At-Home Hemoglobin A1c Testing During COVID-19 Improved Glycemic Control

ABSTRACT

Objectives: COVID-19 has exacerbated barriers to routine testing for chronic disease management. This study investigates whether a home hemoglobin A_1c (HbA_1c) test kit intervention increases frequency of HbA_1c testing and leads to changes in HbA1c 6 months post testing and whether self-reinforcement education improves maintenance of HbA_1c testing.

Study Design: Retrospective analysis of a randomized, controlled quality improvement intervention among members with type 2 diabetes (T2D) in a large commercial health plan.

Methods: Participants were 41,214 commercial fully insured members with T2D without an HbA_1c test in the past 6 months or with only 1 HbA_1c test in the last 12 months. Members were randomly assigned to either a control group or an at-home HbA_1c testing intervention group consisting of either an opt-in test or a direct-to-member opt-out HbA_1c test kit shipment. A third cohort of members was assigned to a self-reinforcement group to encourage continued testing twice per year. Main outcomes were HbA_1c testing rates and HbA_1c levels (in %).

Results: A total of 11.1% (508 of 4590) at-home HbA_1c kits were completed. At-home HbA_1c test kits increased testing rates by 4.9% compared with controls (P < .001). Members with an HbA_1c level of at least 7% who requested and completed at-home HbA_1c testing had a 0.38% reduction in HbA_1c in the 6 months post intervention when controlling for baseline HbA_1c (P < .001). Members who received self-reinforcement messaging had a 0.37% HbA_1c reduction post intervention (P = .015).

Conclusions: This novel, at-home approach to test HbA_1c is an effective intervention to increase testing rates and facilitate HbA_1c reduction over time in patients with T2D.

Am J Manag Care. 2024;30(3):e73-e77. https://doi.org/10.37765/ajmc.2024.89449

Takeaway Points

This randomized, controlled quality improvement intervention among commercial fully insured members with diabetes in a large health plan shows that at-home hemoglobin A_1c (HbA_1c) kits increased the frequency of testing rates and led to changes in HbA_1c 6 months post testing.

• Delivering at-home HbA_1c test kits directly to members increased testing rates by 4.9% compared with controls (P < .001).
• Members with an HbA_1c level of at least 7% who requested and completed at-home HbA_1c testing experienced a 0.38% reduction in HbA_1c on average compared with controls (P < .001).
• This novel at-home approach to test HbA_1c is an effective intervention to increase testing rates and facilitate HbA_1c reduction over time.

Diabetes is a complex chronic condition affecting nearly 11.3% of the US population.¹ Despite guidelines recommending hemoglobin A_1c (HbA_1c) testing at least twice per year in patients whose HbA1c level is at goal and at least 4 times per year in patients with inadequate diabetes control,² less than half of US adults with type 2 diabetes (T2D) report regular HbA_1c testing.³ The COVID-19 pandemic has exacerbated barriers to routine screening and maintenance testing for chronic diseases, leading to a decrease in HbA_1c testing of greater than 90% compared with prepandemic screening rates, placing patients at higher risk for diabetes-related complications.⁴ To date, large-scale efforts to enhance HbA_1c testing have focused on testing within the clinic or laboratory and have not systematically assessed nontraditional sites, including at-home testing.^5-7 In this study, we used data from a large US health insurance provider system to determine whether a home HbA_1c test kit intervention increases the frequency of HbA_1c testing and leads to changes in HbA_1c 6 months post testing among members of a large commercial health plan with diagnosed T2D. Further, we assessed whether self-reinforcement education improved maintenance of testing over time.

METHODS

Our study evaluated 2 interventions: a home HbA_1c testing intervention and a self-reinforcement intervention. The home testing intervention included members of a large commercial health plan with T2D in July 2020 who lacked an HbA_1c test in the past 6 months or who had 1 HbA_1c test in the past 6 months but only tested once per 12 months. Members were mailed educational information about the difference between daily blood glucose and HbA_1c as well as CDC guidelines reinforcing the importance of regular monitoring (eAppendix Figure 1 [eAppendix available at ajmc.com]).⁸ Additionally, outreach included either an automatically shipped direct-to-member test kit for members randomly assigned to an opt-out group or instructions on procuring this test kit subsidized by the health plan at no cost for members assigned to an opt-in group. Each fingerstick HbA_1c test kit contained a lancet and blood spotting card for sample collection. Members mailed their sample within 24 hours in a prestamped package to a certified laboratory and received their laboratory results and personalized health recommendations through a secure portal with the recommendation to follow up with their individual provider. Members were allowed at least 6 months to complete testing.

In the self-reinforcement cohort, members who had an HbA_1c test on record in the past year but were not continually testing according to guidelines were sent outreach messaging describing the importance of regular maintenance testing.² This outreach also contained resources for healthy living, including links to health plan mobile applications for achieving health goals, guides on staying physically active, and details on lesser-known plan features.

For both interventions, HbA_1c (%) or the HbA_1c on file 6 months prior to the intervention was compared with control HbA_1c levels 6 months post intervention. Intervention testing rates were compared with control rates after the intervention. Changes in HbA_1c were also stratified by threshold (HbA_1c < 7% and HbA_1c ≥ 7%) to reflect baseline degree of control.⁹ A linear difference-in-differences model with cluster-robust SEs was used to compare testing rates across groups. Analysis of covariance was used to compare HbA_1c reductions in the at-home HbA_1c kit and self-reinforcement intervention groups with those in the control groups by baseline HbA_1c. All statistical tests controlled for baseline HbA_1c, and all results were evaluated on an intent-to-treat basis. Statistical significance was 2-tailed and set at a threshold of P < .05. All analyses were performed using Python version 3.7 (Python Software Foundation). This study was reviewed by the Sterling Institutional Review Board (IRB ID #9460) and deemed exempt. Informed consent was not required due to voluntary waiver of consent among the commercial fully insured population.

RESULTS

The large commercial health plan population included 32,535 members with T2D in July 2020 who lacked an HbA_1c test in the past 6 months (29.1% of members of the population with T2D) and 8679 members who had 1 HbA_1c test in the past 6 months but only tested once per 12 months. Members without an HbA_1c test in the past 6 months were randomly assigned to either a control group (n = 10,908) or an at-home HbA_1c testing intervention group (n = 21,627) consisting of either an opt-in test (n = 17,776) or a direct-to-member opt-out HbA_1c test kit shipment (n = 3851). A total of 4590 at-home HbA_1c test kits were shipped, and home testing completion rates were 25.7% (192 of 748 requested) for the opt-in cohort and 8.2% (316 of 3842) for the opt-out cohort. There were no significant differences in the proportion of members by HbA_1c threshold in each group, and baseline characteristics were similar between groups at the start of the study (eAppendix Table).

Postintervention testing rates and HbA_1c levels are shown in the Table. Overall, there were no differences in HbA_1c testing rates between the home test kit intervention and control groups at 6 months post intervention. However, testing rates were higher in the home kit opt-out group compared with the control group during the study period. At 6 months post intervention, there was a 4.9% increase in HbA_1c testing among opt-out members compared with those in the control group (4.9%; 95% CI, 0.03%-0.07%; P < .001); on the other hand, there was a nonsignificant 0.5% increase in HbA_1c testing among opt-in members compared with those in the control group (0.5%; 95% CI, –0.01% to 0.02%; P = .422).

For members in the home testing intervention cohort, reductions in HbA_1c post intervention were most profound for those with an HbA_1c level of at least 7% (Figure [B-C]). Although there was no change in HbA_1c among the overall population of members who requested and completed at-home HbA_1c testing (opt-in group) vs the control group, opt-in members with an HbA_1c level of at least 7% experienced a 0.38% reduction in HbA_1c on average when controlling for baseline HbA_1c compared with the reduction for the control group (95% CI, –0.58% to –0.19%; P < .001) (Table). There was no change in HbA_1c levels in members in the opt-out HbA_1c test kit group compared with the control group or stratified by HbA_1c threshold.

The self-reinforcement intervention included 8679 members who had 1 HbA_1c test in the past 6 months but only tested once in the past 12 months (intervention: n = 4391; control: n = 4288). The goal of this intervention was to encourage continued testing twice per year (eAppendix Figure 2). HbA_1c testing rates for the self-reinforcement intervention and control groups were similar prior to the intervention (Figure [A]). When controlling for baseline HbA_1c, there was a 0.1% HbA_1c reduction on average between members who received self-reinforcement messaging and control members; this reduction was primarily driven by those with an HbA_1c level of at least 7%, who had a 0.37% HbA_1c reduction on average post intervention (95% CI, –0.67% to –0.07%; P = .015).

DISCUSSION

The early period of the COVID-19 pandemic generated unprecedented challenges in accessing testing sites for laboratory monitoring, creating a need to expand nontraditional testing routes. Our study examined the impact of self-reinforcement messaging and home HbA_1c test kit interventions on HbA_1c testing frequency and diabetes control. Further, we assessed the impact of opt-in vs opt-out HbA_1c test kit interventions on behavior and a change in HbA_1c level 6 months post testing. Among 32,535 members of a large commercial health plan with T2D, we found that sending HbA_1c testing kits at no cost to members (opt out) modestly increased HbA_1c testing rates. Second, we found that sending the members the option to request a free HbA_1c test kit (opt in) or offering self-reinforcement messaging was associated with decreased HbA_1c levels in members with a higher baseline HbA_1c (≥ 7%).

Cumulatively, the results of this study provide novel insights into potential risk stratification and behavioral strategies to optimize engagement as well as downstream health-related behaviors. The modest decrease in HbA_1c levels among the subset of members who were provided self-reinforcement messaging suggests that a simple reminder about screening may be sufficient to prompt behavior change leading to improved T2D control. This finding bears important clinical implications, particularly in the high-risk population of patients with suboptimal T2D control and gaps in HbA_1c monitoring. It is possible that individuals with suboptimal HbA_1c control (≥ 7%) had higher perceived risk regarding their chronic condition and as a result were more likely to change behavior with a single reminder. Although this is speculative, this hypothesis is corroborated by the finding that reductions in HbA_1c post intervention were greatest for those with an HbA_1c level of at least 7% within the home testing group.

As shown in other campaigns, members randomly assigned to the opt-out HbA_1c testing group had higher HbA_1c home testing engagement.10 However, increased testing in this group did not translate to favorable changes in HbA_1c, possibly because individuals in the opt-out group participated in screening due to lack of willingness to opt out rather than an interest in monitoring and controlling their diabetes. In comparison, patients who needed to opt into screening took an active step to obtain home testing. This result may inform both choice architecture and follow-up efforts for future campaigns, as patients participating in an opt-out campaign may need additional resources to support longitudinal behavior change. In addition, it is important to recognize that any potential downstream health outcomes, such as change in HbA_1c level mediated by these interventions, heavily anchor on provider receipt of the home HbA_1c values to facilitate maintenance of clinical value over time and to arrange patient follow-up and treatment. Primary care counseling on medication usage and adherence; lifestyle optimization, including physical activity and nutritional optimization; and concurrent control of comorbid chronic diseases are essential for sustaining these results long term. Nevertheless, future research should explore behavioral strategies to measure and enhance the patient-provider relationship and examine the incremental clinical and cost-savings value of direct-to-provider outreach.

These findings suggest that a novel, low-cost, feasible, and scalable intervention designed to enhance T2D monitoring with at-home HbA_1c testing bears promise as an effective intervention to increase T2D testing rates and facilitate reductions in HbA_1c levels over 6 months in patients with uncontrolled T2D. Notably, this study provided a mechanism for patients with health inertia to test HbA_1c that alleviates common barriers experienced in ongoing monitoring of T2D. From a public health perspective, this intervention enabled disease-specific education, resources, and support to 8679 individuals with T2D across the nation; expanded access to critical T2D monitoring supplies for more than 5000 adults; and closed gaps in care for more than 500 individuals with T2D by facilitating successful completion of home HbA_1c testing. Given the marked reduction in population-level HbA_1c testing of approximately 90% during COVID-19, this study demonstrates that low-cost, low-touch behavioral outreach interventions such those described herein are operationally feasible to effectuate at scale. Future efforts may expand this testing approach to members who may be uninsured or underinsured, among whom substantial barriers to T2D testing currently exist.

Limitations

Our study is limited by the following: First, significant findings related to HbA_1c home testing and associated reductions in HbA_1c levels were modest. It is possible that the suggestion to obtain a test may require longer than 6 months to drive clinically meaningful reductions in HbA_1c levels in some patients.^11,12 To this point, member follow-up is ongoing. Second, individuals with higher baseline HbA_1c levels experienced greater reductions in HbA_1c, which may reflect regression to the mean. However, this is unlikely because analyses controlled for baseline HbA_1c values. Additionally, HbA_1c testing rates returned to a similar baseline 6 months after the postintervention comparison, suggesting that the intervention was effective in driving differences in testing rates during the study window. Finally, this intervention occurred during the peak of the COVID-19 pandemic, which may have contributed to selection bias from members requesting at-home HbA_1c test kits who may have otherwise opted for more traditional testing routes, yielding challenges in accurately controlling for this factor in our observed testing rates and outreach engagement. Regardless, this afforded a unique opportunity to study the impact of quarantine measures and clinic and/or laboratory testing disturbances on utilization of this novel method of delivering HbA_1c testing through a free, convenient pathway.

CONCLUSIONS

A novel intervention to evaluate HbA_1c levels at home may be an effective strategy to increase testing rates and facilitate HbA_1c reduction over time in certain patients with T2D. Home testing bears significant promise in reducing testing disparities due to cost and travel by providing a convenient, safe, and accessible method to both monitor and manage HbA_1c levels.

Acknowledgments

Lauren C. LaMonica, MPH, and Jingwei Ren, MS, contributed equally to this work and are listed as co–first authors.

The authors would like to acknowledge Dr Gui Woolston, Laure Salomon, Rahul Kak, Miki Duruz, Cathy B. Walsh, Dr Dorothea J. Verbrugge, Dr Kelly J. Thomas Craig, and Dr Amanda Zaleski; members of the Next Generation Transform Diabetes Care team; and members of the Aetna Behavior Change Pods for their involvement in the analytics, conceptualization, copy editing, manuscript and marketing development, publication project management, and/or engineering of this HbA1c campaign and for launching initiatives to improve the health of our member population.

Author Affiliations: CVS Health (LCL, JR, NJ, ESG, KBR, PW, SKC), Wellesley, MA; University of Michigan Medical School (LCL), Ann Arbor, MI; Touro College of Osteopathic Medicine (JR), New York, NY; Brigham and Women’s Hospital (NJ, PW), Boston, MA.

Source of Funding: This study was funded by CVS Health Corporation.

Author Disclosures: Ms LaMonica was employed by CVS Health Corporation from July 2019 to August 2021. Ms Ren was employed by CVS Health Corporation and is an ongoing medical student at Touro College of Osteopathic Medicine. Dr Jain was employed by CVS Health Corporation and is now employed at Brigham and Women’s Hospital. Dr Goldberg was employed by CVS Health Corporation ending in 2021 and has patents pending with CVS Health. Dr Rhee was employed by CVS Health Corporation ending in 2022 and owns shares in CVS Health Corporation. Drs Wickner and Chaguturu are employed by CVS Health Corporation and own shares in CVS Health Corporation.

Authorship Information: Concept and design (LCL, JR, NJ, ESG, KBR, PW, SKC); acquisition of data (LCL, JR); analysis and interpretation of data (LCL, JR, NJ, ESG); drafting of the manuscript (LCL, JR, NJ, ESG, SKC); critical revision of the manuscript for important intellectual content (LCL, NJ, ESG, KBR, PW, SKC); statistical analysis (LCL, JR, ESG); administrative, technical, or logistic support (LCL, ESG, PW); and supervision (LCL, ESG, KBR).

Address Correspondence to: Sreekanth K. Chaguturu, MD, CVS Health, 93 Worcester St, Wellesley, MA 02482. Email: sree.chaguturu@cvshealth.com.

REFERENCES

1. National Diabetes Statistics Report: estimates of diabetes and its burden in the United States. CDC. Updated June 29, 2022. Accessed July 19, 2022. https://www.cdc.gov/diabetes/data/statistics-report/index.html

2. American Diabetes Association. 6. Glycemic targets: Standards of Medical Care in Diabetes—2021. Diabetes Care. 2021;44(suppl 1):S73-S84. doi:10.2337/dc21-S006

3. Kiefer MM, Silverman JB, Young BA, Nelson KM. National patterns in diabetes screening: data from the National Health and Nutrition Examination Survey (NHANES) 2005-2012. J Gen Intern Med. 2015;30(5):612-618. doi:10.1007/s11606-014-3147-8

4. Khan T, Kim J, Ortuno A. The Gap Year for Routine Screenings. Komodo Health.February 2021. Accessed July 19, 2022. https://www.komodohealth.com/hubfs/_RESEARCH_BRIEFS/Gap%20Year%20for%20Routine%20Screenings_Feb%202021.pdf

5. Petersen JR, Finley JB, Okorodudu AO, Mohammad AA, Grady JJ, Bajaj M. Effect of point-of-care on maintenance of glycemic control as measured by A1C. Diabetes Care. 2007;30(3):713-715. doi:10.2337/dc06-1909

6. Miller CD, Barnes CS, Phillips LS, et al. Rapid A1c availability improves clinical decision-making in an urban primary care clinic. Diabetes Care. 2003;26(4):1158-1163. doi:10.2337/diacare.26.4.1158

7. Moffet HH, Parker MM, Sarkar U, et al. Adherence to laboratory test requests by patients with diabetes: the Diabetes Study of Northern California (DISTANCE). Am J Manag Care. 2011;17(5):339-344.

8. All about your A1C. CDC. Updated September 30, 2022. Accessed July 19, 2022. https://www.cdc.gov/diabetes/managing/managing-blood-sugar/a1c.html

9. American Diabetes Association. Standards of Medical Care in Diabetes—2022 abridged for primary care providers. Clin Diabetes. 2022;40(1):10-38. doi:10.2337/cd22-as01

10. Mehta SJ, Khan T, Guerra C, et al. A randomized controlled trial of opt-in versus opt-out colorectal cancer screening outreach. Am J Gastroenterol. 2018;113(12):1848-1854. doi:10.1038/s41395-018-0151-3

11. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med. 2008;359(15):1577-1589. doi:10.1056/NEJMoa0806470

12. Reaven PD, Emanuele NV, Wiitala WL, et al; VADT Investigators. Intensive glucose control in patients with type 2 diabetes — 15-year follow-up. N Engl J Med. 2019;380(23):2215-2224. doi:10.1056/NEJMoa1806802