Endocrinology/Primary Care Pharmacy Collaboration vs Endocrinology Care Alone in Patients With Type 2 Diabetes and A1c >9.

Background: Type 2 diabetes (T2D) requires close collaboration between patients and their care management team, often including endocrinology. Primary care pharmacist impact on diabetes management in collaboration with endocrinology is not well established. Objective: To assess if pharmacy and endocrinology collaboration results in a greater A1c reduction in patients with T2D vs endocrinology alone. Methods: This retrospective, observational cohort study was conducted in adult outpatients with T2D and baseline A1c >9% who saw endocrinology within 1 year preceding the study period (January 1, 2021 to January 1, 2022). Patients were included if they had a follow-up A1c 6 months (±90 days) from index date and completed at least 1 endocrinology visit during the study period. Patients managed by endocrinology/primary care pharmacist collaboration (Endo/PharmD) were compared with those who received endocrinology care alone (Endo). Primary outcome was change in A1c from baseline to 6 months. Secondary outcomes included total number of completed visits and percentage of patients achieving A1c <6.5%, <7%, <8%, and <9% between groups at 6 months. Results: A total of 418 patients were included (22 Endo/PharmD, 396 Endo). The change in follow-up A1c was not significantly different between groups, -0.481% (standard error [SE] = 0.396); P = 0.6179. Endo/PharmD patients had significantly more provider visits during the study period (5.3 ± 2.3 vs 2.3 ± 1.2; P < 0.001). No significant difference was observed in odds of A1c goal attainment between groups at 6 months. Conclusion and Relevance: Endocrinology/primary care pharmacist collaboration occurred infrequently but was associated with a trend toward greater A1c reduction in patients with T2D and A1c >9%.

Evaluation of pharmacist impact on diabetes outcomes for telehealth and hybrid care delivery versus in-office visits.

BACKGROUND: Few studies have examined the effect of pharmacist-led telemanagement on diabetes outcomes during the COVID-19 pandemic. OBJECTIVES: Assess for noninferiority for the absolute change in mean A1C between telehealth and hybrid groups versus the in-office group during the COVID-19 pandemic. Secondary objectives were to compare the percentage of patients achieving population health A1C goals and patient no-show rates between study groups. METHODS: A retrospective, noninferiority analysis was conducted for patients seen by a primary care pharmacist from November 1, 2020 to May 31, 2021 across 17 primary care clinics in the Northeast Ohio region of Cleveland Clinic. The noninferiority margin was prespecified at > 0.3% A1C reduction. Patients with a baseline A1C of 8% or greater were included. Patients were separated into 3 study groups (telehealth, in-office, and hybrid) based on the visit types that were conducted by the pharmacist during the study period. RESULTS: Hybrid care delivery (N = 366) was noninferior to in-office care delivery (N = 180), with regards to absolute change in mean A1C reduction (0.24% [95% CI: -0.13, 0.61], P = 0.002). Similar results were shown when comparing the telehealth group (N = 691) to the in-office group (0.04 [95% CI: -0.28, 0.36], P = 0.02). The mean A1C reduction in the in-office (1.36 ± 1.9), hybrid (1.60 ± 2.2), and telehealth (1.40 ± 2.0) groups were not significantly different (P = 0.23). Subgroup analyses showed that newly consulted patients had a larger reduction in A1C compared to the overall population, in all groups. No-show rates and percentage of patients achieving population health A1C goals were not significantly different based on visit type. CONCLUSION: Telehealth and hybrid visit types were noninferior to in-office visits with regards to mean change in A1C reduction. Results demonstrate the importance of primary care pharmacists continuing to offer diverse visit types based on patient preference.

Thromboembolic and Major Bleeding Events With Rivaroxaban Versus Warfarin Use in a Real-World Setting.

BACKGROUND: Although randomized trials demonstrate the noninferiority of rivaroxaban compared with warfarin in the context of nonvalvular atrial fibrillation (AF), little is known about how these drugs compare in practice. OBJECTIVE: To assess the relative effectiveness and safety of rivaroxaban versus warfarin in a large health system and to evaluate this association by time in therapeutic range (TTR). METHODS: We conducted a retrospective cohort study with propensity matching in the Cleveland Clinic Health System. The study included patients initiated on warfarin or rivaroxaban for thromboembolic prevention in nonvalvular AF between January 2012 and July 2016. The main outcomes were thromboembolic events and major bleeds. Analyses were stratified by warfarin patients' TTR. RESULTS: The cohort consisted of 472 propensity-matched pairs. The mean age was 73.6 years (SD = 11.7), and the mean CHADS2 score was 1.8. The median TTR for warfarin patients was 64%. In the propensity-matched analysis, there was no significant difference in thromboembolic or major bleeding events between groups. Among warfarin patients with a TTR <64% and their matched rivaroxaban pairs, there was also no significant difference in thromboembolic or major bleeding events. CONCLUSIONS: Under real-world conditions, warfarin and rivaroxaban were associated with similar safety and effectiveness, even among those with suboptimal therapeutic control. Individualized decision making, taking into account the nontherapeutic tradeoffs associated with these medications (eg, monitoring, half-life, cost) is warranted.

Development of an Ambulatory Clinical Pharmacy Prioritization Prediction Model for Patients With Diabetes.

BACKGROUND: Health care reform and the projected shortage of primary care physicians necessitate more efficient use of multidisciplinary collaboration. No studies, to date, have evaluated factors associated with treatment success among patients referred to a clinical pharmacist. OBJECTIVE: To develop a prediction model using patient factors to assist in selecting patients with diabetes most likely to benefit from pharmacy interventions. METHODS: A retrospective, nested case-control study was performed. A prediction model using multivariable logistic regression was developed, with model calibration and internal validation. Adult patients with diabetes who had a baseline hemoglobin A1C (A1C) ≥9%, who were consulted for collaborative pharmacy care between July 2009 and July 2014 were included. Success (cases) was defined as a decrease in A1C by 2% or a value of A1C <8% 1 year after the initial visit. Failures were those who did not achieve the A1C goal or were lost to follow-up. RESULTS: A total of 544 unique patients were included, with 243 (44.7%) classified as clinical successes and 301 as clinical failures. Independent factors associated with success included past medical history of cerebrovascular accident and higher baseline A1C, whereas use of short-acting insulin and higher number of classes of diabetic medications at baseline corresponded with failure. A prediction model for success using these factors had an optimism-adjusted C-statistics of 0.629, with good calibration. CONCLUSION: Referred patients with diabetes have baseline characteristics that are predictive of clinical success. A predictive model based on those factors performed well and might be utilized to improve pharmacist efficiency in the face of constrained resources.

Predicting hypertension control using machine learning.

Hypertension is a widely prevalent disease and uncontrolled hypertension predisposes affected individuals to severe adverse effects. Though the importance of controlling hypertension is clear, the multitude of therapeutic regimens and patient factors that affect the success of blood pressure control makes it difficult to predict the likelihood to predict whether a patient's blood pressure will be controlled. This project endeavors to investigate whether machine learning can accurately predict the control of a patient's hypertension within 12 months of a clinical encounter. To build the machine learning model, a retrospective review of the electronic medical records of 350,008 patients 18 years of age and older between January 1, 2015 and June 1, 2022 was performed to form model training and testing cohorts. The data included in the model included medication combinations, patient laboratory values, vital sign measurements, comorbidities, healthcare encounters, and demographic information. The mean age of the patient population was 65.6 years with 161,283 (46.1%) men and 275,001 (78.6%) white. A sliding time window of data was used to both prohibit data leakage from training sets to test sets and to maximize model performance. This sliding window resulted in using the study data to create 287 predictive models each using 2 years of training data and one week of testing data for a total study duration of five and a half years. Model performance was combined across all models. The primary outcome, prediction of blood pressure control within 12 months demonstrated an area under the curve of 0.76 (95% confidence interval; 0.75-0.76), sensitivity of 61.52% (61.0-62.03%), specificity of 75.69% (75.25-76.13%), positive predictive value of 67.75% (67.51-67.99%), and negative predictive value of 70.49% (70.32-70.66%). An AUC of 0.756 is considered to be moderately good for machine learning models. While the accuracy of this model is promising, it is impossible to state with certainty the clinical relevancy of any clinical support ML model without deploying it in a clinical setting and studying its impact on health outcomes. By also incorporating uncertainty analysis for every prediction, the authors believe that this approach offers the best-known solution to predicting hypertension control and that machine learning may be able to improve the accuracy of hypertension control predictions using patient information already available in the electronic health record. This method can serve as a foundation with further research to strengthen the model accuracy and to help determine clinical relevance.

Prevalence of Sodium Glucose Cotransporter 2 (SGLT-2) Inhibitor Prescribing in Patients with Type 2 Diabetes Mellitus and Reduced Estimated Glomerular Filtration Rate.

Sodium glucose cotransporter 2 (SGLT-2) inhibitors have demonstrated benefit in people with type 2 diabetes mellitus (T2DM) and chronic kidney disease (CKD), including slowing the progression of CKD and lowering the risk of kidney failure and death. Despite this evidence, literature suggests SGLT-2 inhibitors are underutilized in this population. To assess prescribing practices and identify potential variables predictive of SGLT-2 inhibitor prescribing, a non-interventional, retrospective, cross-sectional study was conducted in patients with T2DM and reduced estimated glomerular filtration rate (eGFR). The primary outcome compared prevalence of SGLT-2 inhibitor prescribing in patients with T2DM and eGFR of 30-44 mL/min/1.73m2 to patients with T2DM and eGFR 45-59 mL/min/1.73m2. The secondary outcome described possible predictors of prescribing SGLT-2 inhibitors in this population. Of the 9,387 patients identified with T2DM and reduced eGFR, an SGLT-2 inhibitor was prescribed to 324 (12.2%) patients with eGFR of 30-44 mL/min/1.73m2 versus 799 (11.9%) patients with eGFR of 45-59 mL/min/1.73m2. Patients more likely to be prescribed SGLT-2 inhibitors were younger, male, had a higher body mass index (BMI), a higher hemoglobin A1c (HbA1c), were on other antihyperglycemic medications, had concomitant cardiovascular disease, or had concomitant heart failure. This study found no significant difference in prevalence of SGLT-2 inhibitor prescribing between patients with T2DM and eGFR 30-44 mL/min/1.73m2 versus eGFR 45-59 mL/min/1.73m2 (p=0.70). Further exploration into the causes of low SGLT-2 inhibitor prescribing prevalence is warranted given the growing evidence supporting the use of these agents in patients with T2DM and reduced renal function.

Patient and physician factors contributing to polypharmacy among older patients.

OBJECTIVE: Polypharmacy, or use of multiple medications, is associated with patient factors. Less is known regarding variation in polypharmacy by individual physicians. The objective of this study was to assess patient and physician factors associated with polypharmacy among older patients. METHODS: This is a cross-sectional study of patients aged ≥65 years with a primary care visit at Cleveland Clinic Health System in 2015 and their physicians. We collected patient demographics, comorbidities and current medications from the electronic health record, including potentially inappropriate medications (PIMs). We used mixed effects linear regression to estimate adjusted differences in the number of medications by patient factors. We generated adjusted prescribing rates for individual physicians and assessed differences in physician performance on quality measures by their prescribing rate. RESULTS: Our study included 44,570 patients who were prescribed an average of 6.8 medications (standard deviation: 4.0) by 701 physicians. Female sex, higher BMI, having Medicaid insurance, current or former smoking status, comorbidities and seeing a specialist were associated with number of medications. Age was not. Among 267 physicians who saw ≥20 study-eligible patients, the adjusted mean number of medications per patient ranged from 5.2 to 9.6. Compared to physicians who prescribed above the mean, lower prescribing physicians performed significantly better on medication reconciliation (p = .007) and hypertension control (p < .001) and prescribed fewer PIMs (p < .001). CONCLUSIONS: Individual physicians varied in their prescribing practices, even after adjusting for patient demographic and clinical characteristics. Interventions to reduce polypharmacy in older adults should target high prescribing physicians, as physician behavior is more actionable than patient factors.

Starting insulin in patients with type 2 diabetes: An individualized approach.

Because type 2 diabetes mellitus is a progressive disease, most patients eventually need insulin. When and how to start insulin therapy are not one-size-fits-all decisions but rather must be individualized. This paper reviews the indications, goals, and options for insulin therapy in type 2 diabetes.