Evaluation of a Pharmacist-Managed Medication Adjustment Clinic Within an Academic Endocrinology Practice

This was a retrospective observational cohort study evaluating patients with diabetes seen by the pharmacist-led medication adjustment clinic versus those who received usual care at the BBDC adult endocrinology clinic. The study was approved by the University of Kentucky Institutional Review Board.

The cohort seen in the medication adjustment clinic included patients who were 18–99 years of age with diabetes, an A1C >7% at initial (or index) visit, and at least one repeat A1C result measured 1–6 months after the index visit who were referred to and seen within the medication adjustment clinic between 1 February 2019 and 29 February 2020. Any patients seen in the medication adjustment clinic for a nondiabetes-related visits were excluded. The cohort receiving usual endocrinology care included matched patients based on International Classification of Diseases, 10th edition, diabetes code; age (± 5 years); and pre-visit A1C (± 0.5%) who had at least one repeat A1C 1–6 months after the index visit and saw an adult endocrinology provider between 1 February 2018 and 29 February 2020. Patients referred to or seen by the medication adjustment clinic were excluded from the usual endocrinology care cohort.

Cohorts were identified using billing data from Allscripts Patient Manager software. The University of Kentucky’s Center for Clinical and Translational Science matched cohorts in a manner blinded to the authors. Chart review was completed via the Allscripts electronic health record, and data collection was organized within a secured REDCap file. All investigators and supporting staff involved were trained to ensure consistent data collection. The following patient data were collected: age, sex, ethnicity, race, insurance type, weight, height, baseline vital signs, baseline LDL cholesterol, baseline estimated glomerular filtration rate, baseline urine albumin-to-creatinine ratio, baseline comorbidities/complications, diabetes type, duration of diabetes, baseline diabetes medications, pre- and post-index A1C results (± 6 months of index date), number of hypoglycemia episodes per week at baseline and 6 months after the index visit, total endocrinology visits during the 6 months after the index date, and hospital and emergency department (ED) visits at UKHC during the 6 months after index date. All baseline objective information included data closest to or on the index visit date and no more than 1 year before that date. Hypoglycemia occurrences were defined as the frequency of glucose values <70 mg/dL per week based on provider documentation and/or glucose meter reports.

The primary outcome was change in A1C from before to after the index visit. All A1C results within 6 months of the index date were collected, but the A1C closest to 3 months post-index visit was used for the primary outcome if multiple A1C results were measured during this eligibility period. Secondary outcomes included change in A1C at 3 months (including results measured from 1 to 4 months) and 6 months (including results measured from 5 to 8 months) post-index visit; patients achieving an A1C of 7, 8, or 9%; patients with decrease ≥1% in post-index A1C; rates of decreased hypoglycemic frequency; and rates of diabetes-related hospital/ED visits within 6 months post-index visit. Prespecified subgroup analysis of the primary outcome was completed for diabetes type, diabetes duration, sex, insurance type, baseline A1C, and total number of visits within the BBDC.

The authors calculated that 92 patients per group would be needed to detect a 1% difference in A1C change between groups given an SD of 2.4% and using 80% power. Continuous variables were analyzed using independent samples t tests or independent samples difference of medians, as appropriate. Categorical (nominal) variables were analyzed using the Pearson χ² or Fisher exact statistic, as appropriate. Patients with missing data points were excluded from the primary and/or secondary outcomes analyses where data were missing. All statistical analyses was performed using IBM SPSS Statistics, v. 27, software. An independent statistician facilitated blinded analysis of data, as described above.

The BBDC medication adjustment clinic completed a total of 240 encounters with 163 unique patients between 1 February 2019 and 29 February 2020. Inclusion criteria were not met in 54 patients because of missing pre- and/or post-index A1C data (n = 33), baseline A1C <7% (n = 13), and nondiabetes-related reason for the visit (n = 9). A total of 109 patients were included and matched, based on the previously described criteria, within both cohorts.

Baseline characteristics were similar between groups and are summarized in Table 1. The majority of patients had type 2 diabetes (69.7% within each cohort), and mean baseline A1C was 10.1% in medication adjustment cohort and 10.0% in the usual endocrinology care cohort (P = 0.984). Additionally, the majority of patients had a duration of diabetes >20 years and experienced multiple diabetes complications and comorbidities. Baseline diabetes medication therapies were similar between groups, although insulin therapy was used in 100% of patients in the medication adjustment cohort compared with 89% of those in the usual endocrinology care cohort. A higher percentage of patients within the medication adjustment clinic cohort were managed on more complex insulin regimens such as basal-bolus or insulin pump therapy. The only statistically significant difference between groups was a higher percentage of patients experiencing hypoglycemia at baseline within the medication adjustment clinic cohort (51.4 vs. 30.6%, P = 0.002).

As shown in Table 2, change in A1C within 6 months post-index visit was greater within the medication adjustment clinic cohort than in the usual endocrinology care cohort, although this difference did not reach statistical significance (1.27 vs. 0.75%, P = 0.071). Prespecified subgroup analysis of the primary outcome showed statistically significantly greater decreases in A1C within two groups: females and patients with a baseline A1C of 9–12%.

The number of patients with decreased hypoglycemia frequency was statistically higher in the medication adjustment clinic cohort (28.3 vs. 13.5%, P = 0.002). Patients within the medication adjustment clinic cohort experienced statistically fewer diabetes-related hospital and/or ED visits (8.3 vs. 19.3%, P = 0.020). The most common reason for hospitalization within both groups was hyperglycemia-related complications (66.7 vs. 81.0%). Additional reasons for visits included microvascular complications (33.3 vs. 9.5%) and hypoglycemia-related complications (0 vs. 9.5%). Table 3 contains additional subgroup analysis.

This study showed a larger decrease in A1C trending toward significance in endocrinology practice patients referred to and seen within a pharmacist-led medication adjustment clinic. Previous studies have shown improvements in A1C with pharmacist involvement within endocrinology practices in smaller cohorts of patients. Alfayez et al. (12) found significant A1C lowering (8.77 vs. 7.59%, P = 0.040) in a retrospective study of 28 patients with the addition of a pharmacist to an endocrinology interdisciplinary care team. Another study found a significant increase in participants achieving an A1C reduction ≥1% (58 vs. 40.7%, P = 0.041) and those reaching an A1C <8% (32.1 vs. 18.5%, P = 0.047) (13). Although the decrease in A1C in this study was less than in the previous studies, it is important to note that these studies were pre-/post-intervention analyses that lacked a control group, and all had smaller cohorts than the current study.

In comparison with similar studies, improvements in A1C were coupled with statistically lower rates of hypoglycemia and diabetes-related hospital/ED visits in this cohort of medically complex patients in an endocrinology clinic. Because of the increased complexity of diabetes management in endocrinology patients, less stringent A1C goals may be required to minimize the risk of adverse events from medications, including hypoglycemia, and thereby to balance efficacy and safety (14). Because pharmacists are trained to focus on safe medication use, they may be essential members of the diabetes care team for complex patients who are at higher risk for adverse events and complications. Whether in an endocrinology office, a primary care setting, or a community pharmacy, pharmacists can contribute by helping to ensure that patients are working toward attainment of their glycemic goals safely.

Patient populations more likely to benefit from the addition of a pharmacist to the diabetes care team were identified in this study. These key populations include females, individuals with type 2 diabetes, and those with a shorter duration of diabetes; a baseline A1C of 9–12%; hypoglycemia; frequent hospital/ED visits related to diabetes; or less frequent visits to the endocrinology clinic. Interestingly, the medication adjustment clinic cohort saw significantly greater reduction in A1C in the subgroup of patients who were seen less than two times within 6 months, illustrating the impact of a single visit with a pharmacist.

One of the more unexpected subgroup findings included the greater lowering of A1C among females. The literature is mixed with regard to sex differences in diabetes management; some studies have shown females with higher A1C values than their male counterparts (15). This finding could be related to differences in glucose homeostasis, treatment response, or psychological factors (16). Involvement in pharmacist-led medication adjustment services may alleviate psychosocial barriers in females by decreasing stress, increasing social support, and providing closer follow-up. Further exploration of this finding would be beneficial.

These subgroup analysis findings can facilitate a more systematic approach for a health system to determining which patient populations benefit most from receiving care in a pharmacist-led medication adjustment clinic. Population health management is an effective strategy used to proactively manage patient populations with higher-quality care, aiding in the achievement of quality metrics necessary within the pay-for-performance health care model in the United States (17). By focusing pharmacist resources on individuals most likely to achieve improved outcomes, health systems may enhance care for people with diabetes, thereby improving payer reimbursement and ultimately decreasing health care costs.

This study had some limitations. First, a portion of the data collection period occurred was within the coronavirus disease 2019 pandemic, which led to reduced rates of clinic follow-up, fewer documented A1C values, increased patient stress, and impaired lifestyle management (18). All of these factors could have dampened the effect of pharmacist-led intervention, with impaired glucose management overall during the pandemic. Many patients were excluded from the study because of a lack of documented A1C post-index date, likely because of the decreased clinic follow-up. Another limitation was related to electronic health record restrictions. Hospital and ED visits outside of UKHC could not be obtained consistently; therefore, encounters at outside facilities could not be captured. Additionally, rates of hospital/ED visits were not captured before patients’ index date, which may have limited interpretation of those findings. An additional drawback to this study was the possible presence of an inherent selection bias within the clinic referral system. Patients included in the pharmacist-led medication adjustment cohort were referred by their provider because of their likelihood of achieving additional benefit from comprehensive diabetes medication management. Although this process could have resulted in biased cohorts, this referral system is commonly used in pharmacist-led services, providing a practical representation of clinical practice.

In conclusion, this study found that pharmacist involvement in diabetes care through a medication adjustment clinic within an endocrinology practice resulted in a greater decrease in A1C trending toward significance and in decreased rates of adverse events. These findings support the theory that pharmacists can play a vital role on the interdisciplinary diabetes care team in helping patients successfully reach clinical treatment goals and reduce their likelihood of developing diabetes complications.

Emma Payne, a University of Kentucky Doctor of Pharmacy candidate, assisted in data collection.

This project was supported by the National Institutes of Health (NIH) National Center for Advancing Translational Sciences through grant number UL1TR001998. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

No potential conflicts of interest relevant to this article were reported.

M.C.S. researched data, wrote the manuscript, and reviewed/edited the manuscript. K.W.N. researched data and reviewed/edited the manuscript. A.D.S. completed the statistical analysis and contributed to the methods and results sections of the manuscript. K.W.N. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

This research was presented as a poster at the Vizient Pharmacy Network Meeting Virtual Conference in December 2020.