The research team, led by Colleen C. McLaughlin of the Center for Child Health Equity and Outcomes Research at Nationwide Children’s Hospital in Columbus, Ohio, drew on the Transformed Medicaid Statistical Information System, the federal data repository that captures claims from Medicaid and the Children’s Health Insurance Program, known as CHIP. Together these programs insure roughly half of all American children, making them the only data source capable of painting a truly national picture of pediatric healthcare utilization. The investigators analyzed claims from 2017 through 2019, deliberately stopping before the COVID-19 pandemic distorted enrollment figures and care patterns, and identified overnight, attended polysomnography using the standard billing codes for the procedure.

The scale of the analysis was extraordinary. Among more than 51.7 million children enrolled with unrestricted benefits at some point during the three-year window, the researchers identified 482,688 polysomnograms. After careful exclusions for incomplete records and data quality problems, including the removal of Rhode Island and Vermont, 478,568 tests remained in the state-level analysis, representing 99.1 percent of all identified studies. The denominator of 95.5 million person-years of enrollment yielded a national utilization rate of 50.1 sleep studies per 10,000 Medicaid-enrolled children per year. That headline number, however, conceals a startling degree of local variation.

At the state level, age-adjusted rates ranged from a low of 23 polysomnograms per 10,000 person-years in Kansas to a high of 93 per 10,000 in Michigan, a fourfold difference between neighboring Midwestern states. Mapping the data at an even finer resolution, using Census Bureau Public Use Microdata Areas that each contain at least 100,000 residents, revealed that the East North Central Census Division, anchored by Michigan and the Great Lakes industrial states, had the highest mean local rate at 64.9 per 10,000. Sixteen of the 25 microdata areas in the top 1 percent of utilization, defined as rates above 140 per 10,000, sat within that single division, and 15 of those were in Michigan alone. By contrast, the Pacific and West South Central divisions, encompassing states such as California, Oklahoma, and Texas, recorded the lowest rates, at 39.6 and 43.7 per 10,000 respectively. Fourteen of the 25 lowest-rate microdata areas were in Texas.

The statistical machinery behind these findings was rigorous. The team standardized rates to the age distribution of the enrolled population, applied the Jenks natural breaks optimization method to group mapped values, and used the Games–Howell test, which tolerates unequal variances and adjusts for multiple comparisons, to confirm that the differences between Census divisions were statistically significant. Multivariable Poisson regression with clustered error estimators then linked utilization to neighborhood characteristics at the ZIP code level, adjusting for beneficiary age. Because Medicaid data quality varies by state, the researchers also ran a sensitivity analysis excluding the 14 states flagged by the federal Medicaid Data Quality Atlas as having unusable enrollment or claims data for at least one study year, along with six states reporting high rates of missing or invalid procedure codes. The core conclusions held firm, with mean rates for every ZIP code grouping shifting by less than 10 percent.

Two demographic patterns emerged with particular clarity. First, urbanicity mattered: metropolitan ZIP codes, which accounted for 84.1 percent of all the sleep studies in the dataset, had the highest adjusted mean rate at 41.3 per 10,000, and rates declined steadily as communities became more rural, with isolated rural towns showing roughly 6 to 9.7 fewer studies per 10,000 than metropolitan areas. Second, and more troubling, the racial and ethnic composition of a neighborhood was strongly associated with testing rates. ZIP codes where at least 90 percent of residents were White non-Hispanic had the highest rate, 50.8 per 10,000, while ZIP codes where a majority of residents belonged to other racial and ethnic groups, including Asian, Native Hawaiian and Pacific Islander, and American Indian or Alaskan Native populations, had the lowest, at 27.6 per 10,000. Neighborhoods with larger populations of color received between 4.2 and 23.2 fewer sleep studies per 10,000 than the predominantly White areas.

The authors were careful to note that some of these differences, while statistically significant, were modest in absolute terms, and that the analysis relied on ZIP code level demographics rather than individual self-reported race and ethnicity, which most states record too poorly to use. Still, the direction of the disparity is consistent with a growing body of evidence. Prior studies have suggested that polysomnography may itself contribute to racial and ethnic disparities in tonsillectomy rates among Medicaid-insured children, because requiring the test before surgery adds cost, delays treatment, and increases the risk that families lose contact with the care system altogether. Research in Ontario, Canada, has shown that shorter travel times to sleep clinics increase the likelihood that children receive the test, and longer travel distances to pediatric sleep laboratories, which are scarcer than adult facilities, are a plausible explanation for the rural deficit observed in the American data.

The study also situates the geographic patterns within a genuine clinical controversy. Professional societies disagree about when a child with habitual snoring needs a sleep study before tonsillectomy. The American Academy of Sleep Medicine and the American Academy of Pediatrics have recommended polysomnography for all children with suspected obstructive sleep apnea, while the American Academy of Otolaryngology–Head and Neck Surgery takes a narrower position in its 2019 guideline, recommending the test mainly for children with elevated anesthesia or surgical risks, such as obesity or Down syndrome, or when the need for surgery is unclear. Because pediatric polysomnography requires specialized equipment and measurement algorithms distinct from adult studies, and because home sleep tests are not approved for children, capacity is limited, and clinicians who know their local labs have long waitlists may simply skip the test, a workaround the pediatrics guideline explicitly permits when access is a barrier.

What makes the new findings so consequential is that they provide the empirical baseline that has been missing from that debate. Without knowing where testing happens and for whom, professional groups and policymakers could not judge whether guideline recommendations were feasible or equitable across the country. The authors argue that their data can now inform efforts to reduce variation in the evaluation and treatment of children with sleep-disordered breathing, expand access to polysomnography, and refine referral guidelines to reflect real-world availability. They also caution that Medicaid coverage itself varies by state eligibility policy and by the geography of poverty, so the estimates cannot be used to plan care for individual patients or to reconfigure services in any single locality.

For the millions of American families navigating a child’s sleep problems on public insurance, the study is a stark reminder that a diagnosis considered routine in Michigan may be far harder to obtain in Kansas, California, or Texas, and that the odds of getting tested are shaped by the demographics of the ZIP code a child calls home. The researchers call for further investigation into the availability of pediatric sleep laboratories and the structural barriers, from transportation to provider location data gaps, that keep children from the tests that could change their trajectories. Untangling those barriers, they conclude, is essential if the promise of definitive sleep diagnosis is to reach every child who needs it, regardless of geography or background.

Subject of Research: Geographic and racial/ethnic disparities in pediatric polysomnography utilization among US children enrolled in Medicaid

Article Title: Geographic and racial/ethnic patterns of polysomnography use among children enrolled in Medicaid, 2017–2019

Article References: Geographic and racial/ethnic patterns of polysomnography use among children enrolled in Medicaid, 2017–2019. (n.d.). https://doi.org/10.1007/s44470-026-00149-w

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00149-w

Keywords: polysomnography, pediatric sleep apnea, Medicaid, health disparities, geographic variation, obstructive sleep-disordered breathing, rural health, healthcare access, sleep medicine, CHIP, tonsillectomy, health equity