INTRODUCTION
American Indian and Alaska Native (AIAN) communities are under-resourced due to lingering effects of colonialization and other structural barriers to optimal health.1 AIAN individuals are more likely than others to live in poverty, have lower educational attainment, and lack broadband internet.1 These and other factors contribute to inequities in health care access and outcomes among AIAN communities.2,3
During COVID-19, outpatient care shifted toward telemedicine driven in part by stay-at-home orders and COVID-19 infection risk. To support access to care, states implemented policies to broaden permissible telemedicine platforms4 and remove geographic restrictions.5 While these changes were intended to increase access for all individuals, pre-existing access and digital barriers may have prevented AIAN communities from taking up telemedicine.
Unfortunately, there is limited evidence about this issue—and subsequently, limited insight for targeted policy efforts to improve access among AIAN communities. To address this evidence gap, this study examined patient and area-level factors associated with telemedicine use among AIAN individuals in Washington State.
METHODS
This retrospective study analyzed Washington Medicaid claims data from 2019 to 2021. We identified 126,547 adults enrolled in Medicaid for ≥ 11 months in any year who self-reported as AIAN.
The primary outcome was a binary measure of telemedicine use, defined as outpatient visits delivered through real time audio-only or audio–video telemedicine, in a year. Visits were identified using Current Procedural Terminology codes, code modifiers, or place of service codes.6
We applied multivariable logistic regression, with heteroskedastic robust standard errors and separate models for 2020 and 2021, to quantify patient and area-level factors associated with telemedicine use. Inclusion of factors was informed by the Andersen Behavioral Model. A p-value of 0.05 designated statistical significance. This study was approved by the University of Washington IRB.
RESULTS
In 2020, 41,325 AIAN individuals were enrolled in Washington State Medicaid (Table 1). On average, individuals were 39.8 years of age (SD = 15.2), and most were female (55.5%) and non-Hispanic (88.1%) and resided in urban areas (75.4%). Most individuals (73.9%) lived in areas of highest socioeconomic disadvantage, measured by the Area Deprivation Index (ADI). Mean outpatient visits was 2.6 (SD = 4.2). Characteristics were similar in 2021.
Table 1.
Descriptive Statistics of AI/AN Individuals Insured by Medicaid, 2020–2021
| Variable | 2020 N = 41,325 |
2021 N = 49,126 |
|---|---|---|
| Insured by a managed care organization, n (%) | 17,576 (42.53) | 22,080 (44.95) |
| Age, mean ± SD | 39.83 ± 15.24 | 39.48 ± 15.15 |
| Female, n (%) | 22,945 (55.52) | 27,136 (55.24) |
| Hispanic, n (%) | 4670 (11.9) | 5885 (12.7) |
| English as preferred spoken language, n (%) | 40,692 (99.7) | 48,329 (99.7) |
| Area Deprivation Index (ADI), mean ± SD | 8.41 ± 1.44 | 8.40 ± 1.45 |
| ADI quintiles1 | ||
| 1/2 | 689 (1.7) | 833 (1.73) |
| 2/3 | 412 (1.02) | 492 (1.02) |
| 5/6 | 2594 (6.4) | 3131 (6.50) |
| 7/8 | 6885 (17.0) | 8343 (17.31) |
| 9/10 | 29,928 (73.88) | 35,385 (73.44) |
| Homeless, n (%) | 4813 (11.7) | 5594 (11.6) |
| Federal poverty level, mean ± SD | 32.66 ± 48.63 | 35.68 ± 51.20 |
| Weighted Charlson score, mean ± SD | 0.32 ± 0.86 | 0.3 ± 0.83 |
| Comorbidities | ||
| Acute myocardial infarction | 196 (0.47) | 243 (0.49) |
| Cancer | 520 (1.26) | 591 (1.2) |
| Cerebrovascular disease | 525 (1.27) | 673 (1.37) |
| Chronic pulmonary disease | 2529 (6.12) | 2432 (4.95) |
| Congestive heart failure | 684 (1.66) | 827 (1.68) |
| Dementia | 121 (0.29) | 113 (0.23) |
| Diabetes mellitus with complications | 982 (2.38) | 1137 (2.31) |
| Diabetes mellitus without complications | 2679 (6.48) | 3146 (6.40) |
| Hemiplegia paraplegia | 142 (0.34) | 167 (0.34) |
| Mild liver disease | 974 (2.36) | 1080 (2.2) |
| Peptic ulcer | 128 (0.31) | 124 (0.25) |
| Peripheral vascular disease | 280 (0.68) | 372 (0.76) |
| Rheumatic disease | 513 (1.24) | 543 (1.11) |
| Moderate/severe liver disease | 125 (0.3) | 137 (0.28) |
| Area level factors2 | ||
| % households with broadband access, mean ± SD | 88.30 ± 4.31 | - |
| % Households with cellular plan, mean ± SD | 78.4 ± 7.19 | - |
| Primary care shortage area, n (%) | ||
| Whole county has shortage | 3059 (7.4) | |
| Part of county has shortage | 38,263 (92.6) | |
| Per capita income, mean ± SD | 36,447 ± 9549.15 | - |
| % households with no vehicle, mean ± SD | 6.30 ± 2.17 | - |
| Rural residence, n (%) | 10,180 (24.63) | 11,822 (24.06) |
| Health care utilization | ||
| # of outpatient visits, mean ± SD | 2.62 ± 4.18 | 2.57 ± 4.33 |
| # of telemedicine visits, mean ± SD | 1.03 ± 2.77 | 0.92 ± 2.62 |
| Telemedicine visit, n (%) | 12,934 (31.3) | 13,870 (28.23) |
1ADI quintiles capture levels of neighborhood socioeconomic disadvantage as measured by the Area Deprivation Index. Categories measure the relative ranking of an individuals’ residence Census block within the state of Washington
2Area level factors were unavailable for calendar year 2021
In 2020, 31.3% of AIAN individuals had ≥ 1 telemedicine visit. This utilization rate decreased to 28.2% in 2021.
In adjusted analysis of 2020 data, lower age, female gender, preference for spoken English, urban residence, and enrollment in Medicaid-managed care were associated with greater odds of telemedicine use (Table 2). Most comorbidities were associated with telemedicine use, with rheumatic disease, diabetes, and chronic pulmonary disease exhibiting the strongest associations. Among area-level factors, higher ADI, residence in primary care shortage area, and percent of households with broadband internet were associated with lower odds of telemedicine use. Adjusted results were similar in 2021.
Table 2.
Adjusted Results of Factors Predictive of Telehealth Use Among AIAN Medicaid individuals
| Variable | 2020 | 2021 | ||
|---|---|---|---|---|
| OR (95% CI) | p-value | OR (95% CI) | p-value | |
| Enrolled in managed care plan | 1.61 (1.54,1.69) | < 0.01 | 1.50 (1.44,1.57) | < 0.01 |
| Age | 0.99 (0.99,0.99) | < 0.01 | 0.99 (0.99,0.99) | < 0.01 |
| Female | 1.65 (1.58,1.73) | < 0.01 | 1.58 (1.51,1.65) | < 0.01 |
| English as preferred spoken language | 2.05 (1.19,3.56) | 0.01 | 1.71 (1.06,2.76) | 0.03 |
| Homeless | 0.88 (0.82,0.95) | < 0.01 | 0.84 (0.78,0.90) | < 0.01 |
| Comorbidities1 | ||||
| Myocardial infarction | 0.85 (0.61,1.18) | 0.33 | 1.23 (0.92,1.64) | 0.16 |
| Cancer | 2.17 (1.80,2.63) | < 0.01 | 2.24 (1.88,2.67) | < 0.01 |
| Cerebrovascular disease | 1.60 (1.32,1.94) | < 0.01 | 1.58 (1.33,1.87) | < 0.01 |
| Congestive heart failure | 2.00 (1.68,2.38) | < 0.01 | 1.36 (1.16,1.60) | < 0.01 |
| Chronic pulmonary disease | 2.61 (2.39,2.85) | < 0.01 | 2.43 (2.23,2.66) | < 0.01 |
| Dementia | 0.81 (0.53,1.23) | 0.32 | 0.78 (0.50,1.22) | 0.27 |
| Diabetes | 3.00 (2.76,3.26) | < 0.01 | 2.52 (2.34,2.72) | < 0.01 |
| Hemiplegia | 2.04 (1.42,2.93) | < 0.01 | 1.48 (1.06,2.06) | 0.02 |
| Mild liver disease | 2.43 (2.11,2.81) | < 0.01 | 2.62 (2.28,3.00) | < 0.01 |
| Peptic ulcer | 1.65 (1.12,2.43) | 0.01 | 1.52 (1.03,2.24) | 0.04 |
| Peripheral vascular disease | 1.30 (0.99,1.70) | 0.06 | 1.58 (1.26,1.97) | < 0.01 |
| Rheumatic disease | 4.13 (3.38,5.03) | < 0.01 | 2.97 (2.47,3.57) | < 0.01 |
| Liver disease | 1.25 (0.83,1.88) | 0.29 | 0.87 (0.59,1.27) | 0.46 |
| Area-level factors | ||||
| ADI above median | 0.86 (0.81,0.91) | < 0.01 | 0.85 (0.81,0.89) | < 0.01 |
|
% households with broadband access Above median vs. below |
0.85 (0.81,0.90) | < 0.01 | - | - |
| % households with no vehicle for workers age > = 16 | 0.97 (0.96,0.99) | < 0.01 | - | - |
|
Residence in primary care shortage area Whole county vs. > = 1 part |
0.75 (0.69,0.83) | < 0.01 | - | - |
| Urban residence | 1.50 (1.42,1.60) | < 0.01 | 1.21 (1.15,2.27) | < 0.01 |
|
Federal poverty level Zero % vs. all other |
0.90 (0.86,0.94) | < 0.01 | 0.90 (0.86,0.94) | < 0.01 |
1Comorbidities reflect indicators that denote the presence of a given physical or behavioral health condition
DISCUSSION
Utilization of telemedicine among AIAN individuals in Washington declined between April 2020 and December 2021 likely reflecting attenuation of mobility restrictions over time. Telemedicine use rates were similar to those among some racial and ethnic subgroups.7 Use varied by several individual- and area-level factors. Notably, associations between measures of socioeconomic advantage suggest the presence of structural barriers that may reduce the effectiveness of policies to increase access to telehealth use for rural AIAN communities. More broadly, associations identified in this study can inform target areas for future policy development.
Addressing these barriers through policy and practice change is critical because telemedicine has potential to benefit AIAN communities, including via reducing patient travel and childcare expenses, and improving access to subspecialty care. Increased access to fixed broadband internet and culturally humble telemedicine practices could obviate some barriers to effective telemedicine use for rural AIANs. While policy changes were implemented at both the state and federal level during the public health emergency, further efforts to implement these elements may be warranted. Future work should also disentangle differences in determinants of audio-only versus audio–video services.
Study limitations include the inability to measure a broader set of characteristics including patient preferences, and the aggregated nature of area-level characteristics. Analyses focus on a Medicaid population, thus generalizability to other insured populations may be limited. In summary, factors associated with telemedicine use reflect potential facilitators and barriers among AIAN communities. These results can inform tailored efforts to increase access to care in this population.
Acknowledgements:
The authors thank the editor and two peer reviewers for their helpful comments.
Funding
This study was funded by a University of Washington Population Health Initiative Tier 3 grant.
Declarations:
Conflict of Interest:
The authors declare that they do not have a conflict of interest.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
REFERENCES
- 1.Sequist TD. Improving the Health of the American Indian and Alaska Native Population. JAMA. 2021;325(11):1035-1036. [DOI] [PubMed] [Google Scholar]
- 2.Grossman DC, Krieger JW, Sugarman JR, Forquera RA. Health status of urban American Indians and Alaska Natives. A population-based study. JAMA. 1994;271(11):845-850. [PubMed] [Google Scholar]
- 3.Sequist TD, Cullen T, Acton KJ. Indian health service innovations have helped reduce health disparities affecting american Indian and alaska native people. Health Aff (Millwood). 2011;30(10):1965-1973. [DOI] [PubMed] [Google Scholar]
- 4.Schorn MN, Myers CR, Barroso J, et al. Changes in Telehealth Experienced by Advanced Practice RNs During COVID-19: US Survey Results. Comput Inform Nurs. 2023;41(7):507-513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Department of Health and Human Services. Telehealth policy changes after the COVID-19 public health emergency. [Online]. Published 2024. Accessed Aug 1, 2024.
- 6.Harvey JB, Valenta S, Simpson K, Lyles M, McElligott J. Utilization of Outpatient Telehealth Services in Parity and Nonparity States 2010-2015. Telemed J E Health. 2019;25(2):132-136. [DOI] [PubMed] [Google Scholar]
- 7.Morenz AM, Reddy A, Hsu A, Le A, Wong ES, Liao JM. Disparities in telemedicine use among Native Hawaiian and Pacific Islander individuals insured through Medicaid. Health Aff Sch. 2024;2(5):qxae057. [DOI] [PMC free article] [PubMed] [Google Scholar]