Assessment of Self-monitoring of Blood Glucose in Individuals With Type 2 Diabetes Not Using Insulin

Low-value care worsens patient-centered outcomes and imparts a negative economic effect, which has prompted the Choosing Wisely campaign to promote a national dialogue on the judicious use of services that are deemed to be nonbeneficial. One recommendation is “avoid routine multiple daily self-glucose monitoring in adults with stable type 2 diabetes on agents that do not cause hypoglycemia.”¹ This recommendation is based on robust evidence, including a Cochrane review of 12 randomized clinical trials with more than 3000 patients, showing no statistical difference between patients who do not self-monitor their blood glucose multiple times per day and those who do self-monitor their blood glucose multiple times per day in glycemic control, nor evidence of effects on health-related quality of life, patient satisfaction, or decreased number of hypoglycemic episodes.² The aim of this study was to quantify the rate of use and cost of self-monitoring blood glucose supplies that are potentially used inappropriately, specifically focusing on test strips, the most costly supply for regular blood glucose monitoring.

We performed a retrospective analysis of claims data from January 1, 2013, to June 30, 2015, from the deidentified Clinformatics Data-Mart Database (OptumInsight), which includes commercial health insurance beneficiaries and Medicare Advantage beneficiaries. For inclusion in the analysis, individuals were 18 years or older, not pregnant, continuously enrolled for a minimum of 2 years (January 1, 2013, through December 31, 2014), and experienced any of the following in 2013: two office visits less than 181 days apart, or either 1 emergency department visit or 1 hospitalization with a diagnosis of type 2 diabetes. We identified type 2 diabetes using International Classification of Diseases, Ninth Revision codes 250.X0 or 250.X2 (where X ranges from 0 to 9). Patients were excluded if they received any prescription for insulin. Individuals were followed for at least 1 year after filling a prescription for test strips, or for the calendar year of 2014 if no prescription was filled. Only individuals with 3 or more claims for test strips—suggesting routine use—and those with no claims for test strips within the 1-year study period were included (Table).

University of Michigan Institutional Review Board approval was waived as the data were deidentified. Participant consent was waived as the data were deidentified. All tests were 2-sided and results were deemed statistically significant at P < .05.

A total of 370 740 individuals with type 2 diabetes (182 042 women, 188 662 men, and 36 unknown sex; mean [SD] age, 68.5 [12.1] years) met the inclusion criteria (Figure); 86 747 (23.4%) filled 3 or more claims for test strips during the course of the year. More than half of these individuals (51 820 [14.0% of the study population]) were deemed to be potentially using the supplies inappropriately; 32 773 individuals were taking agents not considered to be a risk for causing hypoglycemia (eg, metformin hydrochloride) and 19 047 had no claims for any antidiabetic medications. These 2 patient groups used a median of 2.0 strips per day (interquartile range, 1.2-2.4). The median claims cost for test strips was $325.54 (interquartile range, $0.00-$534.76) per person per year. The mean (SD) consumer copayment for test strips was $18.14 ($54.94) annually (median, $0.00 [interquartile range, $0.00-$0.00]).

Despite a lack of clinical evidence and being identified as a low-value service by the Choosing Wisely initiative, a substantial percentage of patients with type 2 diabetes may still be inappropriately self-monitoring blood glucose. Nearly 1 in 7 patients with type 2 diabetes not using insulin in this claims database filled 3 or more claims for test strips during the course of 1 year. We acknowledge several limitations within our study. The sample is at risk of bias, as it consists of a selected population with private insurance. Furthermore, we are unable to account for changes in medications or lifestyle as well as transient episodes of hypoglycemia that might warrant monitoring. We also cannot include medications for which a claim was not adjudicated (eg, paying cash for the medication).

These limitations notwithstanding, strategies to improve engagement among clinicians and to educate patients are warranted to reduce low-value care. Clinician-facing strategies might include clinical decision support in electronic medical records that create an alert when ordering test strips for patients who are taking nonhypoglycemic medications. Similar tools have reduced inappropriate testing of vitamin D levels.³ Clinical decision support and interventions such as data feedback, physician communication training, clinician scorecards, and pharmacist review should be aligned with patient-centered strategies such as shared decision making and increased consumer cost sharing (eg, value-based insurance design) to effectively reduce low-value care, while preserving the use of self-monitoring of blood glucose when it is clinically indicated.⁴

Accepted for Publication: August 25, 2018.

Corresponding Author: Kevin D. Platt, MD, Department of Internal Medicine, University of Michigan, 1500 E Medical Center Dr, 3116 Taubman Center 5368, Ann Arbor, MI 48109 (plattk@med.umich.edu).

Published Online: December 10, 2018. doi:10.1001/jamainternmed.2018.5700

Author Contributions: Dr Platt and Mr Lin had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Platt, Thompson, Lin, Linden, Fendrick.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Platt, Thompson, Lin.

Critical revision of the manuscript for important intellectual content: Thompson, Lin, Banerjee, Linden, Fendrick.

Statistical analysis: Platt, Lin, Banerjee, Linden.

Administrative, technical, or material support: Thompson, Lin, Banerjee.

Supervision: Linden, Fendrick.

Conflict of Interest Disclosures: Dr Fendrick reported serving as a consultant for AbbVie, Amgen, FMS Inc, Freedman Health, Geisinger Health Plan, Janssen, Lilly, Mallinckrodt, MedZed, Merck, Risalto, Sanofi, Sempre Health, State of Minnesota, Takeda, TriZetto, Wellth, and Zansors; conducting research for the Agency for Healthcare Research and Quality, CalPERS (California Public Employees’ Retirement System), the Gary and Mary West Health Policy Center, the Laura and John Arnold Foundation, National Pharmaceutical Council, Patient-Centered Outcomes Research Institute, Pharmaceutical Research and Manufacturers of America, Robert Wood Johnson Foundation, State of Michigan/Centers for Medicare & Medicaid Services, and State of New York; serving as coeditor of American Journal of Managed Care; being a member of the Medicare Evidence Development & Coverage Advisory Committee; and being a partner in V-BID Health LLC. No other disclosures were reported.