Abstract
Objectives
To examine disparities in the diagnosis of leukodystrophies including geographic factors and access to specialty centers.
Study design
Retrospective cohort study of pediatric patients admitted to Pediatric Health Information System hospitals. Leukodystrophy patients were identified with ICD-10-CM diagnostic codes for any of four leukodystrophies (X-linked adrenoleukodystrophy, Hurler disease, Krabbe disease, and metachromatic leukodystrophy). We used 3-level hierarchical generalized logistic modeling to predict diagnosis of a leukodystrophy based on distance travelled for hospital, neighborhood composition, urban/rural context, and access to specialty center.
Results
We identified 501 patients with leukodystrophy Patients seen at a leukodystrophy center of excellence (COE) hospital were 1.73 times more likely to be diagnosed than patients at non-COE hospitals. Patients that travelled farther were more likely to be diagnosed than those who travelled shorter. Patients living in a Health Professionals Shortage Area (HPSA) zip code were 0.86 times less likely to be diagnosed than those living in a non-HPSA zip code.
Conclusions
Geographic factors affect the diagnosis of leukodystrophies in pediatric patients, particularly in regard to access to a center with expertise in leukodystrophies. Our findings suggest a need for improving access to pediatric specialists and possibly deploying specialists or diagnostic testing more broadly.
Although individually uncommon, with more than 7,000 known distinct genetic causes, rare diseases affect an estimated 8% of the population and 25 million individuals in the U.S. alone.1 Diagnosis has historically been challenging, treatments and therapies are few, yet children with genetic diseases account for the majority of admissions to children’s hospitals.2,3 Although recent advances in use of next-generation sequencing approaches have overcome many technical barriers to diagnosis,4 patients continue to face prolonged “diagnostic odysseys.”5,6 However, there has been a lack of research exploring the reasons for the diagnostic odyssey, and how to overcome the diagnosis gap.
Among the leukodystrophies evidence has recently identified that minority ethnic/racial group status is associated with underdiagnosis.7 Leukodystrophies are a group of rare genetic diseases affecting the white matter of the central nervous system (CNS).8 Leukodystrophies have been well studied over the past decade, including determinations of incidence, mortality, and health care burden.9,10 Diagnosis of leukodystrophies is critical for instituting time-sensitive treatments, clinical trial design, and development of new therapies.11,12 Similar to other rare pediatric diseases, diagnosis rates in leukodystrophies are only about 50%.9,13,14
The finding that minority patients are underdiagnosed with leukodystrophy provides an avenue for exploring causes of the diagnostic odyssey, with the leukodystrophies serving as a proxy more broadly for the numerous different pediatric rare genetic diseases. The goal of our study was to determine whether access to specialty centers was correlated with leukodystrophy diagnosis. We sought to determine rates of leukodystrophy diagnosis, considering geographical factors including distance travelled for hospital and/or specialty center care, neighborhood composition, and urban/rural context, while adjusting for other hospital- and patient-level social factors.
METHODS
We conducted a retrospective cohort study of patients admitted to a Pediatric Health Information System (PHIS) hospital with an International Classification of Diseases, Tenth Revision, clinical modification (ICD-10-CM) diagnosis of one of the following four leukodystrophies: metachromatic leukodystrophy (E75.25); X-linked adrenoleukodystrophy (E71.52x); Krabbe disease (E75.23); and Hurler disease (E76.01). Other leukodystrophies are coded using the ICD-10-CM code “sphingolipidoses” (E75.3), but E75.3 is used for a large number of genetically discrete disorders including vanishing white matter disease, Canavan disease, TUBB4A leukodystrophy, etc., as well as non-specifically for white matter disorders such as occur after chemotherapy or after hypoxic-ischemic injury;19 for this reason, such sphingolipidoses were not included in this study. Patient identification was between October 1, 2015, through September 30, 2018. The patient had to be age 18 years or younger at their first PHIS visit; and they had to be cared for at a PHIS hospital that saw at least one other leukodystrophy visit during the capture window. The Institutional Review Board of the University of Utah approved this study as exempt as non-human research.
The PHIS database contains information from 52 children’s hospitals in the U.S. affiliated with the Children’s Hospital Association (Lenexa, Kansas).15 PHIS data are deidentified but retain a patient specific identifier, permitting tracking of patient medical care and use. Patient and visit demographic data were collected including age, sex, race, insurance type, type of visit, estimated total visit cost, length of stay, and ICD coding and charge information. The majority of data consisted of inpatient hospital admissions; data from additional visit types such as emergency department, observation, ambulatory surgery, clinic and other visit types were also collected when available.
Patients with Leukodystrophy were identified from any PHIS visit (inpatient, observation, emergency department, ambulatory surgery, clinic visit) as having an ICD-10-CM code for one of the four leukodystrophies. Data collected on the cases included demographics at the first available PHIS visit in the collection time frame (quarter 4 of 2015 and onward), and a summary of all available PHIS visits, including visits prior to the patient identification time window, as well as clinic and other visit types. To control for some PHIS hospitals not participating in data collection and reporting for the entire study period, we adjusted data to account for the number of months that the hospital provided data to PHIS during the study period.
We designated 8 hospitals as specialty hospitals for leukodystrophy (“centers of excellence” [COE]), as determined by their listing as a center in the Global Leukodystrophy Initiative (GLIA) and/or in the Leukodystrophy Care Network (LCN). These hospitals were Ann & Robert H. Lurie Children's Hospital; Children's Healthcare of Atlanta; University of Utah/Primary Children's Hospital; Children's Hospital of Colorado; Children's National Health System; Children's Hospital of Pittsburgh; Children's Hospital of Philadelphia; and Lucile Packard Children's Hospital.
For the geographical determinants, we used the distance from the centroid of the zip code of the patient to the hospital. Neighborhood composition was determined by using the zip code’s median household income (for 2015), the zip code’s percentage of residents who were nonwhite, and whether the zip code was a Health Professional Shortage Area (HPSA). For urban/rural designation, we used a categorical variable with six categories, based on Rural Urban Commuting Area (RUCA) categories and accompanying codes: large rural: 4.0, 4.2; 5.0, 5.2, 6.0, 6.1; small rural: 7.0, 7.2, 7.3, 7.4, 8.0, 8.2, 8.3, 8.4, 9.0, 9.1, 9.2; isolated: 10.0, 10.2, 10.3, 10.4, 10.5, 10.6.; urban core: 1.0, 1.1; and other urban: 2.0, 2.1, 3.0, 4.1, 5.1, 7.1, 8.1, 10.1.
For the patient-level individual variables, we included age in years (at first admit during study period) and patient sex. We also adjusted for health care utilization by including the number of inpatient days (per months in PHIS) and the number of admissions (per months in PHIS). We included patient race in the following categories: white, non-Hispanic; Hispanic, all races; black, non-Hispanic; Asian, non-Hispanic; and other races, non-Hispanic. For payer, we included government, commercial/private, and other payer.
We used a three-level hierarchical generalized logistic model (HGLM) to analyze the correlation of different variables with a diagnosis of leukodystrophy. Independent variables at the patient, zip code, and hospital levels were included in the model. We chose to use HGLM for this study because our data had a three-level structure: 1,070,062 children at level 1 (total number of children admitted to a PHIS hospital during study period), nested within 36,274 zip codes at level 2, which are nested within 52 hospitals at level 3. In contrast, traditional regression techniques could result in inaccurate parameter estimates when examining effects at multiple levels.16 Because the dependent variable (leukodystrophy diagnosis) was binary (0 or 1), we used the Bernoulli HGLM, which is the most appropriate for analyzing multi-level data to predict a binary outcome.17
For the sensitivity analysis, we included only patients with complex chronic conditions in the HGLM model. To define the subgroup of children with complex chronic conditions, we included all patients with a patient medical complexity algorithm (PMCA) indicating that the child had a “complex chronic disease” and/or a “complex chronic condition” based on the Feudtner method.18 In this sensitivity analysis, we included 439,024 children with complex chronic conditions, nested in 30,185 zip codes, which were nested in 52 hospitals.
RESULTS
We identified a total of 1,070,062 patients in PHIS during the nearly three year study period with complete data available for all of the analyses variables. The total duration of data collection time was a maximum of 36 months; the mean length of time of data submission of the participating hospitals was 34.65 months.
There were 501 patients with leukodystrophy represented in the PHIS database over the three year time period. The cohort was composed of 197 females (39%); with a mean age of 6.4 years (range 0 - 18); and the most common racial groups represented were white non-Hispanic, white Hispanic, and black non-Hispanic (respectively 241 patients [53%]; 110 [24%]; and 52 [11%]) (Table I). Leukodystrophies included for analyses were Hurler disease (IDUA gene; 38% of cohort), X-linked adrenoleukodystrophy (ALD; ABCD1 gene; 25% of cohort), metachromatic leukodystrophy (MLD; ARSA gene; 25%), and Krabbe disease (GALC gene; 12%). The 4 leukodystrophies used for this study have specific ICD-10-CM designations.
Table 1:
Demographics of the leukodystrophy, overall PHIS and complex chronic condition (CCC) cohorts. Note: Due to missing data at the patient-, zip code- and hospital-levels, the total number of cases included in subsequent tables and analysis (including in the HGLM analyses) are fewer then reported here. The CCC cohort is only used in the sensitivity analysis.
| Leukodystrophy cohort |
PHIS (non- leukodystrophy) cohort |
PHIS CCC (non- leukodystrophy) cohort |
|
|---|---|---|---|
| N | 501 | 1,245,391 | 511,351 |
| Age at first admit in study period | |||
| Mean (min, max) | 6.36 (0-18) | 5.42 (0-18) | 6.62 (0-18) |
| Inpatient days (per time in PHIS) | |||
| Mean (min, max) | 0.045 (.0009-1.03) | 0.186 (.0009-12.23) | 0.340 (.0009-12.23) |
| Number of admits (per time in PHIS) | |||
| Mean (min, max) | 0.113 (.03-2.0) | 0.071 (.03-5) | 0.03 (.082-.122) |
| Sex Number (%) | |||
| Male | 304 (60.7) | 665143 (53.4) | 247148 (53.6) |
| Female | 197 (39.3) | 579908 (46.6) | 237050 (46.4) |
| Race/Ethnicity Number (%) | |||
| White, non-Hispanic | 241 (52.9) | 532250 (48.1) | 232477 (50.6) |
| Hispanic | 110 (23.6) | 262348 (23.2) | 97741 (20.9) |
| Black, non-Hispanic | 52 (11.4) | 195202 (17.6) | 82765 (18.0) |
| Asian, Non-Hispanic | 10 (2.2) | 39588 (3.6) | 14488 (3.2) |
| Other race, Non-Hispanic | 43 (10.8) | 77793 (8.0) | 32340 (7.9) |
| Payer Status Number (%) | |||
| Private Payer | 168 (33.8) | 502058 (41.1) | 204964 (40.8) |
| Government Payer | 293 (59.) | 662388 (54.2) | 274555 (54.6) |
| Other Payer | 168 (33.8) | 58124 (4.8) | 23131 (4.6) |
| Leukodystrophy Type Number (%) | |||
| ALD | 126 (25.1%) | 0 (0) | 0 (0) |
| MLD | 124 (24.8) | 0 (0) | 0 (0) |
| Krabbe | 62 (12.4) | 0 (0) | 0 (0) |
| Hurler | 190 (37.9) | 0 (0) | 0 (0) |
The characteristics of the variable determinations in the 3-level hierarchical generalized logistic model (HGLM) are shown in Table 2. The HGLM sorts and clusters lower level data in a hierarchy of successively higher-level units; analysis of the data takes into consideration their clustering. Patients were nested into zip codes (36,274 zip codes). At the zip code level, average household income was $44,384.85; 29% of the patients were non-White; 35% lived in a HPSA; and the average distance traveled to a PHIS hospital was 262.9 miles. The proportions of patients living in urban/rural areas was (using urban core as a reference): large rural, 13%; small rural 9%; isolated 12%; other urban 15%.
Table 2.
Characteristics of the variables in the three-level hierarchical generalized logistic model (HGLM).
| VARIABLE | N | Mean | SD | Min. | Max. |
|---|---|---|---|---|---|
| Level 3-Hospital | |||||
| Center of Excellence | 52 | 0.15 | 0.36 | 0 | 1 |
| Hospital months in PHIS | 52 | 34.65 | 4.96 | 9 | 36 |
| Level 2-Zip Code | |||||
| Median household Income (U.S. $) | 36274 | 44384.85 | 18528.02 | 6330 | 194591 |
| Proportion non-White | 36274 | 0.29 | 0.25 | 0 | 1 |
| Proportion HPSA | 36274 | 0.35 | 0.48 | 0 | 1 |
| Distance to hospital (miles) | 36274 | 262.92 | 425.96 | 0 | 5091 |
| Urban/Rural Context | |||||
| Urban core | REF | REF | REF | REF | REF |
| Large rural | 36274 | 0.13 | 0.33 | 0 | 1 |
| Small rural | 36274 | 0.09 | 0.28 | 0 | 1 |
| Isolated | 36274 | 0.12 | 0.32 | 0 | 1 |
| Other urban | 36274 | 0.15 | 0.36 | 0 | 1 |
| Level 1-Patient | |||||
| Age at first admit in study period (years) | 1070062 | 5.42 | 6.02 | 0 | 18 |
| Male (v.ersus female) | 1069782 | 0.53 | 0.5 | 0 | 1 |
| Inpatient days (per time in PHIS) | 1066747 | 0.02 | 0.08 | 0 | 12.23 |
| Number of admits (per time in PHIS) | 1066747 | 0.07 | 0.11 | 0.03 | 5 |
| Patient Race/Ethnicity | |||||
| White, non-Hispanic | REF | REF | REF | REF | REF |
| Hispanic | 970094 | 0.22 | 0.42 | 0 | 1 |
| Black, non-Hispanic | 948134 | 0.18 | 0.39 | 0 | 1 |
| Asian, non-Hispanic | 948126 | 0.04 | 0.18 | 0 | 1 |
| Other race, non-Hispanic | 827568 | 0.08 | 0.27 | 0 | 1 |
| Patient Payer Status | |||||
| Private payer | REF | REF | REF | REF | REF |
| Government payer | 1051169 | 0.55 | 0.5 | 0 | 1 |
| Other payer | 1051169 | 0.04 | 0.21 | 0 | 1 |
| ICD-10 Code for leukodystrophy | |||||
| [DEPENDENT VARIABLE] | 1070062 | 0 | 0.02 | 0 | 1 |
Note: this table reports descriptive statistics for the cases included in the HGLM analyses; no missing data were permitted for analysis across the included variables.
Abbreviations: HPSA, Health Professional Shortage Area.
The HGLM analysis predicting odds of a leukodystrophy diagnosis is shown in Table 3. Patients being seen at a leukodystrophy COE hospital were 1.73 times more likely to be diagnosed than patients at non-COE hospitals. Patients that travelled farther were more likely to be diagnosed than those who travelled shorter. Patients living in a HPSA zip code were 0.86 times less likely to be diagnosed than those living in a non-HPSA zip code. We found that as the percentage of minority (non-white) population in the zip code increased by one standard deviation, patients were 0.58 times less likely to be diagnosed. Neighborhood median income was not significantly associated with odds of diagnosis. Patients living in large rural, small rural, and isolated zip codes were 1.18, 1.32, and 1.60 times more likely to have a leukodystrophy diagnosis than patients living in urban core zip codes.
Table 3.
Results from three-level hierarchal generalized logistic model (HGLM) analysis predicting odds of leukodystrophy (n=1,070,062 children)
| Odds | |||||
|---|---|---|---|---|---|
| Coeff. | Ratio | SE | t-ratio | p-value | |
| Hospital Level | |||||
| Intercept | −7.761 | 0.0004 | 0.076 | −101.76 | <0.001 |
| Center of Excellence | 0.550 | 1.734 | 0.064 | 8.539 | <0.001 |
| Hospital months | 0.011 | 1.011 | 0.009 | 1.312 | 0.196 |
| Zip Code Level | |||||
| Median household income | 0.0002 | 1.0000 | 0.0002 | 1.107 | 0.268 |
| HPSA flag | −0.149 | 0.862 | 0.058 | −2.547 | 0.011 |
| Distance to hospital | 0.0007 | 1.0007 | 0.0001 | 5.247 | <0.001 |
| Proportion non-White | −0.539 | 0.583 | 0.174 | −3.102 | 0.002 |
| Urban/Rural Context | |||||
| Urban core | REF | REF | REF | REF | |
| Large rural | 0.169 | 1.184 | 0.066 | 2.552 | 0.011 |
| Small rural | 0.280 | 1.323 | 0.077 | 3.617 | <0.001 |
| Isolated | 0.470 | 1.600 | 0.113 | 4.171 | <0.001 |
| Other urban | −0.028 | 0.973 | 0.060 | −0.457 | 0.648 |
| Patient Level | |||||
| Age at first admit in study period | 0.020 | 1.020 | 0.003 | 6.121 | <0.001 |
| Male (versus female) | 0.299 | 1.348 | 0.048 | 6.192 | <0.001 |
| Inpatient days (per time in PHIS) | 0.385 | 1.470 | 0.113 | 3.395 | <0.001 |
| Number of admits (per time in PHIS) | 1.435 | 4.200 | 0.133 | 10.784 | <0.001 |
| Patient Race/Ethnicity | |||||
| White, non-Hispanic | REF | REF | REF | REF | |
| Hispanic | 0.013 | 1.0133 | 0.061 | 0.216 | 0.829 |
| Black, non-Hispanic | −0.415 | 0.660 | 0.064 | −6.530 | <0.001 |
| Asian, non-Hispanic | −0.349 | 0.706 | 0.160 | −2.172 | 0.030 |
| Other race, Non-Hispanic | 0.172 | 1.188 | 0.063 | 2.722 | 0.006 |
| Patient Payer Status | |||||
| Private payer | REF | REF | REF | REF | |
| Government payer | 0.374 | 1.554 | 0.046 | 9.559 | <0.001 |
| Other payer | 0.374 | 1.1454 | 0.117 | 3.198 | 0.001 |
At the patient-level, male patients were 1.35 times more likely than female patients to be diagnosed. As patients’ age increased, they were more likely to be diagnosed (OR 1.02). Patients who spent more days in the hospital, or who had more admissions, during the study period, were more likely to be diagnosed. Black or Asian (non-Hispanic) patients were less likely to be diagnosed; whereas other race (non-Hispanic) patients were 1.19 time more likely to be diagnosed. Patients with government insurance were 1.55 times more likely, and patients with other insurance were 1.15 times more likely, than patients with private insurance, to be diagnosed.
We also performed an HGLM analysis on the subgroup of PHIS patients with complex chronic conditions.20 These findings were generally the same in terms of direction and significance as when all PHIS children were used. The COE finding retained its statistical significance and positive directionality. In terms of the geographic variables of interest, the distance finding retained statistical significance (p=.029) and positive directionality, as did the findings for “small rural” (p=.018) and “isolated” (p=.003). “Large rural, ” HPSA, and percent non-white retained direction, but were no longer significant (p=.31; p=.082; and p=.163, respectively). At patient level analysis, black (p<.001) and other race (p=.015) stayed negative and significant. The findings for payer status were equivalent in terms of direction and significance (p<.01), as was the finding for number of admissions (p<.001). Inpatient days became negative and no longer significant (p=.191). Age became negative and non-significant (p=.755). Asian was still negative, but lost significance (P = .144).
DISCUSSION
We identified that geographic factors are associated with diagnosis of a leukodystrophy. In particular, we showed that access to a leukodystrophy specialist, as indicated by access to a leukodystrophy center of excellence hospital, was positively associated with being diagnosed with a leukodystrophy. We observed that patients diagnosed with a leukodystrophy travel farther for hospital care. Diagnosed patients were more likely to live in majority white neighborhoods, and less likely to live in HPSA zones.
We also performed our analysis using children with complex chronic conditions, 20 instead of the entire PHIS cohort. This permitted us to discern whether our findings were specific effects related to diagnosis, or were more generally related to any child with complex medical needs. We confirmed our observations with the CCC cohort, indicating that reasons related to diagnosis, and not medical complexity or a chronic health condition, account for our findings.
We did observe that rural children were more likely to be diagnosed than urban children; this was despite controlling for effects of race/ethnicity, distance to hospital, and access to COE hospital. This was observed in both the full PHIS and CCC cohorts. The reason for this increased rate of diagnosis is unclear. One possibility would be higher disease allele incidences in rural populations. Variations in genetic diversity have been suggested to contribute to epidemiology of diseases including of rare genetic diseases,21 but other than specific founder mutation populations and diseases (eg, Holve et al).22 this hypothesis has not been tested more broadly for pediatric conditions.
Our work extends our previous observation of underdiagnosis in minority patients with leukodystrophy.7 In this current analysis, we adjusted for access to COE, but still found lower rates of diagnosis in minority patients. This suggests that more complex or poorly understood factors are responsible for the disparities in diagnosis. For example, there may be phenotypic differences in the clinical symptoms of disease. Disparities in other aspects of pediatric diseases have also been noted. For example, racial and socioeconomic factors contribute to outcomes in congenital heart disease23 and many of these patients have limited access to specialty centers.24 Our study is goes further to identify how geographic factors shape diagnosis in pediatric patients.
Strengths of this work are use of ICD-10-CM codes specific to genetically defined leukodystrophies. This work would not have been feasible using ICD-9-CM codes, which have low specificity for the diagnosis of an inherited leukodystrophy.19 With the use of the PHIS national database we were able to evaluate a much larger number of patients than is available at any one center, and to include data from a variety of location such as urban or rural areas. Limitations of this work include use of retrospective data and the fact that only three years of data were available because ICD-10-CM coding was not started until 2015. Even with the use of the PHIS data, the number of leukodystrophy patients was limited, and that affected our statistical power.
Overall, we identified geographic disparities in the diagnosis of leukodystrophies in pediatric patients, particularly in association with access to a center with expertise in leukodystrophies. Patients who were in medically-underserved areas (HSPA regions), or who could not travel to a specialist, were less likely to be diagnosed with a leukodystrophy. Similar to prior work,7 we showed that minority patients were also less likely to be diagnosed with a leukodytrophy, but in this current study showed that this underdiagnosis is not related to access to a COE. Our findings suggest a need for improving access to pediatric specialists, including in-person evaluation and/or remote assessments and guidance. In addition, diagnosis could be aided by testing the use of broadly deployed diagnostic testing. Some possibilities for example are use of next-generation sequencing in primary care settings for diagnosis25 or in newborn screening,26 although such genomic tools require expert guidance particularly for interpretation. Because of the rapid increase in novel curative medicines including gene therapy available for patients with rare and orphan pediatric conditions,27,28 it is important to characterize and address problems with diagnosis.
Table 4.
Results from three-level hierarchal generalized logistic model (HGLM) analysis predicting odds of leukodystrophy among a subgroup of children with complex chronic conditions (CCC) (n=439,024 children)
| Coeff. | Odds Ratio |
SE | t-ratio | p-value | |
|---|---|---|---|---|---|
| Hospital Level | |||||
| Intercept | −7.051 | −0.001 | 0.087 | −81.314 | <0.001 |
| Center of Excellence | 0.376 | 1.456 | 0.073 | 5.112 | <0.001 |
| Hospital months | 0.004 | 1.004 | 0.011 | 0.406 | 0.687 |
| Zip Code Level | |||||
| Median household Income | 0.001 | 1.000 | 0.001 | 1.206 | 0.228 |
| HPSA flag | −0.110 | 0.896 | 0.063 | −1.739 | 0.082 |
| Distance to hospital | 0.001 | 1.001 | 0.001 | 2.186 | 0.029 |
| Proportion non-White | −0.272 | 0.762 | 0.195 | −1.396 | 0.163 |
| Urban/Rural Context | |||||
| Urban core | REF | REF | REF | REF | REF |
| Large rural | 0.080 | 1.083 | 0.079 | 1.016 | 0.31 |
| Small rural | 0.223 | 1.249 | 0.094 | 2.374 | 0.018 |
| Isolated | 0.383 | 1.467 | 0.129 | 2.970 | 0.003 |
| Other urban | −0.030 | 0.971 | 0.076 | −0.391 | 0.696 |
| Patient Level | |||||
| Age at first admit in study | |||||
| period | −0.001 | 0.999 | 0.004 | −0.312 | 0.755 |
| Male (versus female) | 0.284 | 1.329 | 0.061 | 4.652 | <0.001 |
| Inpatient days (per time in PHIS) | −0.407 | 0.666 | 0.311 | −1.307 | 0.191 |
| Number of admits (per time in PHIS) | 1.164 | 3.202 | 0.182 | 6.401 | <0.001 |
| Patient Race/Ethnicity | |||||
| White, non-Hispanic | REF | REF | REF | REF | REF |
| Hispanic | 0.085 | 1.089 | 0.076 | 2.226 | 0.265 |
| Black, non-Hispanic | −0.464 | 0.629 | 0.083 | −5.595 | <0.001 |
| Asian, non-Hispanic | −0.302 | 0.739 | 0.207 | −1.459 | 0.144 |
| Other race, non-Hispanic | 0.189 | 1.201 | 0.077 | 2.440 | 0.015 |
| Patient Payer Status | |||||
| Private payer | REF | REF | REF | REF | REF |
| Government payer | 0.365 | 1.440 | 0.057 | 6.453 | <0.001 |
| Other payer | 0.415 | 1.514 | 0.154 | 2.688 | 0.007 |
ACKNOWLEDGEMENTS
We thank M. Hall and T. Richards for their assistance with data capture.
Supported by the National Institutes of Health (3UL1TR002538 [to JB., T.C., and S.G.); and the Bray Presidential Chair in Child Neurology Research (to J.B.). There was no involvement of the study sponsors in (1) study design; (2) the collection, analysis, and interpretation of data; (3) the writing of the report; and (4) the decision to submit the paper for publication. J.B.has served as a consultant to Bluebird Bio, Inc; to Calico, Inc; to Neurogene, Inc.; is on the Board of Directors of wFluidx, Inc; and owns stock in Orchard Therapeutics. The other authors declare no conflicts of interest.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Bavisetty S, Grody WW, Yazdani S. Emergence of pediatric rare diseases: Review of present policies and opportunities for improvement. Rare Dis 2013; 1:e23579. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.McCandless SE, Brunger JW, Cassidy SB. The burden of genetic disease on inpatient care in a children's hospital. Am J Hum Genet 2004; 74:121–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tambuyzer E Rare diseases, orphan drugs and their regulation: questions and misconceptions. Nat Rev Drug Discov 2010; 9:921–9. [DOI] [PubMed] [Google Scholar]
- 4.Kingsmore SF, Saunders CJ. Deep sequencing of patient genomes for disease diagnosis: when will it become routine? Sci Transl Med 2011; 3:87ps23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gahl WA, Markello TC, Toro C, Fajardo KF, Sincan M, Gill F, et al. The National Institutes of Health Undiagnosed Diseases Program: insights into rare diseases. Genet Med 2012; 14:51–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Splinter K, Adams DR, Bacino CA, Bellen HJ, Bernstein JA, Cheatle-Jarvela AM, et al. Effect of Genetic Diagnosis on Patients with Previously Undiagnosed Disease. N Engl J Med 2018; 379:2131–2139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bonkowsky JL, Wilkes J, Bardsley T, Urbik VM, Stoddard G. Association of Diagnosis of Leukodystrophy With Race and Ethnicity Among Pediatric and Adolescent Patients. JAMA Netw Open 2018; 1:e185031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gordon HB, Letsou A, Bonkowsky JL. The leukodystrophies. Semin Neurol 2014; 34:312–20. [DOI] [PubMed] [Google Scholar]
- 9.Bonkowsky JL, Nelson C, Kingston JL, Filloux FM, Mundorff MB, Srivastava R. The burden of inherited leukodystrophies in children. Neurology 2010; 75:718–725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Nelson C, Mundorff MB, Korgenski EK, Brimley CJ, Srivastava R, Bonkowsky JL. Determinants of health care use in a population-based leukodystrophy cohort. J Pediatr 2013; 162:624–628.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Helman G, Van Haren K, Escolar ML, Vanderver A. Emerging treatments for pediatric leukodystrophies. Pediatr Clin North Am 2015; 62:649–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Parikh S, Bernard G, Leventer RJ, van der Knaap MS, van Hove J, Pizzino A, et al. A clinical approach to the diagnosis of patients with leukodystrophies and genetic leukoencephelopathies. Mol Genet Metab 2015; 114:501–515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Vanderver A, Simons C, Helman G, Crawford J, Wolf NI, Bernard G, et al. Whole exome sequencing in patients with white matter abnormalities. Ann Neurol 2016; 79:1031–1037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cohen L, Manín A, Medina N, Rodríguez-Quiroga S, González-Morón D, Rosales J, et al. Argentinian clinical genomics in a leukodystrophies and genetic leukoencephalopathies cohort: Diagnostic yield in our first 9 years. Ann Hum Genet 2019; doi: 10.1111/ahg.12345. [Epub ahead of print]. [DOI] [PubMed] [Google Scholar]
- 15.Kittle K, Currier K, Dyk L, Newman K. Using a pediatric database to drive quality improvement. Semin Pediatr Surg 2002; 11:60–63. [DOI] [PubMed] [Google Scholar]
- 16.Hahs-Vaughn DL, Onwuegbuzie AJ. Estimating and Using Propensity Score Analysis With Complex Samples. J Exp Edu 2006; 75:31–65. [Google Scholar]
- 17.Raudenbush SW, Bryk AS. Hierarchical Linear Models. Applications and Data Analysis Methods (2nd ed.). 2002; Thousand Oaks, CA: Sage Publications. [Google Scholar]
- 18.Berry JG, Hall M, Cohen E, O’Neill M, Feudtner C. Ways to Identify Children with Medical Complexity and the Importance of Why. J Peds 2019; 167: 229–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Brimley CJ, Lopez J, van Haren K, Wilkes J, Sheng X, Nelson C, et al. National variation in costs and mortality for leukodystrophy patients in US children's hospitals. Pediatr Neurol 2013; 49:156–162.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Feudtner C, Feinstein JA, Zhong W, Hall M, Dai D. Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and complex medical technology dependence and transplantation. BMC Pediatr 2014; 14:199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jenkins WD, Lipka AE, Fogleman AJ, Delfino KR, Malhi RS, Hendricks B. Variance in disease risk: rural populations and genetic diversity. Genome 2016; 59:519–25. [DOI] [PubMed] [Google Scholar]
- 22.Holve S, Hu D, McCandless SE. Metachromatic leukodystrophy in the Navajo: fallout of the American-Indian wars of the nineteenth century. Am J Med Genet 2001; 101:203–8. [DOI] [PubMed] [Google Scholar]
- 23.Peyvandi S, Baer RJ, Moon-Grady AJ, Oltman SP, Chambers CD, Norton ME, et al. Socioeconomic Mediators of Racial and Ethnic Disparities in Congenital Heart Disease Outcomes: A Population-Based Study in California. J Am Heart Assoc 2018; 7:e010342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Salciccioli KB, Oluyomi A, Lupo PJ, Ermis PR, Lopez KN. A model for geographic and sociodemographic access to care disparities for adults with congenital heart disease. Congenit Heart Dis 2019; doi: 10.1111/chd.12819. [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Vassy JL, Christensen KD, Schonman EF, Blout CL, Robinson JO, Krier JB, et al. The Impact of Whole-Genome Sequencing on the Primary Care and Outcomes of Healthy Adult Patients: A Pilot Randomized Trial. Ann Intern Med 2017; 167:159–169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Holm IA, Agrawal PB, Ceyhan-Birsoy O, Christensen KD, Fayer S, Frankel LA, et al. The BabySeq project: implementing genomic sequencing in newborns. BMC Pediatr 2018; 18:225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Eichler F, Duncan C, Musolino PL, Orchard PJ, De Oliveira S, Thrasher AJ, et al. Hematopoietic Stem-Cell Gene Therapy for Cerebral Adrenoleukodystrophy. N Engl J Med 2017; 377:1630–1638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mendell JR, Al-Zaidy S, Shell R, Arnold WD, Rodino-Klapac LR, Prior TW, et al. Single-Dose Gene-Replacement Therapy for Spinal Muscular Atrophy. N Engl J Med 2017; 377:1713–1722. [DOI] [PubMed] [Google Scholar]