AT THE STS ANNUAL MEETING
HOUSTON – Parents of children with congenital heart disease cite survival statistics, surgeon-specific experience, and complication rates as the three most important congenital heart surgery outcome measures to report publicly, results from a large survey show.
“Recently, an increasing demand for the public reporting of pediatric heart surgery outcomes has led to the development of several different reporting schemes, including a hospital star rating system and procedure-specific mortality data tables for the Society of Thoracic Surgeons benchmark operations,” study investigator Mallory L. Irons, MD , said during a press briefing at the annual meeting of the Society of Thoracic Surgeons. “However, despite the availability of these reporting schemes, there remain unanswered questions about the optimal format and content of public reporting for pediatric heart surgery outcomes.”
Dr. Irons, who is an integrated cardiac surgery resident at the Hospital of the University of Pennsylvania, Philadelphia, noted that the star rating system used by the STS assigns one, two, or three stars to hospitals performing pediatric heart surgery, with two stars representing expected performance, three stars representing better than expected performance, and one star representing worse performance than expected. In contrast to the STS star rating system, she continued, the Pennsylvania Health Care Cost Containment Council has adopted a different approach, reporting tabular hospital-specific discharge mortality data for each STS benchmark operation separately.
“These procedures have been determined to represent more homogenous patient groups,” she said. “How data [are] presented may be just as important as the data itself. Noticeably absent from these frameworks is parent input regarding the information that may be necessary to make an informed choice about their child’s care.
“Failure to consider this perspective may contribute to suboptimal reporting and misunderstanding of the data presented. The goal of the current study was to identify parent preferences regarding the content of pediatric heart surgery outcomes reports, specifically with respect to the type of data that is included as well as the optimal format for presentation of this data.”
She and her associates developed a 43-question survey for 1,862 parents of children born with heart defects. The parents were invited to participate through member lists from patient advocacy groups Mended Little Hearts , the Pediatric Congenital Heart Association , and Sisters by Heart , and from a cohort of patients who underwent surgical correction for an STS benchmark procedure at the Children’s Hospital of Philadelphia after Jan. 1, 2007.
Of the 1,862 parents, 1,281 (69%) provided complete responses for inclusion in the final analysis. The mean age of respondents was 37 years, 92% were mothers of children with congenital heart disease, and 92% were white. “Most reported household incomes in excess of $65,000 per year, but lower income households were also significantly represented,” Dr. Irons added. More than half of the children (57%) were diagnosed with CHD prenatally, 63% underwent an initial repair in the neonatal period, and 60% of families were referred to a cardiac surgical center by a physician, while 23% were transferred from their birth hospital.
When the respondents were asked to rank categories of outcome measures or other types of information to include in an optimal public reporting scheme on a scale of 1 (most important) to 8 (least important), they identified survival statistics, surgeons’ experience with congenital heart surgery, and complication rates as the three most important. These preferences were stable across differences in education levels, household incomes, and race or ethnicity, Dr. Irons said.
Presented with three display formats for hospital-specific mortality rates, most parents (89%) identified a numerical procedure-based approach as the best format, and more than half (60%) identified the hospital star rating system as the worst potential format to display mortality data. These preferences also were stable across differences in education levels, household incomes, and race or ethnicity.
Dr. Irons acknowledged certain limitations of the study, including its retrospective design, and the fact that it lacked input from parents who do not have Internet access. “Similarly, the availability of the survey in English only prevents non-native English speakers from contributing,” she said. “Finally, while we have identified what types of information parents want, we must acknowledge that small case numbers limit the presentation of certain types of data, such as surgeon-specific outcomes, in a statistically meaningful way.”
In her opinion, an optimal reporting system requires a common database in which to collect and analyze data, a robust methodology for risk adjustment, and a way to compare performance across a field that is inherently heterogeneous.
“The ideal public reporting system should be valid as well as easily understood by all stakeholders,” Dr. Irons concluded. “We must recognize that just because parents want certain data, [it] does not mean that we can provide it in a meaningful, statistically valid way. Perhaps the most important takeaway from our study is the importance of involving parents and other stakeholders in the design and planning of methodologies for public reporting of congenital heart surgery outcomes. Ultimately, the optimal platform will represent a melding of what parents want, and what is statistically meaningful and valid.”
Jeffrey P. Jacobs, MD , a pediatric heart surgeon at Johns Hopkins University, Baltimore, who also chairs the STS Workforce on National Databases , characterized the study as “an extremely important paper that examines the format and structure of how it might be best to share information about cardiac surgical outcomes with patients and their families.” He said the STS “has studied multiple different strategies, and currently we use a method where we report outcomes using a categorical system of star ratings and augment that with detailed specific data with point estimates for confidence intervals. We continue to study and explore better ways to share this information with our patients and with their families.”
Dr. Irons reported having no financial disclosures.
