Purpose: To determine whether and how diabetes-trained mid-level clinicians (2 nurse practitioners and a physician assistant) could effectively improve HbA1c levels in poorly controlled, minority patients.

Patients and Methods: In a retrospective, observational study using de-identified data (therefore not requiring IRB approval), 96 patients fulfilled the inclusion criteria of initial Hb A1c levels ≥64 mmol/mol (8.0%), at least 2 visits and 2 HbA1c tests and not documented as non-adherent when treated. The primary outcome was the change in HbA1c levels from baseline. The secondary outcome was the proportion of patients in 1% HbA1c ranges from 97 mmol/mol (11.0%), both outcomes compared between initial and final visits.

Results: Patients were treated for a mean of 7 months before the Covid-19 pandemic stopped in-person visits. Baseline HbA1c levels (mmol/mol) measured at referral fell from 97 mmol/mol (11.0% ± 1.8 [SD]) to 68 mmol/mol (8.4% ± 1.8 [P <10-21]). Treatment intensification did not occur, strongly suggesting that increased medication adherence was mainly responsible. At the last visit, 96% of patients met the American Diabetes Association’s blood pressure target of <140/90 mm Hg and 57% the LDL cholesterol target of <2.6 mmol/L (100 mg/dl).

Conclusion: These culturally competent mid-level clinicians (2 African-Americans and one Latino), were better able to engage patients which greatly improved their medication adherence. Busy primary care practices should consider supporting diabetes-trained mid-level clinicians to interact with all patients with diabetes at important treatment junctures (not just the limited number assigned to them for total care). An added benefit for practices serving a minority, population would be culturally competent mid-level clinicians

Medication non-adherence; Diabetes-trained midlevel clinicians; Cultural competency; Diabetes control; Minority under-resourced populations

Davidson MB, Duran P, Bandla M, Rios E (2022) Role of Medication Adherence in Diabetes Treatment by Mid-Level Clinicians. Ann Nurs Pract 9(2): 1127.

The American Diabetes Association (ADA), recommends the HbA1c goal for most people with diabetes to be 64 mmol/mol (8.0%), while 16% exceed 75 mmol/mol (9.0%) [2]. African-American [3], and Latino [4,5], patients with diabetes have significantly higher HbA1c levels than their white counterparts. This leads to worse outcomes of diabetes care in African-American [3,6,8], and Latino [4,5,7], patients compared with whites.

Ninety percent of people with diabetes are cared for by primary care physicians (PCPs) who may be inexperienced in using the more recent FDA cleared classes of new drugs and are particularly challenged in using insulin. The latter is evidenced [9], by; a) medians of up to 7 years before starting insulin once people with type 2 diabetes failed (HbA1c levels >64 mmol/mol [8.0%]) maximal (tolerated) doses of 2 or 3 non-insulin drugs, b) mean HbA1c levels ranging from 74 mmol/mol (8.9%) to 84 mmol/mol (9.8%) with a mean of 78 mmol/mol (9.3%) when insulin was started, (c) mean HbA1c levels of 83 mmol/mol (9.7%) when insulin was intensified in patients failing basal insulin alone, (d) insulin intensification occurring in only 25- 30% of patients and its discontinuation in a similar number, and e) mean HbA1c levels ranging from 63 mmol/mol (7.9%) to 78 mmol/mol (9.3%) with a mean of 69 mmol/mol (8.5%) in patients receiving “stable” doses of insulin.

A number of studies have demonstrated that diabetes-trained registered nurses (RNs), nurse practitioners (NPs), physician assistants (PAs) and clinical pharmacists (PharmDs) following treatment protocols have substantially improved diabetes outcomes [10].These providers are termed mid-level clinicians here because they are allowed to make independent treatment decisions, either under the supervision of physicians (RNs, PharmDs) or within the scope of their licensed practices (NPs, PAs). For nearly 40 years, the first author has participated in this approach to improving diabetes care by teaching his detailed treatment protocols to 44 mid-level clinicians, the majority of whom provided diabetes care to patients from under-resourced populations. The recently described definition of an underresourced population is one that is characterized by a majority of minority, poor and poorly-educated individuals that have difficulty accessing medical care and experience worse medical outcomes [11].

A large under-resourced population lives in central Los Angeles whose medical care is provided by Los Angeles County in the Martin Luther King (MLK) Outpatient Center. However, it should be noted that the difficulty in accessing medical care would not apply to the under-resourced patients in this study because they were receiving medical care in a primary care clinic (albeit inadequate from a diabetes perspective). A Diabetes Clinic was established in the MLK Outpatient Center to which complicated, poorly controlled patients could be referred for independent treatment by mid-level clinicians (in the present study 2 NPs and a PA). Only an occasional consultation was sought from the first author for particularly difficult situations.

This paper evaluated whether these mid-level clinicians independently following detailed treatment protocols could overcome the barriers faced by PCPs and improve diabetes outcomes in these challenging patients. It also sought to examine which specific changes in the treatments for these newly referred patients might be more effective in lowering HbA1c levels.

The criteria for referral by PCPs to the Diabetes Clinic is an HbA1c level ≥64 mmol/mol (8.0%), in patients receiving either 2 or more classes of non-insulin drugs (without insulin) or taking insulin (regardless of the number of non-insulin drugs). Starting in February, 2019, monthly reports of each patient visit and their latest HbA1c level were provided. In March, 2020, most clinic visits were suspended because of the Covid-19 pandemic and patients were managed by telephone. Since patients were reluctant to come to the laboratory for HbA1c (and other) tests because of the pandemic, the population evaluated in this study were those seen as new referrals from the beginning of February, 2109 until the end of September, 2019 and followed until the end of February, 2020. The kinetics of HbA1c level changes are such that values reflect the average glycemia of the previous 3 months. The termination of recruitment for the study at the end of September, 2019 allowed all patients to have at least two 3-month cycles of HbA1c changes by February, 2020.

The inclusion criteria were patients with first visits to the Diabetes Clinic after January, 2019, with initial HbA1c levels ≥64 mmol/mol (8.0%), having ≥2 visits, at least 2 HbA1c tests and not being identified by the mid-levels as “non-adherent.” The latter was recorded only after 3 or more reminders (and documentation), that patients were not taking prescribed medications correctly, or for those on insulin, not providing enough glucose readings to adjust insulin doses. If a patient failed an appointment, at least 3 attempts were made to re-schedule them. Patients documented as “non-adherent” were excluded to allow a valid cohort in which to accurately determine which treatments might be more effective in improving diabetes control.

The primary outcome was change in HbA1c levels between initial and final visits. Given the kinetics of HbA1c changes, baseline HbA1c values were accepted within 2 months before and up to one month following the initial visit and the final HbA1c values were accepted 1 month before and up to 3 months after the last visit. After the initial visit, HbA1c levels were usually measured approximately every 3 months. A secondary outcome was the proportion of the final HbA1c levels in the following ranges; <53 mmol/mol (7.0%), 53 -63 mmol/mol (7.0-7.9%), 64- 74 mmol/mol (8.0 – 8.9%), 75-85 mmol/mol (9.0 – 9.9%), 86-96 mmol/mol (10.0 – 10.9%) and ≥97 mmol/mol (11.0%).

Good diabetes care requires meeting blood pressure (BP) and low density lipoprotein (LDL) cholesterol targets. Therefore, another secondary outcome was the percent of patients who met the ADA’s [12], BP and LDL cholesterol goals of <140/90 mm Hg and <2.6 mmol/L (100 mg/dl), respectively, at study end. Because patients often wish to limit their trips to the clinic, they sometimes delay refilling their prescriptions until a scheduled visit with a provider. In that case, their BP may be high at the visit. If an elevated BP were recorded at the last visit, the value at the previous visit was accessed. If it were <140/90 mm Hg, the patient was considered to have met the ADA BP goal since BPs rapidly increase if anti-hypertensive medications are missed. Although the delay in renewing prescriptions would also apply to statin therapy, the relationship between LDL cholesterol concentration changes and not taking this medication plays out over a much longer period than BP changes. Consequently, the LDL cholesterol concentration nearest to the last visit was used.

The study was approved by the Institutional Review Board at Charles R. Drew University. Because it was a retrospective, observational one and the data were de-identified, informed consents were not required. Since the patients were cared for in the Department of Health Services of Los Angeles County, their Institutional Review Board also had to confirm that informed consents were not required which they did. A 2-tailed paired Student’s t test was used to analyze differences in HbA1c levels between the first and last visits. Significance was accepted at the P <0.05 level.

The number of patients referred to the Diabetes Clinic, those evaluated after their last visit and the reasons why all the referred patients were not evaluated are shown in Figure 1.

Figure 1 Flow diagram from patient referral to evaluable patients.

The 96 evaluable patients were 51.5 ± 11.4 (SD) years old, 47 females, 90 with type 2 diabetes, had diabetes for 11.0 ± 8.5 years, were 84% Latino and 11% African-American and with BMIs of 31.5 ± 8.4. The mid-level clinicians followed the patients for 35 to 393 days with a mean of 212 days (7 months).

The changes in the outcomes between the first and last visits are shown in Table 1.

The HbA1c level fell by 29 mmol/ mol from 97 to 68 mmol/mol (a 2.6% fall from 11.0% to 8.4%). By definition, no patients at the initial visit had an HbA1c level <64 mmol/mol (8.0%) while 66% had values ≥86 mmol/mol (10.0%). At the last visit, 42% had HbA1c levels <64 mmol/ mol (8.0%) with 24% meeting the ADA goal of <53 mmol/mol (7.0%) while only 18% had values ≥86 mmol/mol (10.0%). Strikingly, this marked improvement occurred in the absence of pharmacological treatment intensification. For instance, 7% fewer patients were taking insulin at the last visit compared with the first one and only 8 patients not receiving insulin increased their number of non-insulin drugs. The very few patients on more than 3 non-insulin drugs at the last visit reflect the limitation Los Angeles County has placed on the 3 newer classes of them, e.g., DPP-4 inhibitors, GLP-1 agonists and SGLT-2 inhibitors. The patients gained 3.5 kg in the 7 months of the program. At the end of the study, 96% met the ADA BP goal and 57% met the LDL cholesterol goal.

Diabetes control markedly improved under the care of the 2 NPs and the PA over a mean of 7 months. HbA1c levels fell by 29 mmol/mol (2.6%). Twenty-four percent of the patients met the ADA HbA1c level goal of <53 mmol/mol (7.0%) and 42% achieved HbA1c levels of <64 mmol/mol (8.0%). This occurred in the absence of additional pharmacological treatment strongly suggesting that increased medication adherence was responsible for the improvement. Patients’ weight did increase significantly by 3.5 kg but this is commonly seen when poorly controlled patients are brought under better control with most of the weight gain due to the storage of glucose calories previously lost in the urine [13], Before the pandemic precluded in-person clinic visits, a remarkable 96% of the patients met the ADA BP goal of <140/90 mm Hg. Fifty-seven percent of these patients achieved the ADA LDL cholesterol goal of <2.6 mmol/L (100 mg/dl) compared to 36% in a large number of patients with diabetes cared for in 6 Public Hospitals that serve a similar under-resourced population as does the MLK Outpatient Center [14].

Medication non-adherence is an important issue in outcomes of medical care. It is usually assessed by examining large pharmacy databases and calculating the medication possession ratio in retrospective observational studies [15]. Values over 80% are considered necessary to achieve the full benefit of medication in chronic diseases [16]. Medication non-adherence is very common in people with diabetes [17,18]. Although medication non-adherence could not be directly documented in this retrospective, observational study, its results provided a unique way to evaluate medication non-adherence, ie, by a marked improvement in glycemia in the absence of increased prescribed medication. Medication non-adherence leads to worsening diabetes control over time [19], increased vascular complications [20-22], and increased all-cause mortality, medical care costs, emergency department visits and hospitalizations [23- 25].The disappointing effects of new medications in real world experience compared with the promising results of randomized control trials are largely due to medication non-adherence [18,26].

It is well documented that the outcomes of diabetes care are worse in African-American [3,6,8], and Latino [4,5,7], patients compared with whites. The 24% of referred patients who were eligible for treatment but never returned for a second appointment in spite of at least 3 attempts to re-schedule them (Figure 1) illustrate the challenges of caring for the minority patients served by Los Angeles County. The 67% who persisted in their treatment by the mid-level clinicians and received at least 2 HbA1c tests achieved a marked glycemic improvement because of increased medication adherence which was likely enhanced by the cultural competency of these mid-level clinicians (2 AfricanAmericans and a Latina), who were from the same minority populations as those they are caring for. Cultural competence can be defined as the understanding of cultural and linguistic barriers that ethnic and racial minority patients may be facing in achieving quality healthcare. Culturally competent diabetes interventions significantly improved clinical outcomes [7,27]. Within the additional time focusing exclusively on diabetes care, the NPs and PA were better able to engage with the patients in their diabetes care. Being more invested in their health, the patients built partnerships with these mid-level clinicians to seek better longer-term clinical outcomes. Unfortunately, cultural competency training of primary care clinicians did not improve clinical outcomes [28].

It is unrealistic to expect busy primary care practices to support a diabetes-trained mid-level clinician who is assigned the total care of patients, only a limited number of whom may have diabetes. Rather, a diabetes-trained mid-level clinician, either working directly under the supervision of PCPs or under approved diabetes protocols (RNs, PharmDs) or within the scope of their licensed practices (NPs, PAs), could improve diabetes care throughout the practice. An important reason for poor outcomes of diabetes care is the relatively infrequent interactions between patients with diabetes and their primary care clinicians [29]. Patients, especially those from under-resourced populations, are usually seen only every 3 to 6 months which can lead to problems with medication adherence, missed visits and laboratory tests and long intervals in which intensification of therapy may have been appropriate but could not occur. Some examples of how a diabetes–trained, mid-level clinician could improve outcomes would be periodically reviewing the electronic health record to assure ongoing care as well as necessary LDL cholesterol and HbA1c testing, both of which would help to monitor medication adherence; visits for BP monitoring by the mid-level clinician with appropriate medication changes if necessary; ensuring that FPG tests are carried out when used to titrate non-insulin drugs with doses appropriately increased if warranted; and obtaining, organizing and reviewing glucose meter readings (from patients at visits or from home), for possible medication adjustments. Ongoing frequent interactions should allow relationships and partnerships between mid-level clinicians and patients that will result in better clinical outcomes. An added benefit for practices that serve an under-resourced population would be a culturally competent mid-level clinician.

There are several limitations to this study. The absence of a control group introduces the possibility of unrecognized confounding. Because of its retrospective, observational nature, medication non-adherence could not be directly measured but only inferred in light of the marked improvement in glycemia in the absence of pharmacological treatment intensification. A longer period of observation that was unfortunately limited by the pandemic might have produced even more of an improvement in glycemia. As with all studies that evaluate diabetes control, examination of the microvascular complications per se requires much longer periods of follow up. However, since these patients had probably been poorly controlled for many years, these complications were very likely to have been present at the time of referral to the Diabetes Clinic.

In conclusion, this retrospective, observational study confirms the effectiveness of mid-level clinicians to improve diabetes outcomes. The marked glycemic improvement in these poorly controlled patients with diabetes from an under-resourced population was most likely due to increased medication adherence as there was no pharmacological treatment intensification. It is suggested that employing a diabetes-trained mid-level clinician in busy primary care practices would result in more frequent interactions with all patients with diabetes, markedly improve their diabetes outcomes and free up time for PCPs to attend to other demanding issues.

The authors gratefully acknowledge Apollonia Godoy, NP, Pandora McDaniel, NP and Christine Turner, PA for their devoted efforts on behalf of their patients. Mohit Bandli and Esteban Rios each gathered and analyzed the data on one-half of the patients which they wrote up for their Masters of Science theses. Petra Duran taught and supervised the students on extracting data from the Electronic Health Record and entering and organizing it into an Excel database.

The authors have no conflict of interests. The corresponding author has received permission to list the others in the Acknowledgments.

1. American Diabetes Association. Glycemic targets: Standards of Medical Care in Diabetes – 2021. Diabetes Care. 2021; 44: S73-S84.

2. Carls G, Huynh J, Tuttle E, Yee J, Edelman SV. Achievement of glycated hemoglobin goals in the US remains unchanged through 2014. Diabetes Care. 2017; 8: 863-873.

3. Kirk JK, D’Agostino RB, Bell RA, Passmore LV, Bonds DE, Karter AJ, et al. Disparities in HbA1Clevels between African-American and nonhispanic white adults with diabetes” a meta-analysis. Diabetes Care. 2006; 29: 2130-2136.

4. Gallegos-Macias AR, Macias, SR, Kaufman E, Skipper B, Kalishman N. Relationship between glycemic conrol, ethnicity and socioeconomic status in Hispanic and white non-Hispanic youths with type 1 diabetes mellitus. Pedatric Diabetes. 2003; 4: 19-23.

5. Getaneh A. Light, LS, Brillon DJ, Escandon JC, Felicetta J, Ewans GW, et al. Diabetes control among Hispanics in the Action to Control Cardiovascular Risk in Diabetes Trial. J Gen Intern Med. 2012; 27: 1499-1505.

6. Parrinello CM, Rastegar I, Godino JG, Miedema MD, Matsushito K, Selvin E. Prevalence of and racial disparities in risk factor control in older adults with diabetes: the Atherosclerosis Risk in Communities Study. Diabetes Care. 2015: 38: 1290-1298.

7. Peek ME, Cargill A, Huang ES. Diabetes health disparities: a systemic review of health care interventions. Med Care Res Rev. 2007; 64: 101S-156S.

8. Ayanian JZ, Landon BE, Newhouse JP, Zaslavsky AM. Racial and ethnic disparities among enrollees in Medicare Advantage Plans. N Engl J Med. 2014; 371: 2288-2297.

9. Davidson MB, Davidson, SJ. Use of computerized insulin dose adjustment algorithms to facilitate adjusting insulin doses by primary care providers. Medical Research Archives. 2021; 9(2). doi: https:// doi.org/10.18103/mra.v9i2.2335.

10.Davidson MB. Improving diabetes care: a personal journal. On the Cutting Edge: Diabetes Dietetic Practice Group. 2022; 42: 8-11.

11.htpps://icic.org > wp-content > uploads > 2020/10 pdf.

12. American Diabetes Association. Cardiovascular disease and risk management: Standards of Medical Care – 2021. Diabetes Care. 2021; 44: S125-S150.

13. Carlson MG, Campbell PJ. Intensive insulin therapy and weight gain in IDDM. Diabetes. 1993; 42: 1700-1707.

14. Chew LD, Schillinger D, Maynard C, Lessler DS, Consortium for Quality Improvement in Safety Net Hospitals. Glycemic and lipid control among patients with diabetes at six U.S. public hospitals. J Health Care Poor Underserved. 2008; 19:1060-1075.

15. Hess LM, Raebel MA, Connor D, Malone DC. Measurement of adherence in pharmacy administrative databases: a proposal for standard definitions and preferred measures. Ann Pharmacother. 2006; 40: 1280-1288.

16. Karve S. Cleves MA, Helm M. Hudson TJ, West DS, Martin BC. Good and poor adherence: optimal cut-point for adherence measures using administrative claims data. Curr Med Res Opin. 2009; 25: 2303-2310.

17. Cramer JA. A systematic review of adherence with medications for diabetes. Diabetes Care. 2006; 27: 121812-12124.

18. Edelman SV, Polonsky WH. Type 2 diabetes in the real world: the elusive nature of glycemic control. Diabetes Care. 2017; 40: 1425- 1432.

19. Egede LE, Gebregziabher M, Echois C. Lynch CP. Longitudinal effects of medication nonadherence on glycemic control. Ann Pharmacother. 2014; 48: 562-570.

20. Gibson TB, Song X, Alemayehu B, Wang SS, Waddell JL, Bouchard JR, et al. Cost sharing, adherence, and health outcomes in patients with diabetes. Am J Manag Care. 2010; 16: 589-600.

21. Simpson AH, Lin M, eurich DT. Medication adherence affects risk of new diabetes complications: a cohort study. Ann Pharmacother. 2016; 50: 741-746.

22. Beernink JM, Oosterwijk MM, Khunti K, Gupta P, Patel P, van Boven JFM, et al. Biochemical urine testing of medication adherence and its association with clinical markers in an outpatient population of type2 diabetes patients: analysis in the Diabetes and LifE style Cohort Twente (DIALECT). Diabetes Care. 2021; 44: 1419-1425.

23. Kirkman MS, Rowan-Martin MT, Levin R, Fonseca VA, Schmittdiel JA, Herman WH, et al. Determinants of adherence to diabetes medications: findings from a large pharmacy claims database. Diabetes Care. 2015; 38: 604-609.

24. Polonsky WH, Henry RR. Poor medication adherence in type 2 diabetes: recognizing the scope of the problem and its key contributors. Patient Prefer Adherence. 2016; 10: 1299-1307.

25. Khunti K, Seidu S, Kunutsor, Davies M. Association between adherence to pharmacotherapy and outcomes in type 2 diabetes: a meta-analysis. Diabetes Care. 2017; 40:1588-1596.

26. Carls GS, Tuttle E. Tan RD, Huynh J, Yee J, Edelman SV, et al. Understanding the gap between efficacy in randomized controlled trials and effectiveness in real-world use of GLP-1 RA and DPP4-4 therapies in patients with type 2 diabetes. Diabetes Care. 2017; 40: 1469-1478.

27. Attridge M, Creamer J, Ramsden M, Cannings-John R, Hawthorne K. Culturally appropriate health education for people in ethnic minority groups with type 2 diabetes mellitus. Cochrane Database Syst Rev. 2014; 4: CD006424.

28. Sequist TD, Fitzmaurice GM, Marshall R, Shaykevich S, Marston A, Safran DG, et al. Cultural competency training and performance reports to improve diabetes care for black patients: a cluster randomized, controlled trial. Ann Intern Med. 2010; 152: 40-46.

29. Pantalone KM, Misra-Hebert AD, Hobbs TM, Kong SX, Ji X, Ganguly R, et al. The probability of A1C goal attainment in patients with uncontrolled type 2 diabetes in a large integrated delivery system: a prediction model. Diabetes Care. 2020; 43: 1910-1919.