Abstract
PURPOSE:
Existing evidence identifies racial and ethnic disparities impacting the prevalence and severity of cerebral palsy (CP). There is a paucity of literature examining the impact on associated treatment.
METHODS:
In this retrospective cohort study, an institutional database search identified outpatient encounters for pediatric patients with spastic CP. Additional filters were used to determine treatments received. For each treatment, the proportion of African American (AA) patients receiving treatment was compared to the proportion of Caucasian (C) patients receiving the same treatment.
RESULTS:
3,686 children with spastic CP were seen in outpatient clinics associated with an academic tertiary hospital over a 21-year period. There was no significant difference between the proportion of any treatment compared to the entire sample for AA or C patients.
CONCLUSION:
In this sample, there was no significant evidence of a racial disparity for AA patients receiving treatments for spasticity. This data is limited by several factors. Further research is needed to determine whether pediatric patients with disabilities are receiving equitable care. Clinicians should consider systematically monitoring their practices to identify areas of bias or inequity in accessing care.
Introduction
A recent issue of the Journal of Pediatric Rehabilitation Medicine focused on health disparities and access limitations for patients with cerebral palsy (CP). In it, the editors called for further exploration of these factors to improve the care and long-term management of patients with CP [1]. Much of the existing literature has focused on risk factors for CP and for severity of motor impairment and comorbidities [1–3]. Further, implicit bias may impact healthcare providers’ practices and associated outcomes for people of color [4]. Disparities in health care have been identified in Black physical medicine and rehabilitation (PM&R) patients [5], though to date this has not been described for pediatric patients with CP. Racial disparities have been described with regard to quality of life [6] and unmet therapy needs [7] in children with special needs. Additionally, racial health inequities in children are well documented with regard to health care utilization such as emergency department visits or hospitalizations [8]. To date, there has been little evidence describing patterns of utilization and acceptance of treatments for CP symptoms such as spasticity. In an effort to assess how race impacted treatment patterns in a single practice, the primary objective of this study was to test if African American (AA) patients with CP were more or less likely to receive spasticity treatment as compared to Caucasian (C) patients with CP.
Materials and methods
An institutional database search identified all outpatient encounters for patients 0–17 years old with International Classification of Diseases 9 and 10 codes for spastic CP over a 21-year period. Other types of CP, such as ataxic or choreoathetoid CP, were not included. Additional filters were used to determine spasticity treatments utilized prior to turning 18 years old, including enteral baclofen, intrathecal baclofen (ITB), selective dorsal rhizotomy (SDR), phenol, and botulinum toxins. Race/ethnicity information in this database was obtained from the electronic medical record demographic section and categorized as AA, C, Asian, other, or unknown/refused. The type of data available for analysis was at the cohort-level and not patient-level.
To address the primary objective, unadjusted odds ratios (ORs) with 95% confidence intervals (CIs) were estimated using logistic regression comparing AA to C patients for each treatment. As this study did not have access to patient-level data, models were unable to be adjusted by covariates (e.g., age).
Statement of informed consent and human/animal rights
Institutional Review Board exemption was granted. Individual informed consent was not required, as all data was viewed in aggregate and there was no access to individuals’ protected health information.
Results
3,686 children with spastic CP, 0-17 years old, were seen in an outpatient clinic associated with an academic quaternary care center from 1/1/2000 - 12/31/2020. Sample characteristics are described in Table 1. C and AA comprised the two largest racial groups within the sample, 76.5% and 13.4% respectively. Asian patients made up less than 3% of the sample population, a sample size too small from which to draw meaningful results.
Sample characteristics
Sample characteristics
The unadjusted OR for each treatment comparing AA to C patients is presented in Table 2. Compared to C patients, the ORs of baclofen and ITB for AA patients were 1.21 (95% CI = 0.98–1.48) and 0.62 (95% CI = 0.35–1.08), respectively.
Results
Results of the unadjusted logistic regression to estimate the odds ratio with 95% confidence intervals (CI) for each treatment comparing African American to Caucasian children with spastic cerebral palsy.
In this sample, there was no clear evidence of a racial disparity for AA patients receiving common treatments for spasticity. The percentage of C (76.5%) and AA (13.4%) patients was similar to White and Black representation in 2019 census data for the state of Michigan, with 79.2% identifying as White and 14.1% Black, as well as for the United States with 76.3% White and 13.4% Black [9]. This data is limited for several reasons. Notably, there was access to only cohort-level data. This data does not inform whether treatment plans were initiated at similar ages or reassessed with similar frequency in individuals, nor does it account for any difference in severity of spasticity or functional impairment between groups. This data includes only treatments prescribed and procedures completed within the study period, not those that were offered but declined or deferred to a date beyond the study period. This information was not discernible within the database. Given that minorities are more likely to have more severe motor impairment [2, 3], one might infer that more treatment utilization would be expected if employed equitably. Factors such as insurance coverage and socioeconomic status were not included in analysis, nor was individual follow-up interval or frequency. Given that the entire cohort had been evaluated in an academic subspecialty clinic, this group likely represents a portion of the population with greater access to care. Regional variation and individual physician factors could not be elucidated from this single center work. Similarly, the single center design limits generalization to other settings or clinics, or the broader US population with CP. Other races were not included in the study due to sample size limiting interpretation. Finally, racial complexities may not be accurately described within the constructs of databases.
Language is important for obtaining accurate information about race and ethnicity. The database utilized for this study included the word “Caucasian,” which may not be equivalent to the National Institute of Health’s recommended categorization of “White.” The word Caucasian was used here to maintain consistency with the database.
PM&R professionals should consider systematically monitoring their clinical practices to identify areas of bias or inequity in accessing care, with the aim of implementing quality improvement initiatives to address those inequities. Individual practices may utilize varying review periods and available patient-level data to better understand trends within their own care. Larger datasets may be utilized to study similar trends within the US and internationally. Further research is needed to determine whether pediatric patients with disabilities of all races and ethnicities are receiving equitable rehabilitation care.
Footnotes
Acknowledgments
The authors have no acknowledgements.
Ethical considerations
Institutional Review Board exemption was granted. Individual informed consent was not required, as all data was viewed in aggregate and there was no access to individuals’ protected health information.
Conflicts of interest
The authors have no conflicts of interest to report.
Funding
No outside funding sources were required for this research.