Skip to main content
NCBI home page
Lippincott Open Access logoLink to Lippincott Open Access
. 2026 Mar 9;49(2):E82–E95. doi: 10.1097/JAC.0000000000000551

Diabetes, Dementia, and Disruptions in Health Care Use in 2020 for Low-Income Medicare Beneficiaries

Avantika Saraf Shah 1,2, Hannah E Peterson 3, Rosette Chakkalakal 4, Stacie Dusetzina 3, Chanee Fabius 2, John Graves 3, Jennifer Kim 5, Michael Mumma 3, Loren Lipworth 3, David G Stevenson 3,6, Laura M Keohane 3,✉
PMCID: PMC12975019  PMID: 41807336

Abstract

Background:

Despite widespread interruptions to health care in 2020, little evidence demonstrates how populations at highest risk for poor outcomes fared across a comprehensive scope of services.

Methods:

Among a predominantly low-income population of individuals ages 65 and older with diabetes (n = 4,187), we measured health care service use in Medicare and Medicaid claims data from 2018 to 2020. Stratified analyses included individuals with Alzheimer disease and related dementias (ADRD, n = 1,025), individuals who used Medicaid-funded home- and community-based services (HCBS, n = 264), and long-term nursing home services (n = 365).

Results:

Relative to 2018–2019, adjusted quarterly rates of evaluation and management visits dropped by 26% (95% confidence interval [CI]: 23%–28%) in Q2 2020 and remained 7% lower (95% CI: 4%–10%) in Q4 2020. Persistent declines occurred for inpatient discharges and emergency room visits (relative risk Q4 2020 vs. 2018–2019: 0.87 [95% CI: 0.76–0.99] and 0.77 [95% CI: 0.69–0.87], respectively). Insulin fills declined in later 2020 (relative risk Q4 2020 vs. 2018–2019: 0.87 [95% CI: 0.79–0.95]) while annual wellness visits rebounded (relative risk Q4 2020 vs. 2018–2019: 1.19 [95% CI: 1.06–1.34]). Individuals who used Medicaid-funded HCBS or long-term nursing home services before the pandemic had large declines in evaluation and management visits (relative risk Q4 2020 vs. 2018–2019: 0.80 [95% CI: 0.69–0.93] and 0.63 [95% CI: 0.43–0.94], respectively). Nursing home residents also had notable declines in insulin fills (relative risk Q4 2020 vs. 2018–2019: 0.73, 95% CI: 0.55–0.96). Individuals with ADRD had increased skilled nursing facility admissions (relative ratio Q3 2020 vs. 2018–2019 1.60, 95% CI: 1.21–2.13). Telehealth usage in 2020 did not differ based on ADRD diagnosis.

Conclusions:

Extended disruptions in routine care highlight opportunities to improve support for older adults with diabetes.

Keywords: Alzheimer disease and related dementias, diabetes, long-term services and supports, pandemic, telehealth

KEY POINTS.

  • As policymakers consider effective strategies to respond to health shocks, evidence from service disruptions in 2020 can highlight which populations could benefit from targeted support.

  • Across a predominantly low-income cohort of Medicare beneficiaries with diabetes, there were significant declines across a broad range of institutional and outpatient services in the second quarter of 2020 compared with previous use in 2018–2019. By the third and fourth quarter of 2020, service volume increased but largely remained below prepandemic levels. Insulin fills declined throughout 2020.

  • Individuals with long-term nursing home service use leading up to the pandemic had some of the largest disruptions in use.

BACKGROUND

The coronavirus disease 2019 (COVID-19) pandemic caused widespread disruptions to health care services, especially for populations at higher risk of severe COVID-19 (Czeisler et al., 2020; Mehrotra et al., 2021; Whaley et al., 2020). Although swift implementation of telehealth services mitigated some access issues, the overall volume of outpatient visits took months to return to prepandemic levels (Patel et al., 2021a, 2021b). Without routine monitoring, adults with diabetes may be at increased risk of complications and hospitalizations (Chan et al., 2021; Valabhji et al., 2022), which is concerning given widespread decreases in monthly Hb1Ac measurements and primary care visits among individuals with diabetes during the pandemic (Adhikari et al., 2023; Amsah et al., 2023; Czeisler et al., 2021). Telehealth services increased in 2020, but not enough to completely offset the decreased volume of in-person office visits among those with diabetes (Patel et al., 2021c).

Concerns about care disruptions are heightened for older adults with diabetes and Alzheimer disease and related dementias (ADRD). Low continuity of care for individuals with ADRD is associated with higher rates of hospitalizations and emergency room visits (Amjad et al., 2016). Relative to individuals without cognitive impairment, during the pandemic, Medicare beneficiaries with ADRD had higher mortality rates (Gilstrap et al., 2022), and individuals with cognitive impairment had larger declines in inpatient and outpatient use (Tannous et al., 2022). Adults with dementia are also more likely to use long-term services and supports (LTSS), one of the hardest hit sectors in the pandemic (Lin et al., 2024; Shen et al., 2021; Shippee et al., 2020; Yi et al., 2020). Under public health emergency measures designed to address overwhelming demand for inpatient services and broaden access to alternative treatment settings, the Centers for Medicare and Medicaid Services waived the skilled nursing facility (SNF) requirement for a prior 3-day inpatient hospital stay. While SNF use declined among community-dwelling Medicare beneficiaries (Ulyte et al., 2023) and commercially insured individuals (Werner & Bressman, 2021), these decreases were offset by increased SNF use among those residing in long-term care facilities (Ulyte et al., 2023). However, there is limited evidence on disruptions in routine health care delivery for LTSS users during the pandemic, especially among those who use Medicaid-funded home- and community-based services (HCBS) (Kaye & Caldwell, 2023).

This study aimed to provide evidence about where greater resources and planning efforts may be needed to ensure continuity of care for older adults with diabetes and dementia during emergencies and unexpected interruptions in health care access. We evaluate 2020 changes in health care utilization across a broad range of services in a predominantly low-income population of older adults in the southeastern United States. Surveys of this population show that 40% reported missing or delaying a health appointment during the pandemic (Ni et al., 2021). Five percent reported avoidance of emergency or urgent care due to fear of COVID-19 (Gettler et al., 2022). We focus on outcomes for older individuals diagnosed with diabetes before 2018 and evaluate whether a subset with ADRD experienced larger disruptions in health care use. By collectively examining outpatient, inpatient, postacute, prescription drug, and telehealth use, we can identify potential relationships and tradeoffs between these services. To assess which populations may be at greater risk of gaps in care, we stratified results for critical subgroups such as those using Medicaid-funded HCBS and nursing home residents.

METHODS

Study population and data sources

We examined a subset of the Southern Community Cohort Study, a predominantly low-income, racially diverse cohort of adults ages 40 to 79 recruited across the southeastern United States over the years 2002–2009 with periodic survey follow-up (Signorello et al., 2010). Most recruitment occurred at community health centers (86%), with 14% of the sample recruited via mail-based population sampling. The baseline survey captured demographic characteristics, including race, ethnicity, and education level. Linked Medicare and Medicaid claims identify participants’ use of health services and diagnostic information over time. Data sources included Medicare Master Beneficiary Summary File data, MEDPar records on inpatient and SNF use, outpatient and carrier claims, Part D prescription drug event data, Minimum Data Set assessments, and Medicaid claims and enrollment data. The Southern Community Cohort Survey (SCCS) study protocol is approved by the institutional review board of Vanderbilt University.

Our main study population included SCCS participants with diabetes who were age 65 or older with traditional Medicare from 2018 to 2020 or from 2018 until death (Supplementary Appendix Figure 1, https://links.lww.com/JACM/A168 for sample details). Main analyses further limited the sample to participants who had at least 24 months of traditional Medicare coverage between 2015 and 2017 to be able to verify the presence of diabetes, ADRD, and other comorbid conditions in claims data. Since the Chronic Condition Warehouse algorithms we used vary in whether there is a 1- or 3-year lookback period for disease diagnoses (Chronic Conditions Warehouse, 2023), we examined claims data from 2013 onwards. We ascertained diabetes and ADRD diagnoses as of the end of 2017, since the pandemic may have delayed new diagnosis of incident diabetes or ADRD. Participants were identified as having diabetes if claims included diabetes diagnoses by the end of 2017. Sensitivity analyses for Medicare Advantage enrollees also included individuals who self-reported diabetes on their baseline survey or had at least two Part D insulin fills in 2017. Participants were identified as having ADRD if claims data indicated ADRD diagnoses by the end of 2017.

We also performed stratified analyses for several populations that might be at greater risk for health care disruptions during the pandemic (Czeisler et al., 2020; Gettler et al., 2022; Ni et al., 2021): participants who were Black, female, dual-eligible with Medicare and Medicaid coverage, had a high school education or less, or had at least two outpatient insulin Part D fills prescriptions in 2017. Using Medicaid claims data and Minimum Data Set nursing home assessments, we also report results for individuals who used long-term nursing home services or Medicaid-funded HCBS in 2017.

In sensitivity analyses, we examined participants with Medicare Advantage enrollment in 2015–2017 whose hospital and prescription drug use are captured in MEDPar and Part D prescription drug data. These results should be compared with results for individuals with traditional Medicare with caution as we did not have complete outpatient claims data to establish diabetes diagnoses. The traditional Medicare cohort includes individuals who are identified in outpatient claims data as having a diabetes diagnosis, even if they were not using insulin. The Medicare Advantage cohort includes individuals who self-reported at SCCS baseline that they had diabetes, individuals who had two or more insulin fills in 2017, and individuals who had an inpatient claim with diabetes or a prior outpatient claim indicating diabetes (for example, a claim from before Medicare Advantage enrollment). Because of these differences, the Medicare Advantage cohort is more likely to have longer-standing diabetes diagnoses and possibly more advanced diabetes than the traditional Medicare cohort. The most comparable groups are the subset of traditional Medicare and Medicare Advantage beneficiaries with at least two insulin fills in 2017.

Study measures

We examined the quarterly rates per beneficiary, or number of events per beneficiary per quarter, of several outcomes between 2018 and 2020: number of inpatient discharges, SNF admissions, emergency room visits (categorized by whether visits transitioned into inpatient stays or not), observation stays, office-based evaluation and management visits, and annual wellness visits (measures detailed in the Supplementary Appendix, https://links.lww.com/JACM/A168). For participants with Part D prescription drug coverage, we captured the total number of insulin fills. We measured telehealth services in the year 2020 only, as these services were rarely used before the pandemic. Beneficiaries contribute to the denominator of health services use measures until death. We also present quarterly mortality rates, or the number of deaths among beneficiaries alive at the beginning of each quarter, stratified by ADRD status.

Claims diagnoses identified whether participants had comorbid conditions associated with diabetes: chronic obstructive pulmonary disease, kidney disease, heart failure, ischemic heart disease, hyperlipidemia, or hypertension (Chronic Conditions Warehouse, 2023). We also identified whether participants had any history of amputation or severe hypoglycemia that was treated in an inpatient or emergency room setting between 2015 and 2017.

Long-term nursing home users were identified based on the presence of at least one quarterly or annual nursing home assessment, a measurement approach used in prior work (Werner et al., 2013). Medicaid HCBS users were identified based on meeting any of these conditions: enrollment in a 1915c waiver program, presence of a claim funded by a 1915c waiver program, or presence of a claim for personal care services. This HCBS measure is a modified version of the approach detailed by Stepanczuk et al., 2022.

Analytic approach

For each outcome, we report the relative risk of using these services in each quarter of 2020 compared with the average rate of service use in 2018 and 2019. The underlying unit of observation for this model is beneficiary-month. Because analyses followed the same cohort over time, we expected that individuals’ health care service use might increase over time as they age. We used negative binomial models that adjusted for age, sex, education level, number of chronic conditions, and insulin use. We tested whether trends during the pandemic were different for individuals with pre-existing ADRD by including an interaction term between ADRD status and the quarter indicators for 2020.

RESULTS

Baseline characteristics

Our sample included 4,187 Medicare participants diagnosed with diabetes by December 2017. Table 1 presents the demographic and clinical characteristics of the population at the start of 2018 based on whether they were diagnosed with ADRD by December 2017. Overall, the cohort included 2,763 (66%) women, 2,297 (54.9%) Black individuals, and 2,369 individuals (56.6%) with a high school education or less. In addition, 1,881 (44.9%) individuals were dual-eligible, 365 used long-term nursing home services (8.7%), and 264 (6.3%) used Medicaid-funded HCBS in 2017. Individuals averaged 3.3 (SD: 1.4) comorbidities.

Table 1.

Characteristics of Southern Community Cohort Survey Traditional Medicare beneficiaries with diabetes by ADRD diagnosis, 2018

Characteristics Total No ADRD Diagnosis ADRD Diagnosis P Value
N = 4,187 N = 3,162 N = 1,025
Female 2,763 (66.0%) 2,060 (65.1%) 703 (68.6%) 0.044
Age 74.8 (6.3) 73.8 (5.8) 77.8 (6.9) <0.001
Race 0.75
 White 1,669 (39.9%) 1,250 (39.5%) 419 (40.9%)
 Black 2,297 (54.9%) 1,744 (55.2%) 553 (54.0%)
 Other/missing 221 (5.3%) 168 (5.3%) 53 (5.2%)
Education <0.001
 <9 years 500 (11.9%) 339 (10.7%) 161 (15.7%)
 9 to 11 years 734 (17.5%) 538 (17.0%) 196 (19.1%)
 High school 1,135 (27.1%) 848 (26.8%) 287 (28.0%)
 Some college or technical training 939 (22.4%) 727 (23.0%) 212 (20.7%)
 College or advanced degree 755 (18.0%) 618 (19.5%) 137 (13.4%)
 Missing/not reported 124 (3.0%) 92 (2.9%) 32 (3.1%)
Has Part D benefits 3,273 (78.2%) 2,408 (76.2%) 865 (84.4%) <0.001
Original Medicare eligibility due to disability or ESRD 1,543 (36.9%) 1,178 (37.3%) 365 (35.6%) 0.34
Dual-eligible (2017) 1,881 (44.9%) 1,247 (39.4%) 634 (61.9%) <0.001
Nursing home use (2017) 365 (8.7%) 58 (1.8%) 307 (30.0%) <0.001
Medicaid-funded HCBS use (2017) 264 (6.3%) 160 (5.1%) 104 (10.1%) <0.001
Insulin use (2017) 762 (18.2%) 560 (17.7%) 202 (19.7%) 0.15
Amputation or severe hypoglycemia (2015–2017) 155 (3.7%) 85 (2.7%) 70 (6.8%) <0.001
COPD 1,097 (26.2%) 790 (25.0%) 307 (30.0%) 0.002
Kidney disease 2,596 (62.0%) 1,947 (61.6%) 649 (63.3%) 0.32
Heart failure 1,308 (31.2%) 906 (28.7%) 402 (39.2%) <0.001
Ischemic heart disease 1,874 (44.8%) 1,373 (43.4%) 501 (48.9%) 0.002
Hyperlipidemia 3,195 (76.3%) 2,522 (79.8%) 673 (65.7%) <0.001
Hypertension 3,856 (92.1%) 2,941 (93.0%) 915 (89.3%) <0.001
Number of chronic conditions 3.3 (1.4) 3.3 (1.4) 3.4 (1.6) 0.34
Health care use per 100 participants, Q1 2018
 Inpatient discharges 18.9 (56.4) 16.1 (53.4) 27.4 (64.0) <0.001
 SNF admissions 5.2 (27.1) 3.0 (22.1) 11.8 (37.9) <0.001
 Emergency room outpatient visits 22.3 (63.7) 20.3 (59.3) 28.4 (75.3) <0.001
 Inpatient discharges that included ER use 14.2 (47.6) 11.6 (43.8) 22.4 (57.1) <0.001
 Observation stays 2.8 (17.4) 2.8 (17.3) 3.0 (17.7) 0.66
 Annual wellness visits 5.7 (24.8) 5.9 (24.9) 5.0 (24.3) 0.29
 Office evaluation and management visits 264.5 (265.1) 290.9 (268.9) 182.8 (235.2) <0.001
 Insulin fillsa 56.2 (151.2) 48.0 (121.0) 79.0 (212.3) <0.001
Open in a new tab

Demographic and clinical characteristics of the study population based on whether they were diagnosed with ADRD according to claims data by December 2017. Data sourced from SCCS data, Medicare and Medicaid claims and enrollment data, and Minimum Data Set records. Race and education based on self-reported survey data. All other variables are derived from claims or Minimum Data Set records. Individuals with diabetes were identified based on the presence of a diabetes diagnosis in claims data in 2016 or 2017. Individuals with Alzheimer disease and related dementias were identified based on the presence of corresponded diagnosis in claims data in 2015–2017. History of dual-eligible status, long-term nursing home use, Medicaid-funded home- and community-based services use, and insulin use were based on 2017 data. A history of amputation or severe hypoglycemia was established based on 2015 to 2017 claims data.

COPD, chronic obstructive pulmonary disease; ER, emergency room.

a

Number of insulin fills per 100 participants limited to Part D participants (N = 3,238).

A total of 1,025 (24.5%) individuals had a diagnosis of ADRD. Relative to those without ADRD, individuals with ADRD were significantly older, had lower education levels, were more likely to be dual-eligible, and to use LTSS. Regarding clinical factors, older adults with ADRD were also more likely to have heart failure (39.2% vs. 28.7%), ischemic heart disease (48.9% vs. 43.4%), and experience a severe hypoglycemic episode or undergo an amputation (6.8% vs. 2.7%) between 2015 and 2017 compared with individuals without ADRD.

Relative changes in health care utilization for individuals with diabetes

Relative to the same time period in 2018–2019, the study population experienced significant reductions during Q2 2020 in inpatient discharges [Figure 1 and Supplementary Appendix Table 1, https://links.lww.com/JACM/A168 relative risk of 0.73, 95% confidence interval (CI): 0.64–0.83], emergency room visits (0.64, 95% CI: 0.57–0.71), observation stays (0.55, 95% CI: 0.41–0.73), annual wellness visits (0.71, 95% CI: 0.61–0.82), and office-based evaluation and management visits (0.74, 95% CI: 0.72–0.77). In Q3 and Q4 2020, use of almost all services increased from earlier in the year, but still remained below prepandemic levels. Only annual wellness visits showed evidence of rebounding to levels that were higher than observed in prior years (relative risk for Q3: 1.25, 95% CI: 1.10–1.42) and Q4 (1.19, 95% CI: 1.06–1.34). Unlike other services, the rate of SNF admissions did not significantly change in any quarter of 2020. Relative to prior years, the rate of insulin fills was 9% higher in Q1 2020 (95% CI: 1%–18%), but declined over the course of 2020. By Q4 2020, the rate of insulin fills was 13% lower than in previous years (95% CI: 5%–21%).

Figure 1.

Figure 1.

Open in a new tab

Relative risk ratio for changes in health care use for Southern Community Cohort Survey traditional medicare beneficiaries with diabetes, 2020 vs. 2018–2019. Data sourced from SCCS data, Medicare and Medicaid claims and enrollment data, and Minimum Data Set records. Results based on negative binomial models adjusted for age, sex, education level, number of chronic conditions, and any history of insulin use. The unit of analysis is person-month. ER, emergency room; Eval. & Man., evaluation and management visits.

Relative changes in health care utilization for individuals with diabetes and Alzheimer disease and related dementias

Overall, the relative changes in health care use for participants with diabetes and ADRD (Figure 2) were consistent with the results seen in the overall population with diabetes, with a notable exception for SNF use. Among individuals with diabetes and ADRD, SNF admissions increased in Q2 through Q4 relative to 2018–2019 (Q2 relative ratio: 1.33, 95% CI: 0.99–1.78; Q3 relative ratio: 1.60, 95% CI: 1.21–2.13; Q4 relative ratio: 1.36, 95% CI: 0.98–1.9). According to interaction terms in our model that tested for differences based on ADRD status (Supplementary Appendix Table 2, https://links.lww.com/JACM/A168), increases in SNF use were significantly greater for ADRD patients than the differences observed over the same time period for those without ADRD in Q2 (1.91, 95% CI: 1.20–3.03) and Q3 (2.02, 95% CI: 1.29–3.16). Trends in health care use also reflect increased mortality in 2020: older adults with ADRD had higher mortality rates than those without ADRD in all study years, but especially in Q2 to Q4 of 2020 (Supplementary Appendix Figure 2, https://links.lww.com/JACM/A168).

Figure 2.

Figure 2.

Open in a new tab

Relative risk ratio for changes in for Southern Community Cohort Survey traditional Medicare beneficiaries with diabetes and ADRD, 2020 vs. 2018–2019. Data sourced from Southern Community Cohort Survey data, Medicare and Medicaid claims and enrollment data, and Minimum Data Set records. Results based on negative binomial models adjusted for age, sex, education level, number of chronic conditions, and any history of insulin use. Unit of analysis is person-month. ER, emergency room; Eval. & Man., evaluation and management visits.

Subgroup analyses

For most health care services, patterns did not vary across the following subgroups of individuals with diabetes (Figure 3, Supplementary Appendix Figure 3, https://links.lww.com/JACM/A168): individuals who were Black, female, dual-eligible, had a high school education or less, used insulin, and had long-term nursing home use or Medicaid-funded HCBS use before the pandemic. Notably, individuals who used LTSS before the pandemic had the largest disruptions in office-based evaluation and management visits. Relative to prior years, the rate of evaluation and management visits for HCBS users was 0.59 (95% CI: 0.49–0.70) times lower in Q2 2020 and remained 0.80 (95% CI: 0.69–0.93) times lower in Q4 2020. Compared with 2018–2019, the rate of evaluation and management visits for long-term nursing home residents was 0.49 (95% CI: 0.33–0.73) times lower in Q2 2020 and remained 0.61 (95% CI: 0.43–0.85) times lower in Q4 2020. Supplementary Appendix Figure 3, https://links.lww.com/JACM/A168 shows that long-term nursing home residents also had large declines in the rate of insulin fills by Q4 2020 (0.73, 95% CI: 0.55–0.96).

Figure 3.

Figure 3.

Open in a new tab

Relative changes in evaluation and management visits by Southern Community Cohort Survey subgroup, 2020 vs. 2018–2019. Analysis created from Southern Community Cohort Survey data, Medicare and Medicaid claims and enrollment data, and Minimum Data Set records. Results based on stratified negative binomial models adjusted for age, sex, education level, number of chronic conditions, and any history of insulin use. Unit of analysis is person-month. *Presence of ADRD diagnosis, insulin use, and use of nursing home or Medicaid HCBS services was identified based on 2017 claims data.

When we evaluated changes in inpatient use among Medicare Advantage enrollees, we found similar trends as observed among the main cohort of traditional Medicare enrollees (Supplementary Appendix Figure 4, https://links.lww.com/JACM/A168). Relative to 2018–2019, inpatient admissions declined by 27% (relative risk: 0.73, 95% CI: 0.63–0.86) in Q2 2020 relative to years 2018–2019 but rebounded by quarter 4 (relative risk: 0.88, 95% CI: 0.76–1.03). The rates of insulin fills in 2020 were fairly consistent with previous years among Medicare Advantage enrollees. Only one subgroup of Medicare Advantage enrollees, individuals with a history of insulin use, had a significant decline in the rate of insulin fills in Q3: relative risk: 0.91, 95% CI: 0.85–0.98.

Telehealth utilization for individuals with diabetes and ADRD

As shown in Supplementary Appendix Figure 5, https://links.lww.com/JACM/A168, the mean number of telehealth visits increased sharply between the first and second quarter of 2020 regardless of ADRD diagnosis. In Q2 2020, the mean number of telehealth visits per 100 participants was 82.5 (95% CI: 71.5–93.4) for participants with ADRD and 78.0 (95% CI: 72.9–83.2) for participants without ADRD. This rate decreased for both populations in subsequent quarters, declining to 40.1 (95% CI: 33.1–48.5) in Q4 for participants with ADRD and to 36.8 (95% CI: 33.1–40.6) for participants without ADRD. The percentage of participants with any telehealth use peaked in Q2 2020 at 38.4% and 39.8%, respectively, among participants with and without ADRD (data not shown). After adjusting for age, sex, education, number of chronic conditions, and insulin use the mean number of telehealth visits was slightly higher among participants with ADRD as compared with participants without ADRD in Q2 (Supplementary Appendix Table 3, https://links.lww.com/JACM/A168, relative risk: 1.18, 95% CI: 1.01–1.38) and Q3 (1.26, 95% CI: 1.04–1.53).

DISCUSSION

Within a predominantly low-income cohort of older adults with diabetes, we found widespread disruptions in health care use in 2020 relative to 2018–2019, including significant reductions in inpatient discharges, emergency room visits, observation stays, annual wellness visits, and outpatient evaluation and management visits. While utilization partially rebounded in later quarters, health services use remained lower than prepandemic levels for inpatient discharges, emergency room visits, office-based evaluation and management visits, and insulin fills. Notably, the rate of insulin fills fell over the year, with larger declines in use by the end of 2020. Only annual wellness visits showed greater volume in the latter half of 2020, offsetting decreases earlier in 2020.

Adults with diabetes and ADRD largely had comparable trends to the overall cohort in health care utilization, including similar uptake of telehealth services. The most notable exception was SNF admissions. Adults without ADRD did not experience significant changes in SNF use, while adults with ADRD had significantly more SNF admissions in quarter three of 2020 compared with the preceding years. Individuals with a history of using Medicaid-funded HCBS or long-term nursing home services also stood out for large, persistent declines in the rates of evaluation and management visits. Nursing home residents also experienced exceptionally large declines in the number of insulin fills.

Our findings are consistent with other studies observing significant disruptions in health care utilization during 2020, particularly among older adults with chronic conditions and among lower-income groups (Czeisler et al., 2020; Mehrotra et al., 2021; Smith et al., 2022; Whaley et al., 2020). These studies, like ours, noted that while telehealth services increased, they did not fully offset reduced in-person care. However, our study extends this body of work by examining a broader range of services and stratifying results by ADRD, HCBS status, and long-term nursing home use. These subgroup analyses reveal important differences in care disruptions that have not been well-documented, particularly the persistent declines in outpatient care among individuals using LTSS, insulin fills among nursing home residents, and the increase in SNF admissions among individuals with ADRD.

As the United States considers what policies can prepare for future pandemics and health emergencies, our findings highlight several important areas to consider for older adults with diabetes. First, SNF use was the only service where older adults with comorbid ADRD had substantially distinct patterns in 2020. This work complements previous findings demonstrating that SNF use decreased among community-dwelling Medicare beneficiaries but increased among nursing home residents (who are more likely to have ADRD) when Medicare waived prior hospital stay requirements for SNF coverage (Ulyte et al., 2023). Policy efforts should consider what resources SNFs need, both in terms of funding and workforce, to provide high-quality care during emergencies as their patient populations shift towards individuals with greater levels of frailty or cognitive impairment. Second, individuals with diabetes who were using long-term nursing home residents before the pandemic had persistent and large disruptions in office-based evaluation and management visits and insulin fills. These findings underscore the heightened vulnerability of nursing home residents during public health emergencies, likely due to facility lockdowns, staffing shortages, and infection control measures that limited access to routine care. Medicaid-funded HCBS users also experienced large disruptions in evaluation and management visits, but no disruptions in insulin fills. This distinction is important for policy planning, as it suggests that different LTSS populations may require tailored strategies to maintain continuity of care during crises. For nursing home residents, this may include bolstering on-site primary care capacity and ensuring safe access to essential services. For HCBS users, strengthening remote care infrastructure and caregiver support may be more critical.

Finally, despite concerns that older adults with ADRD might have more difficulties navigating telehealth services, rates of telehealth use did not differ between these two populations. This finding, which is consistent with prior evidence on telehealth use among community-dwelling Medicare beneficiaries (Wang et al., 2024), speaks to the importance of continuing access to telehealth on a permanent basis for older Medicare beneficiaries who might have difficulties traveling to see providers in-person. However, telehealth use for individuals with ADRD often involves additional complexities, such as the need for provider support or caregiver participation to navigate virtual platforms (Weiss et al., 2021). Future research should assess the preferences of individuals of ADRD and their caregivers regarding specific modalities of telehealth and options for caregivers to participate virtually. Moreover, additional evidence is needed on the overall effectiveness of telehealth for patients with diabetes and ADRD, as some evidence points towards mixed outcomes among patients with diabetes who use telehealth (Quinton et al., 2022; Zupa et al., 2023), and its convenience might not always translate to sufficient care for patients with complex care needs (Ladin et al., 2021; Sun et al., 2023).

Limitations of this work include that our results are not representative of the general population. We focus on disruptions in health care service use, but do not evaluate downstream effects on long-term health outcomes. Some studies have suggested that disruptions in routine diabetes care during the pandemic are linked to long-term adverse health outcomes such as more diabetes-related amputations (Pride et al., 2023), increased risk for kidney disease (Harding et al., 2023), and increased mortality (Lv et al., 2022; Ran et al., 2021; Valabhji et al., 2022), although other studies find no or small changes in glycemic control (Chen et al., 2022; Patel et al., 2021). Other limitations include a limited sample size to evaluate smaller changes in health care utilization among older adults with diabetes and ADRD. Data on health care services and diagnoses for Medicare Advantage enrollees is incomplete, restricting our ability to identify populations of interest through claims-based algorithms and to measure health care service use other than inpatient care and prescription drug use. Due to data availability constraints, we could only establish whether individuals were using LTSS before the pandemic, but could not evaluate changes in the use of LTSS during the pandemic. We cannot observe insulin use among the cohort members who did not have Part D coverage (about one in five traditional Medicare beneficiaries) and may undercount the number of prior insulin users.

CONCLUSION

Our results offer insights for policymakers seeking to strengthen continuity of care for older adults with diabetes during future pandemics or emergencies. Among a largely low-income cohort, we found persistent disruptions across a wide range of health care services in 2020. Several findings, including increased SNF admissions for individuals with ADRD and large and persistent decreases in evaluation and management visits for LTSS users, highlight how future emergency planning efforts should address the exceptional circumstances of frailer older adults.

Supplementary Material

jamcm-49-e82-s001.pdf (810.7KB, pdf)
Open in a new tab

Footnotes

The authors would like to thank Taofik Oyekunle for assistance with claims data coding. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

This study was funded by the National Institute on Aging (R01AG068606 and R01AG068606-01S1). The Southern Community Cohort Study was funded by the National Cancer Institute (U01CA202979, U01CA202979-05S1, U01CA202979-03S1). Dr. Keohane was additionally funded by Award Number K01AG058700 from the National Institute on Aging. Dr. Dusetzina was additionally funded as a research scientist through the National Cancer Institute (P30 CA068485). Southern Community Cohort Survey data collection was performed by the Survey and Biospecimen Shared Resource, which is supported in part by the Vanderbilt-Ingram Cancer Center (P30 CA68485). Dr. Dusetzina receives funding from Arnold Ventures and the Commonwealth Fund for work related to drug pricing and access.

Dr. Dusetzina is a member of the Institute for Clinical and Economic Review’s (ICER) Midwestern Comparative Effectiveness Advisory Council and a member of the Medicare Payment Advisory Commission (MedPAC). This work does not necessarily represent the official position of ICER or MedPAC. The remaining authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article.

At the time this work was completed, Hannah E. Peterson worked for the Vanderbilt University School of Medicine, Nashville, Tennessee.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.ambulatorycaremanagement.com).

REFERENCES

  1. Adhikari S. Titus A. R. Baum A. Lopez P. Kanchi R. Orstad S. L., … Schwartz M. D. (2023). Disparities in routine healthcare utilization disruptions during COVID-19 pandemic among veterans with type 2 diabetes. BMC Health Services Research, 23(1, 41. doi:10.1186/s12913-023-09057-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Amjad H. Carmichael D. Austin A. M. Chang C. H., & Bynum J. P. W. (2016). Continuity of care and health care utilization in older adults with dementia in fee-for-service medicare. JAMA Internal Medicine, 176(9, 1371–1378. doi:10.1001/jamainternmed.2016.3553 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Amsah N. Md Isa Z. Ahmad N., & Abdul Manaf M. R. (2023). Impact of COVID-19 pandemic on healthcare utilization among patients with type 2 diabetes mellitus: a systematic review. International Journal of Environmental Research and Public Health, 20(5, 4577. doi:10.3390/ijerph20054577 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chan K. S. Wan E. Y. F. Chin W. Y. Cheng W. H. Ho M. K. Yu E. Y., & Lam C. L. (2021). Effects of continuity of care on health outcomes among patients with diabetes mellitus and/or hypertension: a systematic review. BMC Family Practice, 22(1, 145. doi:10.1186/s12875-021-01493-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen J. L. Krupp G. R., & Lo J. Y. (2022). The COVID-19 pandemic and changes in health care utilization among patients with type 2 diabetes. Diabetes Care, 45(4, e74–e76. doi:10.2337/dc21-2248 [DOI] [PubMed] [Google Scholar]
  6. Chronic Conditions Warehouse. (2023). 27 CCW Chronic Conditions Algorithms: MBSF_CC_{YYYY} File. Accessed January 7, 2025. https://www2.ccwdata.org/documents/10280/19139421/ccw-chronic-condition-algorithms.pdf [Google Scholar]
  7. Czeisler M. E. Marynak K., & Clarke K. (2020). Delay or avoidance of medical care because of COVID-19–related concerns — United States, June 2020. Morbidity and Mortality Weekly Report, 69, 1250–1257. doi:10.15585/mmwr.mm6936a4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Czeisler M. E. Barrett C. E. Siegel K. R. Weaver M. D. Czeisler C. A. Rajaratnam S. M. W., … Bullard K. M. (2021). Health Care access and use among adults with diabetes during the COVID-19 pandemic – United States, February-March 2021. Morbidity and Mortality Weekly Report, 70(46, 1597–1602. doi:10.15585/mmwr.mm7046a2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Gettler E. Stern R. Ni B. Munro H. M. Steinwandel M. Aronoff D. M., … Lipworth L. (2022). Avoidance of emergency care in the Southeastern United States during the COVID-19 pandemic. Open Forum Infectious Diseases, 9(5, ofac161. doi:10.1093/ofid/ofac161 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gilstrap L. Zhou W. Alsan M. Nanda A., & Skinner J. S. (2022). Trends in mortality rates among medicare enrollees with Alzheimer disease and related dementias before and during the early phase of the COVID-19 pandemic. JAMA Neurology, 79(4, 342–348. doi:10.1001/jamaneurol.2022.0010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Harding J. L. Wang Y. McDonald B. Gander J. C. Oviedo S. A., & Ali M. K. (2023). Disruption to routine diabetes care processes during the pandemic: evidence from a large integrated health system in the Southeast United States. BMJ Open Diabetes Research and Care, 11(4, e003466. doi:10.1136/bmjdrc-2023-003466 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Kaye H. S., & Caldwell J. (2023). Excess deaths of medicaid home and community-based services recipients during COVID-19. Health Affairs (Project Hope), 42(1, 115–120. doi:10.1377/hlthaff.2022.00457 [DOI] [PubMed] [Google Scholar]
  13. Ladin K. Porteny T. Perugini J. M. Gonzales K. M. Aufort K. E. Levine S. K., … Weiner D. E. (2021). Perceptions of telehealth vs in-person visits among older adults with advanced kidney disease, care partners, and clinicians. JAMA Network Open, 4(12, e2137193. doi:10.1001/jamanetworkopen.2021.37193 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Lin S. C. Zheng J. Epstein A. Orav E. J. Barnett M. Grabowski D. C., & Joynt Maddox K. E. (2024). The extended impact of the COVID-19 pandemic on long-term care residents in Medicare with frailty or dual Medicaid enrollment. Journal of the American Geriatrics Society, 72(11, 3457–3466. doi:10.1111/jgs.19131 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Lv F. Gao X. Huang A. H. Zu J. He X. Sun X., … Ji F. (2022). Excess diabetes mellitus-related deaths during the COVID-19 pandemic in the United States. eClinicalMedicine, 54, 101671. doi:10.1016/j.eclinm.2022.101671 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Mehrotra A. Chernew M. Linetsky D. Hatch H. Cutler D. M., & Schneider E. C. (2021). The impact of COVID-19 on outpatient visits in 2020: visits remained stable, despite a late surge in cases. Commonwealth Fund. doi:10.26099/bvhf-e411 [Google Scholar]
  17. Ni B. Gettler E. Stern R. Munro H. M. Steinwandel M. Aldrich M. C., … Lipworth L. (2021). Disruption of medical care among individuals in the southeastern United States during the COVID-19 pandemic. Journal of Public Health Research, 11(1, 2497. doi:10.4081/jphr.2021.2497 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Patel S. Y. Mehrotra A. Huskamp H. A. Uscher-Pines L. Ganguli I., & Barnett M. L. (2021. a). Trends in outpatient care delivery and telemedicine during the COVID-19 pandemic in the US. JAMA Internal Medicine, 181(3, 388–391. doi:10.1001/jamainternmed.2020.5928 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Patel S. Y. Mehrotra A. Huskamp H. A. Uscher-Pines L. Ganguli I., & Barnett M. L. (2021. b). Variation in telemedicine use and outpatient care during the COVID-19 pandemic in the United States. Health Affairs, 40(2, 349–358. doi:10.1377/hlthaff.2020.01786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Patel S. Y. McCoy R. G. Barnett M. L. Shah N. D., & Mehrotra A. (2021. c). Diabetes care and glycemic control during the COVID-19 pandemic in the United States. JAMA Internal Medicine, 181(10, 1412–1414. doi:10.1001/jamainternmed.2021.3047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Pride L. Kabeil M. Alabi O. Minc S. D. Fakorede F. A. Ochoa L. N., … Wohlauer M. V. (2023). A review of disparities in peripheral artery disease and diabetes–related amputations during the COVID-19 pandemic. Seminars in Vascular Surgery, 36(1, 90–99. doi:10.1053/j.semvascsurg.2022.12.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Quinton J. K. Ong M. K. Sarkisian C. Casillas A. Vangala S. Kakani P., & Han M. (2022). The impact of telemedicine on quality of care for patients with diabetes after March 2020. Journal of General Internal Medicine, 37(5, 1198–1203. doi:10.1007/s11606-021-07367-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Ran J. Zhao S. Han L. Ge Y. Chong M. K. C. Cao W., & Sun S. (2021). Increase in diabetes mortality associated with COVID-19 pandemic in the U.S. Diabetes Care, 44(7, e146–e147. doi:10.2337/dc21-0213 [DOI] [PubMed] [Google Scholar]
  24. Shen K. Loomer L. Abrams H. Grabowski D. C., & Gandhi A. (2021). Estimates of COVID-19 cases and deaths among nursing home residents not reported in federal data. JAMA Network Open, 4(9, e2122885. doi:10.1001/jamanetworkopen.2021.22885 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Shippee T. P. Akosionu O. Ng W. Woodhouse M. Duan Y. Thao M. S., & Bowblis J. R. (2020). COVID-19 pandemic: exacerbating racial/ethnic disparities in long-term services and supports. Journal of Aging and Social Policy, 32(4-5, 323–333. doi:10.1080/08959420.2020.1772004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Signorello L. B. Hargreaves M. K., & Blot W. J. (2010). The Southern Community Cohort Study: investigating health disparities. Journal of Health Care for the Poor and Underserved, 21(Suppl 1, 26–37. doi:10.1353/hpu.0.0245 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Smith M. Vaughan Sarrazin M. Wang X. Nordby P. Yu M. DeLonay A. J., & Jaffery J. (2022). Risk from delayed or missed care and non-COVID-19 outcomes for older patients with chronic conditions during the pandemic. Journal of the American Geriatrics Society, 70(5, 1314–1324. doi:10.1111/jgs.17722 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Stepanczuk C. Murray C. Carpenter A., & Wysocki A. (2022). Methodology for identifying medicaid long-term services and supports expenditures and users. Accessed October 8, 2025. https://www.medicaid.gov/medicaid/long-term-services-supports/downloads/ltss-users-expenditures-method-2022.pdf [Google Scholar]
  29. Sun C. A. Shenk Z. Renda S. Maruthur N. Zheng S. Perrin N., … Han H. -R. (2023). Experiences and perceptions of telehealth visits in diabetes care during and after the COVID-19 pandemic among adults with type 2 diabetes and their providers: qualitative study. JMIR Diabetes, 8(1, e44283. doi:10.2196/44283 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Tannous J. Pan A. Bako A. Potter T. Jones S. L. Janjan N., … Vahidy F. S. (2022). COVID-19 associated disruptions in routine health care of people with mild cognitive impairment or dementia. Alzheimer’s & Dementia (Amsterdam, Netehrelands), 14(1, e12323. doi:10.1002/dad2.12323 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Ulyte A. Waken R. J. Epstein A. M. Orav E. J. Barnett M. L. Joynt Maddox K. E., & Grabowski D. C. (2023). Medicare skilled nursing facility use and spending before and after introduction of the public health emergency waiver during the COVID-19 pandemic. JAMA Internal Medicine, 183(7, 637–645. doi:10.1001/jamainternmed.2023.0770 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Valabhji J. Barron E. Gorton T. Bakhai C. Kar P. Young B., … Wareham N. J. (2022). Associations between reductions in routine care delivery and non-COVID-19-related mortality in people with diabetes in England during the COVID-19 pandemic: a population-based parallel cohort study. Lancet Diabetes & Endocrinology, 10(8, 561–570. doi:10.1016/S2213-8587(22)00131-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wang N. Seale M., & Chen J. (2024). Availability and use of telehealth services among patients with ADRD enrolled in traditional Medicare vs. Medicare Advantage during the COVID-19 pandemic. Frontiers in Public Health, 12, 1346293. doi:10.3389/fpubh.2024.1346293 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Weiss E. F. Malik R. Santos T. Ceide M. Cohen J. Verghese J., & Zwerling J. L. (2021). Telehealth for the cognitively impaired older adult and their caregivers: lessons from a coordinated approach. Neurodegenerative Disease Management, 11(1, 83–89. doi:10.2217/nmt-2020-0041 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Werner R. M., & Bressman E. (2021). Trends in post-acute care utilization during the COVID-19 pandemic. Journal of the American Medical Directors Association, 22(12, 2496–2499. doi:10.1016/j.jamda.2021.09.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Werner R. M. Konetzka R. T., & Polsky D. (2013). The effect of pay-for-performance in nursing homes: evidence from state Medicaid programs. Health Services Research, 48(4, 1393–1414. doi:10.1111/1475-6773.12035 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Whaley C. M. Pera M. F. Cantor J. Chang J. Velasco J. Hagg H. K., … Bravata D. M. (2020). Changes in health services use among commercially insured US populations during the COVID-19 pandemic. JAMA Network Open, 3(11, e2024984. doi:10.1001/jamanetworkopen.2020.24984 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Yi S. H. See I. Kent A. G. Vlachos N. Whitworth J. C. Xu K., … Stuckey M. J. (2020). Characterization of COVID-19 in Assisted Living Facilities – 39 States, October 2020. Morbidity and Mortality Weekly Report, 69(46, 1730–1735. doi:10.15585/mmwr.mm6946a3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zupa M. F. Vimalananda V. G. Rothenberger S. D. Lin J. Y. Ng J. M. McCoy R. G., & Rosland A. -M. (2023). Patterns of telemedicine use and glycemic outcomes of endocrinology care for patients with type 2 diabetes. JAMA Network Open, 6(12, e2346305. doi:10.1001/jamanetworkopen.2023.46305 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of Ambulatory Care Management are provided here courtesy of Wolters Kluwer Health

RESOURCES