Medical, dental, and optical care utilization among community-living people with spinal cord injury in the United States

Abstract

Context/Objective

Information about patterns of healthcare utilization for people living with spinal cord injury (SCI) is currently limited, and this is needed to understand independent community living after SCI. This study investigates self-reported healthcare utilization among community-living people with SCI and assesses disparities across demographic, socioeconomic, and injury-related subgroups.

Design

Secondary analysis of cross-sectional survey data administered via telephone interview.

Setting

6 SCI Model Systems centers in the United States (California, Colorado, New Jersey, New York, Ohio, and Pennsylvania).

Participants

Adults with chronic, traumatic SCI who were community-living for at least one year after the completion of an inpatient rehabilitation program (N = 617).

Interventions

Not applicable.

Outcome Measures

Utilization of a usual source of 4 types of health care in the past 12 months: primary, SCI, dental, and optical.

Results

84% of participants reported utilizing primary care in the past year. More than half reported utilizing SCI (54%) and dental (57%) care, and 36% reported utilizing optical care. There were no significant differences across key subgroups in the utilization of primary care. Participants who had been injured for 5 years or less and participants with greater educational attainment were more likely to report utilizing SCI care. Participants with higher household income levels were more likely to report using dental care. Female participants and older age groups were more likely to report using optical care.

Conclusion

Rates of healthcare utilization among people with SCI are below recommended rates and vary across demographic, socioeconomic, and injury-related subgroups. This information can inform future research to target barriers to using healthcare services among community-living people with SCI.

Introduction

Routine health screenings are essential for early detection of potentially preventable illnesses. For people living with spinal cord injury (SCI), regular healthcare is also critical for mitigating the risk of secondary medical complications, which remain common contributors to rehospitalization and early mortality.^1,2 Furthermore, as the proportion of older adults living with SCI increases, routine care is also vital for management of age-related chronic conditions – including cardiovascular and metabolic conditions – and maintaining independent community living.³

Information about healthcare utilization for people with SCI is limited, but evidence suggests that this population receives services below recommended rates.^4,5 People with SCI also report high levels of unmet health needs,^6,7 and are less satisfied with their care than the general population.⁸ Reported reasons for forgoing care among people with SCI are, in part, similar to that of the general population, including personal choice and lack of knowledge about the importance of preventive services.⁴ However, people with SCI also face unique barriers to obtaining healthcare, including inaccessible exam rooms and equipment,^8,9 inaccessible transportation,¹⁰ financial barriers,¹¹ physicians unfamiliar with SCI concerns,^8,11 and short appointment times for routine visits, which may limit opportunities to address both injury-related complications and other general concerns.¹² Additionally, people with SCI benefit from regular visits with an SCI specialist trained to manage the risk of common secondary complications and provide services that general practitioners may not.^13,14 This may include evaluations of bowel and bladder health, early identification of skin breakdown, assessments for pain and spasticity, and management of autonomic dysfunction, respiratory complications, and overuse injuries.¹⁵ Limited evidence from community samples⁸ and hospital-based registries¹⁶ suggests that relatively high proportions of people with SCI in the United States seek care from SCI specialists. However, there is significant geographic variation in the availability of specialty outpatient services for people with SCI in the United States,¹⁷ and people with SCI describe the management of secondary complications as an “uphill battle”.¹⁸ More information is needed about which members of the SCI community access services from specialists trained in the management of their care.

Less is known about use of other ancillary services like dental and optical care among people living with SCI. Prior evidence on rates of regular dental care among people with SCI is scant and inconsistent.^5,19 Dental care is particularly important for people with SCI, as medications commonly taken to manage spasticity and pain can produce dry mouth, a risk factor for tooth decay,²⁰ and people who have difficulty brushing and flossing independently especially benefit from regular professional cleanings. Additionally, people living with SCI who rely on teeth to supplement impaired upper limb function have a critical need to maintain good oral health.²¹ Despite this, people with SCI may have difficulty accessing dental care. Safely assisting a patient with transfers into a dental chair may require special equipment and staff trained to meet the needs of their patients with mobility impairments,²² but most dentists in the United States are not trained in special needs dentistry.²³ Furthermore, Medicare, the primary payer for many people with SCI in the United States,²⁴ has historically not covered routine dental services for adults.

Routine optical screenings are important for not only early detection of eye disease and preventing vision loss, but also for detection of chronic conditions like high blood pressure²⁵ and diabetes.²⁶ Very limited research investigates gaps in optical care among people with physical disabilities, but findings suggest that rates of routine eye examinations are low.^4,27 Routine optical examinations typically require transfer into a hydraulic chair, and Medicare does not cover routine optical care for adults; these are likely under-explored barriers to optical care for people with SCI.

This exploratory analysis uses secondary data to extend existing knowledge of healthcare use after SCI in two ways. First, the current investigation includes dental and optical care use, as these are under-explored aspects of healthcare utilization among people with physical disabilities. Second, potential barriers to healthcare utilization include key social determinants of health, and so the aim of this analysis is to identify at-risk groups for poor healthcare utilization by analyzing who accesses care among people with SCI and examining differences across demographic, socioeconomic, and injury-related subgroups.

Materials and methods

Data

This investigation is a secondary analysis of data from the “Residential Instability in Chronic SCI” study (2017-2020), a collaborative module at six U.S. SCI Model Systems (SCIMS) centers in California, Colorado, New Jersey, New York, Ohio, and Pennsylvania. The SCIMS program has been continuously funded by what is now the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR) since 1970 to provide comprehensive SCI care and conduct innovative research to improve quality of life for people living with SCI. The national SCIMS database is one of the largest and longest-running SCI research databases in the world, and SCIMS centers must meet high benchmarks of at least 80% for enrollment and follow-up.²⁸ Further details on the SCIMS program and database have been extensively documented in prior publications.^29–32 Collaborative module studies like the Residential Instability study are multi-site demonstration projects intended to address topics outside of the core SCIMS database and are offered once per data collection cycle. Participants have the option of enrolling in these additional module studies when they are contacted to complete their routine follow-up data collection for the core SCIMS database.

The original study data were collected to investigate the effect of residential mobility on health and quality of life for people with SCI (data collection instruments available from the senior author by request). The data included a novel subset of variables on healthcare utilization intended to study the effect of residential mobility on continuity of care; these variables are the focus of this secondary analysis. Eligible participants were adults with chronic impairment (ASIA Impairment Scale A-D) who had been community-living for at least one year after completing an inpatient rehabilitation program for traumatic SCI. Participants completed one cross-sectional survey via telephone between July 2017 and October 2020. A total of 690 eligible participants completed the survey. The majority (n = 551) were recruited to complete the survey as part of their longitudinal follow-up interview for the SCIMS database. The sample also included a subset (n = 139) of people with traumatic SCI who were not enrolled in the SCIMS database. These participants were recruited via referrals and flyers to increase geographic variation in the sample. This subset of participants completed an additional questionnaire to assess key demographic, injury, and health information included in the longitudinal follow-up assessment for the SCIMS database. The present, secondary analysis of healthcare utilization uses a final analytic sample of N = 617 cases with complete data on all key variables, to aid comparability across multivariate regression models used in the analytic strategy (see Statistical Analysis).

Measures

Dependent variables

Four types of healthcare utilization were assessed: primary, SCI, dental, and optical care. Participants were asked whether 1) they had a regular source for each type of healthcare and 2) they utilized this source in the past 12 months. Participants who answered “yes” to both questions were coded as 1 = yes for each type of healthcare.

Independent variables

Injury-related covariates include injury level (tetraplegia or paraplegia) and injury type (incomplete or complete). Time since injury was measured as < 5 years or > 5 years. Wheelchair/scooter use was a binary measure comparing participants’ use of a wheelchair or scooter (vs. ambulation).

Demographic covariates were self-reported using standard survey items and include participants’ sex, race/ethnicity (non-Hispanic white, non-Hispanic Black, Hispanic, or other), marital status (married vs. not married), educational attainment (≤ high school/GED or some post-secondary education), household income (< $25,000, $25,000-$49,999, $50,000-$74,999, or $75,000 or more), and age. Participant age was recoded in 15-year intervals in accordance with the International Spinal Cord Injury Core Data Set.³³ Race and ethnicity are assessed in the SCIMS database in accordance with U.S. Census guidance.³⁴ Participants who identified as Asian or Pacific Islander, Native American, or “some other race” were collapsed into a single category; these data are included for comparison purposes only and are not interpreted due to the small sample size (n = 30) and the heterogeneous identities included in the group.

Three measures of health-related covariates were included. Health insurance compares participants with a private plan, Medicare, Medicaid, or another plan as their primary payer; participants with “other” insurance are included for comparison purposes only and are not interpreted due to the small sample size (n = 40) and the fact that the plan is unidentified. Secondary complication measures participants’ reports of either a urinary tract infection (UTI) or pressure injury in the past 12 months (vs. none). Chronic condition measures participants’ self-report of any of the following medical diagnoses in the past 12 months: diabetes, hypertension, hyperlipidemia, arthritis, osteoarthritis, rheumatoid arthritis, gout, lupus, or fibromyalgia (vs. none).

Statistical Analysis

All data analysis was conducted in Stata/SE version 17.³⁵ Descriptive and bivariate statistics (i.e. Chi-square tests, t-tests) were used to assess the distribution of key variables and compare differences across injury-related, demographic, and health-related characteristics. Multivariate logistic regression was used to estimate the association of each type of healthcare utilization, adjusted for all covariates. Likelihood ratio tests and Bayesian Information Criterion (BIC) were used to compare model fit of non-nested models in order to determine the best specified models. All models were estimated using odds ratios with 95% confidence intervals and robust standard errors. For comparability across models, all regression analyses were run on complete cases (n = 617). To account for the comparison of multiple tests, we assess statistical significance using a study-wide alpha level of p < 0.0125.³⁶ Finally, to address the potential influence of the initial months of the COVID-19 pandemic on participants’ healthcare utilization, we conducted a sensitivity analysis that replicates our analytic strategy after omitting participants interviewed during the pandemic.

Results

Table 1 presents the sample characteristics. The majority of participants had been injured for > 5 years at the time of interview (79.6%), and most used a wheelchair or scooter as their primary means of mobility (77%). Consistent with the national SCIMS database,²⁴ approximately half were classified as having an incomplete injury (54.1%), though a greater proportion than the national database were classified as people with paraplegia than tetraplegia (55.6%). Also consistent with national estimates, the majority were male (78.1%) and non-Hispanic white (61.3%). At the time of interview, 39.2% reported being married, and approximately half (51.1%) had attained at least some post-secondary education. More than one third (34.9%) lived in a household with an income of less than $25,000. Approximately two-thirds of participants were insured by either a private health insurance plan (36%) or a Medicare plan (31%), with an additional 26% covered by a Medicaid plan. Most participants reported at least one secondary medical complication in the past year (61.4%), whereas 43.8% reported a diagnosis of at least one chronic condition.

Table 1.

Sample Characteristics (n & %) (N = 617).

	n	%
Injury-related covariates
Injury level
paraplegia	343	55.6
tetraplegia	274	44.4
Injury type
incomplete	334	54.1
complete	283	45.9
Injured > 5 yrs. (vs. ≤ 5 yrs.)	491	79.6
Wheelchair/scooter user (vs. ambulatory)	475	77.0
Demographic covariates
Age group
18–34 yrs.	131	21.2
35–44 yrs.	120	19.5
45–54 yrs.	155	25.1
55–64 yrs.	138	22.4
65 + yrs.	73	11.8
Male (vs. female)	482	78.1
Race/ethnicity
Non-Hispanic White	378	61.3
Non-Hispanic Black	125	20.3
Hispanic	84	13.6
Other	30	4.9
Married (vs. never or previously married)	242	39.2
Socioeconomic status covariates
Educational attainment
HS/GED or less	302	48.9
Some college or more	315	51.1
Household income
<$25,000	215	34.9
$25,000-$49,999	139	22.5
$50,000-$74,999	86	13.9
$75,000+	177	28.7
Health-related covariates
Primary health insurance
private	223	36.1
Medicare	193	31.3
Medicaid	161	26.1
other	40	6.5
Secondary complication in past 12 mos. (vs. none)	379	61.4
Chronic health condition (vs. none)	270	43.8

Figure 1 presents the reported rates of healthcare utilization in the past 12 months. The majority of participants reported utilizing a regular source for primary care in the past 12 months (84.4%). Approximately half reported utilizing a regular source for SCI care (53.7%) and dental care (57.1%). Approximately one-third (36%) of participants reported utilizing a regular source for optical care in the past 12 months.

Figure 1.

Rates of Reported Healthcare Utilization in the Past 12 Months (%) (N = 617).

Table 2 presents logistic regression models estimating the odds of utilizing primary and SCI care, adjusted for all covariates. In the model predicting the odds of primary care, there were no differences observed by demographic characteristics, injury-related characteristics, or reports of secondary complications. Participants with a chronic condition were more likely to report utilizing a regular source for primary care in the past 12 months (OR = 2.28, P = 0.004, 95% CI =  1.30-3.98). In the model predicting odds of SCI care, participants who had been injured for more than 5 years had 69% lower odds of utilizing SCI care compared to participants who had been injured for 5 years or less (OR = 0.31, P = 0.00, 95% CI =  0.19-0.49), whereas participants who use a wheelchair or scooter had more than double the odds of utilizing SCI care compared to ambulatory participants (OR = 2.38, P = 0.00, 95% CI = 1.47-3.85). Participants with at least some college education had 77% greater odds of utilizing SCI care than participants with a high school diploma/GED or less (OR =  1.77, P = 0.005, 95% CI = 1.19-2.63).

Table 2.

Logistic Regression: Associations of Primary Care and SCI Care Utilization with All Covariates (N = 617).

Primary Care					SCI Care
	OR	p	95% CI			OR	p	95% CI
Paraplegia (ref: tetraplegia)	0.86	0.559	0.51	1.44	Paraplegia (ref: tetraplegia)	0.73	0.100	0.50	1.06
Complete SCI (ref: incomplete)	0.78	0.346	0.46	1.31	Complete SCI (ref: incomplete)	0.95	0.797	0.64	1.42
Injured > 5 yrs. (vs. ≤ 5 yrs.)	0.83	0.559	0.45	1.53	Injured > 5 yrs. (vs. ≤ 5 yrs.)	0.31	0.000	0.19	0.49
Wheelchair/scooter user (ref: ambulatory)	1.02	0.940	0.55	1.92	Wheelchair/scooter user (ref: ambulatory)	2.38	0.000	1.47	3.85
Age group (ref: 18–34 yrs.)					Age group (ref: 18–34 yrs.)
35-44yrs	0.84	0.631	0.41	1.71	35-44yrs	0.99	0.999	0.58	1.73
45-54yrs	0.57	0.106	0.28	1.13	45-54yrs	0.84	0.509	0.49	1.42
55-64yrs	0.92	0.835	0.40	2.10	55-64yrs	0.58	0.071	0.32	1.05
65 + yrs	1.12	0.826	0.40	3.17	65 + yrs	0.71	0.339	0.34	1.44
Male (ref: female)	1.02	0.954	0.57	1.80	Male (ref: female)	1.25	0.300	0.82	1.92
Race/ethnicity (ref: Non-Hispanic White)					Race/ethnicity (ref: Non-Hispanic White)
Non-Hispanic Black	0.96	0.910	0.49	1.89	Non-Hispanic Black	1.84	0.015	1.13	3.01
Hispanic	1.28	0.541	0.58	2.81	Hispanic	1.64	0.100	0.91	2.95
Other	0.83	0.733	0.28	2.43	Other	1.60	0.281	0.68	3.78
Married (ref: never or previously married)	1.58	0.083	0.94	2.63	Married (ref: never or previously married)	0.75	0.164	0.49	1.13
Some college or more (ref: HS/GED or less)	1.60	0.082	0.94	2.73	Some college or more (ref: HS/GED or less)	1.77	0.005	1.19	2.63
Household income (ref: < $25,000)					Household income (ref: < $25,000)
$25,000-$49,999	0.64	0.165	0.34	1.20	$25,000-$49,999	1.16	0.554	0.71	1.88
$50,000-$74,999	0.76	0.498	0.35	1.67	$50,000-$74,999	1.74	0.081	0.93	3.24
$75,000+	0.67	0.270	0.33	1.36	$75,000+	1.34	0.311	0.76	2.33
Primary health insurance (ref: private)					Primary health insurance (ref: private)
Medicare	1.35	0.311	0.75	2.42	Medicare	1.21	0.406	0.77	1.90
Medicaid	1.54	0.234	0.76	3.14	Medicaid	1.15	0.612	0.67	1.98
Other	1.23	0.673	0.47	3.26	Other	4.51	0.001	1.86	10.92
Secondary complication in past 12 mos.	1.85	0.014	1.13	3.01	Secondary complication in past 12 mos.	1.16	0.436	0.79	1.71
Chronic health condition	2.28	0.004	1.30	3.98	Chronic health condition	1.36	0.120	0.92	2.01
Constant	3.32	0.05	0.99	11.08	Constant	0.73	0.470	0.31	1.72

Notes: OR = odds ratio; P = p-value; CI = confidence interval

Table 3 presents the fully adjusted logistic regression models predicting odds of dental and optical care utilization. In the model predicting odds of dental care, participants with household incomes of $75,000 or more had triple the odds of utilizing dental care compared to those with incomes of less than $25,000 (OR = 3.22, P = 0.00, 95% CI = 1.83-5.67). In the model predicting odds of optical care, male participants had approximately half the odds of utilizing optical care of female participants (OR = 0.52, P = 0.00, 95% CI = 0.34-0.81). We also find higher odds of optical care utilization for older age groups; specifically, participants who were 55–64 years old (OR = 2.84, P = 0.00, 95% CI = 1.51-5.35) or 65 + years (OR = 4.50, P = 0.00, 95% CI = 2.11-9.61) had significantly greater odds of optical care utilization compared to those 18–34 years old.

Table 3.

Logistic Regression: Associations of Dental Care and Optical Care Utilization with All Covariates (N = 617).

Dental Care					Optical Care
	OR	p	95% CI			OR	p	95% CI
Paraplegia (ref: tetraplegia)	0.84	0.374	0.58	1.23	Paraplegia (ref: tetraplegia)	0.76	0.177	0.51	1.13
Complete SCI (ref: incomplete)	1.08	0.701	0.73	1.60	Complete SCI (ref: incomplete)	0.99	0.998	0.65	1.53
Injured > 5 yrs. (vs. ≤ 5 yrs.)	0.96	0.884	0.62	1.51	Injured > 5 yrs. (vs. ≤ 5 yrs.)	1.07	0.782	0.66	1.73
Wheelchair/scooter user (ref: ambulatory)	0.95	0.836	0.59	1.53	Wheelchair/scooter user (ref: ambulatory)	0.98	0.950	0.60	1.62
Age group (ref: 18–34 yrs.)					Age group (ref: 18–34 yrs.)
35-44yrs	1.22	0.464	0.71	2.10	35-44yrs	1.03	0.934	0.54	1.96
45-54yrs	1.08	0.762	0.64	1.84	45-54yrs	2.13	0.013	1.17	3.87
55-64yrs	1.19	0.555	0.67	2.13	55-64yrs	2.84	0.001	1.51	5.35
65 + yrs	1.63	0.189	0.79	3.37	65 + yrs	4.50	0.000	2.11	9.61
Male (ref: female)	0.73	0.163	0.48	1.13	Male (ref: female)	0.52	0.004	0.34	0.81
Race/ethnicity (ref: Non-Hispanic White)					Race/ethnicity (ref: Non-Hispanic White)
Non-Hispanic Black	0.88	0.592	0.55	1.41	Non-Hispanic Black	0.77	0.326	0.45	1.30
Hispanic	0.81	0.465	0.46	1.42	Hispanic	0.81	0.530	0.41	1.58
Other	0.30	0.006	0.13	0.71	Other	0.43	0.106	0.16	1.20
Married (ref: never or previously married)	0.80	0.282	0.53	1.21	Married (ref: never or previously married)	1.38	0.132	0.91	2.10
Some college or more (ref: HS/GED or less)	1.36	0.115	0.93	1.99	Some college or more (ref: HS/GED or less)	1.31	0.199	0.87	1.98
Household income (ref: < $25,000)					Household income (ref: < $25,000)
$25,000-$49,999	1.02	0.949	0.64	1.61	$25,000-$49,999	0.92	0.749	0.53	1.57
$50,000-$74,999	1.38	0.269	0.78	2.47	$50,000-$74,999	1.39	0.309	0.74	2.61
$75,000+	3.22	0.000	1.83	5.67	$75,000+	1.35	0.313	0.75	2.42
Primary health insurance (ref: private)					Primary health insurance (ref: private)
Medicare	0.74	0.188	0.47	1.16	Medicare	0.65	0.074	0.41	1.04
Medicaid	0.81	0.444	0.48	1.38	Medicaid	0.59	0.087	0.32	1.08
Other	1.14	0.734	0.53	2.45	Other	1.82	0.125	0.85	3.89
Secondary complication in past 12 mos.	0.76	0.162	0.52	1.12	Secondary complication in past 12 mos.	1.34	0.161	0.89	2.02
Chronic health condition	1.22	0.323	0.82	1.79	Chronic health condition	1.28	0.231	0.85	1.93
Constant	1.54	0.322	0.65	3.62	Constant	0.37	0.032	0.15	0.92

Notes: OR = odds ratio; P = p-value; CI = confidence interval

Sensitivity Analysis

The COVID-19 pandemic greatly disrupted the healthcare sector in the United States, including access to and utilization of routine outpatient visits. To address the potential effect of the COVID-19 pandemic on participants’ healthcare utilization, we replicated the analysis after removing participants (n = 101) who completed the interview on or after March 1, 2020 (results not tabled). This date was selected to coincide with the initial months of stay-at-home orders introduced to slow community transmission of COVID-19 in the United States. Compared to participants interviewed prior to the pandemic, post-pandemic participants included significantly higher proportions of non-Hispanic whites and those in the highest household income quartiles (i.e. $50,000-$74,999 and $75,000 or more). Rates of reported healthcare utilization were not significantly different for participants interviewed pre- and post-pandemic. In the analysis of primary and dental care utilization, the associations observed in the full sample were unchanged. In the analysis of SCI care utilization, the observed associations for years injured and mobility were unchanged, but the association with educational attainment was no longer significant at the P = 0.0125 level. In the analysis of optical care utilization, the associations of age group were unchanged, but the association with sex was no longer significant at the P = 0.0125 level.

Discussion

This analysis uses a sample of people living with chronic SCI to investigate patterns of healthcare utilization and variation across injury-related and demographic subgroups. We find that the majority of participants reported relatively high levels of care utilization overall, but that rates of some types of healthcare were below recommended rates, and there was some evidence of socioeconomic disparities across groups.

In terms of primary care, 15% of participants reported that they did not seek primary care in the past year. This is consistent with estimates from the general population, in which 82-85% of adults report having seen a doctor in the past year.³⁷ While there are currently no national recommendations for annual physical examinations among people living with SCI, it is widely recognized that people with SCI may require more frequent and lengthier primary care visits than the general population to manage comorbidities and make referrals for specialist care as needed.^13,15

We find that 54% of participants reported seeing an SCI care provider in the past year, and that this was significantly more likely among those who had been injured for 5 years or less. We also find that people with at least some post-secondary education were significantly more likely to report having utilized SCI care than those with a high school diploma or GED. These differences persist despite controlling for health insurance and household income. This analysis is not able to disentangle whether participants utilized SCI care instead of, or in addition to, primary care. Prior reports suggest that while many people with SCI use a primary care physician as their principal medical provider,^38,39 many are dissatisfied with their provider’s knowledge of SCI concerns.¹⁶ By contrast, those who do see an SCI specialist are highly satisfied with their care.^10,16 Medical providers who are knowledgeable about SCI are necessary to maximize opportunities for independent living,³ and as such, more information about access to SCI specialist care and how people with SCI make decisions about the type of care provider they choose for their medical needs is needed.

Findings about dental care show that about 57% of participants reported accessing dental care in the past 12 months. This is lower than assessments from the general population, which estimate 65% of U.S. adults saw a dentist in the past 12 months.⁴⁰ Additionally, we find some evidence of socioeconomic disparities, in that participants with a household income of more than $75,000 were significantly more likely to report utilizing dental care than those with incomes of less than $25,000, which is consistent with disparities in the general population. The American Dental Association currently recommends that the frequency of routine dental cleanings be determined by the dental care provider based on patient needs and risk factors.⁴¹ Recent evidence suggests that annual cleanings may be appropriate for patients with low risk of periodontal disease, and that individuals with one or more risk factors may require multiple cleanings per year.^42,43 In the general U.S. population, low socioeconomic groups are more likely to have unmet dental needs due to financial and health insurance barriers to accessing dental care.⁴⁴ Given that many people with SCI are at risk for financial hardship,^45,46 as well as the fact that some people with SCI are at higher risk for periodontal disease, more information about barriers to dental care for people with SCI should be explored in order to assess the prevalence of unmet needs in this population.

We find that most of our sample had not sought optical care in the past year, with 36% reporting utilization of optical care in the past 12 months, which is lower than estimates in the general population of 55%.⁴⁷ The American Academy of Ophthalmology recommends routine eye examinations for individuals without optical problems or risk of eye disease based on age, specifically 1–3 years for people 55–64 years and every 1–2 years for people 65 or older.⁴⁸ This is consistent with our finding that the odds of optical care utilization were significantly higher for participants 55 years of age or older. Additionally, we find that male participants had lower odds of accessing optical care than female participants, which is consistent with estimates from the general population.⁴⁹ Vision screening guidelines are patient- and needs-specific, and this analysis was not able to assess how patient risk factors might explain the relatively low prevalence of optical care utilization reported. Future research should explore whether the lower prevalence of eye exams can be explained by patient needs or by barriers to accessing care that have been reported in other groups of people with physical disabilities.²⁷

Finally, while data collection for the parent study overlapped with the initial 8 months of the COVID-19 pandemic in the United States, we find that rates of healthcare utilization were not significantly different for participants interviewed prior to and after the onset of the pandemic. However, the sensitivity analysis also found that the post-pandemic participant sample included fewer racial/ethnic minorities and people in the lower household income quartiles. In the general population, emerging evidence suggests that racial/ethnic minorities had lower rates of routine, outpatient care visits during the initial months of the pandemic.^50,51 The lack of diversity in the post-pandemic sample may have precluded us from capturing similar disparities in the SCI population. We do find other significant differences in the rates of healthcare utilization across demographic groups when comparing participants interviewed before and after the start of the pandemic. Specifically, the difference in the odds of accessing SCI care between participants with and without post-secondary education was no longer significant when only pre-pandemic participants were included in the model, though the overall educational attainment of pre-pandemic and post-pandemic participants was not significantly different. Additionally, we find no significant differences in the odds of accessing optical care by sex when only pre-pandemic participants are included. Some groups of people with disabilities, including people with physical disabilities, were disproportionately vulnerable to negative effects resulting from efforts to curtail the pandemic, such as closures of community programs and services.⁵² Given that some of the findings of this analysis were sensitive to the COVID-19 pandemic, future research is needed to understand how the COVID-19 pandemic may have affected healthcare access for certain individuals with SCI.

Limitations

This investigation has several limitations. As a secondary analysis of existing data, the study was limited to available measures in the source data. The objective of the parent study was to understand post-injury residential mobility among people with SCI and the effect of moving on their health and quality of life. A subset of variables on healthcare utilization was included to assess whether moving was disruptive to participants’ continuity of care, but a full assessment of participants’ healthcare utilization patterns was beyond the scope of the study. Therefore, in the present analysis, reasons why participants did or did not utilize healthcare are not included, nor are telehealth services or the full range of medical services that may be relevant to patients with SCI (e.g. pain management). Additionally, a broad range of rehabilitation professionals may provide specialty outpatient services for people with SCI, and participants were not asked to further specify the type of SCI specialist they saw. The analytic sample was obtained from 6 regional SCIMS centers and may not be representative of all people with SCI living in the United States. While recent research suggests that the SCIMS database has high levels of demographic comparability with the U.S. population with incident SCI,⁵³ rates of reported SCI care utilization in particular may be overestimated given that the majority of participants in the sample are patients affiliated with a SCIMS center, which are recognized centers of excellence for SCI care. This may limit the study’s ability to detect healthcare use disparities that may be evident in the general population, particularly for historically vulnerable groups who may not have access to high-quality care like SCIMS rehabilitation facilities. We also use a stringent alpha level of P = 0.0125 to reduce the likelihood of a Type I error across multiple tests in the analysis. However, this approach also increases the likelihood of making a Type II error. Some patterns in the data were significant at the P = 0.05 level, which is a conventional threshold for statistical significance in public health and psychosocial research. Future research with larger samples should further explore these patterns to reduce the likelihood of a Type II error in understanding patterns of healthcare utilization after SCI. Finally, we cannot eliminate the possibility of recall bias or social desirability bias in our interpretation of the results, which are risks of self-report measures for which there is a socially favorable response, including healthy behaviors.^54,55 The original study data were collected as self-report measures via telephone interview, and some participants may have inaccurately reported higher levels of healthcare utilization.

Conclusion

Information on healthcare utilization for people with SCI is limited, especially for ancillary services like dental and optical care. We find lower than recommended rates of SCI and dental care use among community-living people with SCI, as well as some evidence of socioeconomic disparities in care utilization. Understanding patterns of healthcare utilization for people with SCI living in the community can inform future research to target modifiable barriers to care access, which in turn can promote early intervention for chronic conditions and prevent secondary medical complications and rehospitalizations in this population.

Funding Statement

This work was supported by the Craig H. Neilsen Foundation under the Psychosocial Research (PSR) Postdoctoral Fellowship [#639798]; and the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR) under the Northern New Jersey Spinal Cord Injury System [#90SI5026-01-00]. NIDILRR is a center within the Administration for Community Living (ACL), United States Department of Health and Human Services. The contents of this manuscript do not necessarily represent the policy of NIDILRR, ACL, or HHS, and you should not assume endorsement by the federal government.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Contributors

All authors contributed significantly to the study design, data analysis, and/or writing of this manuscript. LM designed the objectives for the present analysis, conducted all data analysis, interpreted the results, and wrote the manuscript. AB designed and implemented the original study, interpreted the results of the present analysis, and wrote the manuscript. TB, JC, MS, MJR, and LW assisted with the design and implementation of the original study, assisted with the interpretation of the results of the present analysis, and provided feedback on the final manuscript.