Researchers from the Georgia Sickle Cell Data Collection (SCDC) Program, housed at the Georgia Health Policy Center, recently had two new publications highlighting the use of administrative data to ascertain sickle cell disease cases.
In the first paper, published in Public Health Reports, the currently used public health surveillance case definition — that relies on multiple sources of data — identified twice as many people with sickle cell disease as the single-source administrative database definitions during the same period.
- In California, 7,117 people met the surveillance case definition of sickle cell disease from 2016 through 2018, with 48% of these cases captured by the Medicaid case definition and 41% by the discharge case definition.
- In Georgia, 10,448 people met the surveillance case definition of SCD from 2016 through 2018, with 45% of cases captured by the Medicaid case definition and 51% by the discharge case definition.
- Proportions of cases captured differed by year, year of birth, and length of Medicaid enrollment.
“These findings show us that there are trade-offs in using single-source administrative databases for policy and programmatic decisions for sickle cell disease,” says coauthor Angela Snyder, Ph.D., principal investigator of Georgia SCDC Program. “When policymakers are trying to understand the prevalence of sickle cell disease in a geographic region for allocating health care resources, it is important to use surveillance data. But, these single-source administrative databases can provide actionable information on opportunities for improving quality of care in specific subpopulations or standardizing health care practices across hospitals.”
In the second paper, published in JMIR Public Health and Surveillance, the researchers found that adults identified as having sickle cell disease from administrative claims data — using the SCDC case definition — have a high probability of truly having the disease, especially among hospitals that have active sickle cell disease programs.
- There were 354 people from Alabama and 865 from Georgia who were identified with sickle cell disease over a five-year time period, yielding a positive predictive value of 91% for Alabama and 87% for Georgia.
- Ove a three-year period, using data from three states (Alabama, Georgia, and Wisconsin), a total of 1,432 individuals from these states were identified. The three-year positive predictive values were 92%, 93%, and 81%, respectively.
“Positive predictive values consistently over 80%, even among older age groups, are important to ensure we can track survival and other sickle cell disease outcomes over individuals’ lifespans with reasonable accuracy,” says Snyder, who was a coauthor on this paper, too. “These findings add to the data supporting state programs seeking optimal resource allocation to improve outcomes and quality of life for those living with sickle cell disease. Administrative data are especially important in identifying adults who are too old to have been screened for sickle cell disease at birth and for those adults who have little access to specialized sickle cell disease treatment centers.”