior project manager for PBGH. "It is the easiest way because of the paucity of information you can get via administrative data systems. But we don’t want to rely exclusively on patient surveys." She says. "What we would like to see over the long term is medical groups making the commitment to building data system infrastructure so that evaluation of clinical data becomes far easier, even easier than getting it from patient surveys."
How accurate are patient reports on the care they received? PBGH plans to find out with a special survey of a subgroup of patients with diabetes, heart disease, low back pain or sciatica, and asthma or lung disease. "We are asking for their consent to go into their medical chart to validate what they’re telling us [about their care on the surveys]," says Castles. "We will have a lot more information about the accuracy of patient report on these indicators."
Based on prior research, FACCT considers the patient reports on diabetic foot exams and eye exams to be reliable, says von Glahn. Those items, which are difficult to collect from administrative data or even medical records, have been incorporated into FACCT/ONE.
Who answers the surveys?
Even if patients remember the care they received and respond accurately on surveys, that method has its drawbacks, notes Castles. "If you’re not capturing the whole population [through the survey] then you always have to be concerned about which patients are responding and how they may differ from patients who are not responding. That’s a key methodological concern," she says.
"The best way to capture high-quality data is for providers to make investments in information systems," Castles says. "We’ll always have to use them to measure satisfaction, but in terms of other things, we very much prefer relying on high quality clinical data systems. Surveys are enlightening in many respects about quality. But they’re not going to give us the whole picture."
Still, the surveys can bolster the outcomes information that is available — without the burden caused by medical record reviews, says von Glahn. "It’s particularly nice to have this as another tool, given the [current] shortcomings of all of electronic information systems."
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content