RESEARCH. What is already known about this subject. What this study adds

Transcription

1 RESEARCH Impact of a Clinical Pharmacy Program on Changes in Hemoglobin A1c, Diabetes-Related Hospitalizations, and Diabetes-Related Emergency Department Visits for Patients with Diabetes in an Underserved Population Nancy Chung, PharmD, BCACP; Karen Rascati, PhD; Debra Lopez, PharmD, CDE, BCPS, BCACP; Jason Jokerst, PharmD, BCPS; and Aida Garza, PharmD, CDE, BCACP ABSTRACT BACKGROUND: Diabetes mellitus is associated with substantial morbidity and mortality. With the rise in prevalence of diabetes, there has been an increased need for clinical pharmacy services focused on diabetes management in ambulatory clinics. However, more data IS needed to determine the overall impact that clinical pharmacists have on preventing diabetesrelated inpatient admissions and emergency department (ED) visits for patients with diabetes, especially in an underserved population. OBJECTIVES: To assess the impact of clinical pharmacy services on the change in hemoglobin A1c measurements, the number of diabetes-related hospitalizations, and the number of diabetes-related ED visits for patients with uncontrolled diabetes. METHODS: This was a retrospective study that evaluated outcomes for patients referred to a clinical pharmacist for management of diabetes, compared with patients who were not seen by a clinical pharmacist. Adult patients aged between 18 and 89 years with a diagnosis of type 1 or type 2 diabetes mellitus were identified, using the electronic medical records from CommUnityCare outpatient clinics in central Texas during the period July 1, 2007, through July 1, Patients enrolled had poor glycemic control, defined as an A1c 9% at baseline (index), with at least 3 visits with a clinical pharmacist or 3 visits to usual care. Patients with at least 1 year of pre-index data, 1 year of post-index follow-up, and a post-index A1c measure were included in the study. Propensity score (PS) matching was used to create a 1:1 cohort. T-tests were used to calculate results for the main outcome variables (change in A1c, change in number of diabetes-related hospitalizations, and change in number of diabetes-related ED visits). In addition, general linear models (GLM) were used to control for baseline demographic and clinical characteristics. RESULTS: A total of 782 patients met inclusion criteria, 557 in the usual care (control) group and 225 in the clinical pharmacy (intervention) group. PS matching provided a 1:1 matched sample of 220 patients per cohort. When assessing the change in the number of diabetes-related hospitalizations from the pre-index year to the post-index year, patients in the control group had an increase of 8 hospitalizations (8 visits per 220 patients, mean = 0.036, SD = 0.284), while the intervention group had a decrease of 1 hospitalization (-1 visit per 220 patients, mean = , SD=0.278). Both the t-test (P = 0.06) and GLM model (P = 0.06) indicated that the difference was statistically significant. When assessing the change in the number of diabetes-related ED visits from the pre-index year to the post-index year, we found patients in the control group had an increase of 16 ED visits (16 visits per 220 patients, mean = 0.073, SD = 0.584), while the interven- tion group had an increase of 4 ED visits (4 visits per 220 patients, mean = , SD=0.641). Both the t-test (P = 0.18) and GLM model (P = 0.28) indicated that the difference was not statistically significant. A1c levels were reduced in the post-index period for both groups. For the control group, A1c reduction was 1.50 (from to 9.67, SD = 2.49). For the intervention group, A1c reduction was 1.90 (from to 9.19, SD = 2.44). Both the t-test (P = 0.04) and GLM model (P = 0.05) indicated that the A1c difference was statistically significant. CONCLUSIONS: Underserved patients with baseline uncontrolled diabetes who were managed by a clinical pharmacist in the outpatient setting had a higher decrease in A1c compared with usual care. The changes in diabetes-related hospitalizations and diabetes-related ED visits were in the hypothesized direction, but the comparison for ED visits was not statistically significant. J Manag Care Pharm. 2014;20(9): Copyright 2014, Academy of Managed Care Pharmacy. All rights reserved. What is already known about this subject Diabetes mellitus imposes a substantial economic burden on society, and hospitalization costs comprise the largest expenditure. Additionally, elevated hemoglobin A1c is associated with increased diabetes-related complications, including hospitalizations. Clinical pharmacist interventions have been associated with mostly positive clinical outcomes on A1c; however, there is limited data on the impact of clinical pharmacy services in preventing diabetes-related hospitalizations and diabetes-related emergency department (ED) visits. Indigent populations have barriers to diabetes care that may include, but are not limited to, access to health care, cultural beliefs, and language. What this study adds Patients managed by a clinical pharmacist working under a collaborative drug therapy management protocol had a decrease in the number of hospitalizations during the 1-year follow-up period, while those in the usual care group had an increase in the number of hospitalizations. 914 Journal of Managed Care & Specialty Pharmacy JMCP September 2014 Vol. 20, No. 9

2 What this study adds (continued) While both cohorts (usual care vs. clinical pharmacy group) showed a decrease in A1c scores at the 1-year follow-up, there was greater A1c reduction in the clinical pharmacy group. Clinical pharmacist management of patients with uncontrolled diabetes was associated with improved clinical outcomes in a federally qualified health center. Diabetes mellitus is a major cause of illness in the United States and is a growing public health problem. The diabetes epidemic affects approximately 8.3% of the U.S. population, and its prevalence is increasing, with approximately 79 million Americans at increased risk for diabetes. 1 Besides the risk of complications and the burden of disease on the patient, the cost associated with diabetes is a significant economic burden. Medical expenses for people living with diabetes are more than 2 times higher than for people living without diabetes, and inpatient care continues to be the largest expenditure (43% of total cost). In fact, the American Diabetes Association (ADA) estimates that the total annual cost of diagnosed diabetes rose to $245 billion in 2012 from $174 billion in ,3 Furthermore, the ADA estimated that nearly one-third of patients with diabetes may require 2 or more hospitalizations per year. 4 With the prevalence of diabetes expected to rise, inpatient costs will likely remain a major contributor to diabetes-related costs. Pharmacy services have expanded from traditional dispensing roles to include comprehensive clinical pharmacy services as part of the health care team. Currently, at least 70% of state boards of pharmacy authorize collaborative drug therapy management (CDTM) protocols between a clinical pharmacist and prescriber in the outpatient setting. 5 In the state of Texas, pharmacists working with a CDTM protocol can initiate, modify, or continue drug therapy for a referred patient. Previous studies have shown the value of clinical pharmacy services in improving health outcomes and reducing economic costs for patients with diabetes. 6 Benefits of CDTM include, but are not limited to, improving patient adherence, decreasing adverse events, avoiding drug interactions, and using the knowledge of the clinical pharmacist to promote enhanced services via direct patient care. 6-8 Previous studies have shown that patients with diabetes who receive clinical pharmacist interventions demonstrate significantly improved hemoglobin A1c Furthermore, the data have shown that increased A1c is associated with increased costs per hospitalization. In fact, patients with a mean A1c of at least 10% or greater have significantly higher rates of diabetes-related hospital utilization, compared with patients with a mean A1c of < 7%. 13 However, there is a lack of data for clinical pharmacists impact on inpatient utilization, especially in underserved populations. The purpose of this study was to assess the effect of clinical pharmacist involvement in a federally qualified health center (FQHC) on the change in A1c and frequency of diabetesrelated hospitalizations and emergency department (ED) visits. Methods Study Setting CommUnityCare is a FQHC with outpatient clinics in Austin, Texas, serving about 66,000 underserved and uninsured patients ( The majority of patients are enrolled in the Medical Access Program, which is available to Travis County residents who have a household income below the federal poverty guidelines, followed by patients who have federal insurance (i.e., Medicaid, Medicare). The clinical pharmacy program at CommUnityCare is comprised of pharmacists who have completed at least 1 year of a postdoctoral residency training program. Many of them hold advanced certifications in patient care (i.e., certified diabetes educator) and medication management (i.e., board certified ambulatory care pharmacist). CommUnityCare clinical pharmacists and participating prescribers update the CDTM protocol annually. Some services provided by the individual clinical pharmacists include, but are not limited to, implementing new medications, titrating medications, ordering laboratory panels, counseling on lifestyle modification, providing diabetes education, and managing associated comorbid conditions (i.e., hypertension, dyslipidemia). Under a CDTM protocol, patients referred to the clinical pharmacists for diabetes management are scheduled for a 30-minute visit as often as required to reach the patient s goals (i.e., A1c) and as clinically necessary for patient safety (i.e., to monitor laboratory values). Data from CommUnityCare covering the period between July 1, 2007, through July 1, 2011 (study period), were used to explore the relationship between clinical pharmacists interventions and changes in A1c, diabetes-related hospitalizations, and diabetes-related ED visits. This study was a retrospective review via electronic medical records from CommUnityCare, and the patients data were de-identified. The study was approved by the University of Texas Institutional Review Board and by CommUnityCare. Patient Selection Adult patients aged between 18 and 89 years with a diagnosis of type 1 or type 2 diabetes mellitus were identified during the time period of July 1, 2007, through July 1, Documentation of an A1c value at baseline (defined as within 3 months before or after initial visit with clinical pharmacy or usual care) was required. For enrollment in the study, patients must have had poor glycemic control, defined as an A1c 9% at baseline, and at least 3 visits with a clinical pharmacist or usual care. Usual care patients were followed by their primary care providers and had no appointments with a clinical pharmacist during the study period. Patients must have been Vol. 20, No. 9 September 2014 JMCP Journal of Managed Care & Specialty Pharmacy 915

3 TABLE 1 ICD-9-CM Codes for Diabetes and Diabetes-Related Complications Complication Identifier (ICD-9-CM Code) Description Hyperglycemia 250.1x Diabetes with ketoacidosis 250.2x Diabetes with hyperosmolarity 250.3x Diabetes with other coma Diabetic nephropathy 250.4x Diabetes with renal manifestations Diabetic retinopathy 250.5x Diabetes with ophthalmic complications 362.xx Diabetic retinopathy (e.g., proliferative) Diabetic neuropathy 250.6x Diabetes with neurological manifestations 357.2x Neuropathy in diabetes Diabetic peripheral circulatory disorders 250.7x Diabetes with peripheral circulatory disorders Hypoglycemia 250.8x Diabetes with other specified manifestations ICD-9-CM = International Classification of Diseases, Ninth Edition, Clinical Modification. eligible for at least 1 year before and 1 year after their index dates. In addition, a follow-up A1c (at least 1 year post-index) must have been documented in the database. Patients were excluded if they had a clinical pharmacy visit for any disease state before the study initiation, or if they had a diagnosis of cancer or human immunodeficiency virus, were pregnant, or had diagnosis codes for motor vehicle accidents or chronic pain (i.e., back pain, generalized pain, myalgia). Study Outcomes The outcomes of this study were 1-year post-index change in A1c, diabetes-related hospitalizations, and diabetes-related ED visits. Follow-up A1c values at least 1 year post-index were used to calculate the change in A1c from baseline for each patient. The International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) was used to identify diabetes-related diagnoses (Table 1). The ICD-9-CM codes identified are shortterm complications of diabetes that include hyperglycemia and hypoglycemia, as well as long-term complications of diabetes that include microvascular and macrovascular conditions. A hospitalization or ED visit was considered diabetes-related if 1 of the designated ICD-9-CM codes was documented in the diagnosis fields (up to 5 ICD-9-CM codes could be listed). Diabetes-related ED and hospital visits for 1 year pre-index and 1 year post-index were identified, and the difference in number of events for each patient was calculated. Statistical Analysis Clinical pharmacy and usual care groups were propensity score (PS) matched using a 1:1 ratio. Variables used for PS matching were age, gender, pre-index hospitalizations, preindex ED visits, and index A1c values. Sample size estimations FIGURE 1 Patient Sample Extraction Data extracted for patients with A1c baseline > 9% and at least 3 provider visits n = 1,716 (C = 1,256, I = 460) At least 1 year eligibility post-index visit n = 1,231 (C = 903, I = 328) At least 1 year follow-up and A1c follow-up n = 782 (C = 557, I = 225) Exclusion: less than 1 year eligibility post-index visit n = 485 (C = 353, I = 132) Exclusion: no follow-up A1c measure n = 449 (C = 346, I = 103) C = control group (usual care group); I = intervention group (clinical pharmacy group). were based on change in A1c, which was the primary outcome variable. For a 1-sided t-test of 2 independent means, with an effect size of 0.25, an alpha error P of 0.10, and a power of 95%, we estimated that 275 patients were needed in each group. In addition, to assess the change in A1c values, a general linear model (GLM) was conducted to control for baseline demographics (age, gender, race/ethnicity) and clinical characteristics (pre-index diabetes-related hospitalizations, pre-index diabetes-related ED visits, and index A1c). For the outcomes with count results and a high number of zeros (hospitalizations and ED visits), zero-inflated negative binomial regressions were conducted, again controlling for baseline characteristics. All analyses were conducted with an a priori alpha level of P < 0.10, using SAS version 9.2 (SAS Institute, Inc., Cary, NC). Results Patient Characteristics A total of 782 patients met inclusion criteria (Figure 1), with 225 intervention patients and 557 control patients. PS matching produced a 1:1 match of 220 patients per group. Demographics characteristics of the 2 cohorts before and after matching are listed in Table 2. Diabetes-Related Outcomes Table 3 provides information on the change in outcomes for both unmatched and matched samples to show similarity for both. T-test results for the matched groups (220 in each group) will be discussed. For the usual care group, the number of 916 Journal of Managed Care & Specialty Pharmacy JMCP September 2014 Vol. 20, No. 9

4 TABLE 2 Baseline Demographic Characteristics for Unmatched and Matched Groups Variable Unmatched (N = 557) Clinical Pharmacy Group Unmatched (N = 225) Matched Clinical Pharmacy Group Matched Age, mean ± SD 51.1 ± ± ± ± 11.5 Female, n (%) 304 (54.6) 140 (62.2) 132 (60.0) 135 (61.4) Hispanic, n (%) 418 (75.0) 156 (69.3) 167 (75.9) 153 (69.6) White, n (%) 58 (10.4) 30 (13.3) 24 (10.9) 29 (13.2) Black, n (%) 69 (12.4) 33 (14.7) 25 (11.4) 32 (14.5) Other, n (%) 12 (2.2) 6 (2.7) 4 (1.8) 6 (2.7) SD = standard deviation. hospitalizations for 220 patients was 8 for the pre-index year and 16 for the post-index year, for an increase of 8 hospitalizations (8 per 220, mean = 0.036, standard deviation [SD] = 0.284, or 3.6 hospitalizations per 100 patients). For the clinical pharmacist group, the number of hospitalizations for 220 patients was 6 for the pre-index year and 5 for the post-index year for a decrease of 1 hospitalization (-1 per 220, mean = , SD = 0.278). The difference in these changes was statistically significant (P = 0.06). For the usual care group, the number of ED visits for 220 patients was 28 for the pre-index year and 44 for the post-index year, for an increase of 16 visits (16 per 220, mean = 0.073, SD = 0.584, or 7.3 ED visits per 100 patients). For the clinical pharmacist group, the number of ED visits for 220 patients was 30 for the pre-index year and 34 for the post-index year, for an increase of 4 visits (4 per 220, mean = 0.018, SD = 0.641). The difference in these changes was not statistically significant (P = 0.18). For the matched cohorts, the mean change in A1c was (SD = 2.49) for the usual care group, while the mean change for the clinical pharmacist group was (SD = 2.44). The difference in A1c reduction was statistically significant (P = 0.04). In addition to t-tests, GLMs were conducted for these outcomes, controlling for the following covariates: gender, race/ ethnicity, age, number of pre-index diabetes-related ED visits, number of pre-index diabetes-related hospitalizations, and baseline A1c values. The difference in the number of diabetesrelated hospitalizations and the number of ED visits was not normally distributed, as both variables had a high number of zeros and a standard deviation higher than the mean; thus, a negative binomial distribution was assumed for these 2 models. Similar to the t-test results, the difference in the change in hospitalizations and the difference in A1c values were significantly different at the a priori level of significance (P = 0.06 and P = 0.05, respectively), while the change in the number of ED visits was not statistically different (P = 0.26). Discussion This study demonstrated that patients with uncontrolled type 1 or type 2 diabetes who had at least 3 visits with a clinical pharmacist had an additional improvement in A1c and diabetes-related hospital visits. In addition, the change in the number of diabetes-related ED visits was different in the expected direction, but this difference was not statistically significant. Previous studies have shown that patients who received pharmacist-managed care demonstrated improved A1c values Increased A1c is associated with increased costs per hospitalization, and these inpatient costs account for the major expenditure in health spending for patients with diabetes in the United States. Hospitalizations were considered to be diabetes-related if patients had an ICD-9-CM code listed in any of their diagnosis fields that matched at least 1 of 10 of the study s selected ICD-9-CM codes. Up to 5 ICD-9-CM codes could be listed per event. In contrast, a previous study required the ICD-9-CM code to be listed as the primary or secondary diagnosis to be included for analysis. 8 In our case, all diagnosis field entries were accepted for analysis, since the goal of the study was to identify diabetes-related events, regardless of prioritization of the ICD-9-CM codes. The 10 ICD-9-CM codes identified microvascular complications (i.e., nephropathy, neuropathy, retinopathy) because pharmacologic interventions to normalize glucose can reduce the risk of microvascular complications by 37% for every 1% drop in A1c. 14 A 2010 study used a broader range of ICD-9-CM codes that accounted for macrovascular complications such as ischemic heart disease and stroke, as well as acute complications such as infections. 15 However, a meta-analysis revealed that intensive glucose-lowering therapy modestly reduced major macrovascular events after an average follow-up of 4.4 years in patients with type 2 diabetes. 16 Thus, the length of time to see an effect on macrovascular outcomes precluded the necessity of including ICD-9-CM codes that identified these complications. Because there is large variability in the design of studies that rely on procedural codes for analysis, these are important considerations in designing future strategies to document disease-state complications. Our study identified important findings regarding clinical pharmacy intervention and the effect on inpatient usage in a FQHC. This study did not look at secondary outcome measures that have been documented in previous studies assessing Vol. 20, No. 9 September 2014 JMCP Journal of Managed Care & Specialty Pharmacy 917

5 TABLE 3 Comparison of Change in Outcomes for Unmatched and Matched Groups Variable Unmatched (N = 557) Clinical Pharmacy Group Unmatched (N = 225) Matched Clinical Pharmacy Group Matched Pre-index diabetes hospitalization: n, mean ± SD ± ± ± ± Post-index diabetes hospitalization: n, mean ± SD ± ± ± ± Change in diabetes hospitalizations: a mean ± SD ± ± b ± ± b Pre-index diabetes ED visits: n, mean ± SD ± ± ± ± Post-index diabetes ED visits: n, mean ± SD ± ± ± ± Change in diabetes ED visits: c mean ± SD ± ± d ± ± d Baseline A1c: mean, SD ± ± ± ± 1.55 Follow-up A1c: mean, SD 9.42 ± ± ± ± 2.30 Change in A1c: e mean, SD ± ± 2.50 b ± ± 2.44 b at-tests: unmatched comparison of change, P = 0.01; matched comparison of change, P = b P < c T-tests: unmatched comparison of change, P = 0.17; matched comparison of change, P = d Not significant at P < e T-tests: unmatched comparison of change, P = 0.09; matched comparison of change, P = A1c = hemoglobin A1c; ED = emergency department; SD = standard deviation. pharmacist impact on diabetes care. However, clinical pharmacist intervention has been shown to improve patient adherence with the ADA guidelines for preventive care. Significant improvement in health indicators after clinical pharmacy intervention include a decrease in low-density lipoprotein, increase in frequency of microalbumin screening, increase in annual eye and foot examination, and increase in daily aspirin use as appropriate. 17 Furthermore, improvement of quality-of-life measures (i.e., patient satisfaction) have been documented in the literature. 18 Limitations Since the analysis was retrospective in nature, only associations can be reported, not causality. In addition, retrospective studies do not allow for randomization, so selection bias is possible. Matching was conducted to address this issue, but unmeasured confounders could have played a role in group differences. One common limitation to similar studies comes from small sample sizes. Since less than 300 patients met the inclusion criteria for the intervention group, this limitation reduced the power of the analyses, and a P value less than 0.10 was chosen, based on expected sample size and expected effect size. The use of ICD-9-CM codes rather than medical records may raise concerns about clinical accuracy. For example, it is possible that the ICD-9-CM code was incorrectly documented, since it cannot be ascertained who input the code into the diagnosis field and whether the code matched the final diagnosis at discharge. Additionally, flaws in the ICD-9-CM system could not be avoided. For example, if a patient was admitted for a diabetes-related foot infection, the provider may not have included the diagnosis because a code does not exist for this exact wording. Another limitation was the number of hospitals contracted with the database that submitted information relating to hospitalizations and ED visits. Only 2 hospital systems submitted this information; thus, the outcomes were affected by the availability of only local data if a patient was hospitalized or went to the ED at a noncontracted hospital outside of the city. We assumed that patients only had clinical pharmacist intervention during this time period, but we did not know whether patients received other services that could impact outcomes. For example, if a patient saw a dietician for diabetes nutrition counseling, this interaction could be a variable to improvement of A1c. In addition, the degree/level of clinical pharmacist intervention is not standardized. Clinical pharmacists who work under a CDTM have a variety of roles, depending on the nature and need of the visit. In many instances, the first visit focuses on diabetes education, which includes, but is not limited to, insulin pen administration, glucose monitoring recommendations, nutrition, and preventive services (i.e., immunization recommendations). This study attempted to control for this unknown by requiring that a patient have 3 visits to be included in the study to determine the impact of clinical pharmacist intervention as shown in previous studies. 8,11 This study was conducted in a FQHC setting; thus, generalizability of the results is limited to underinsured and underserved populations. The majority of the study population was Hispanic 918 Journal of Managed Care & Specialty Pharmacy JMCP September 2014 Vol. 20, No. 9

6 (about 70%). Minority populations, especially Hispanic patients, have a higher prevalence of diabetes, worse glycemic control, poor adherence, and more complications than non-hispanic white patients with diabetes. 11 Previous studies have observed that Hispanics have a lower rate of diabetes-related hospitalizations. 11,19,20 Multiple factors contribute to this finding, such as cultural stigma associated with the hospital setting. In this study s demographic, socioeconomic status could prohibit a patient from going to the hospital or ED even when needed. Barriers to health care access also include low English proficiency, low health literacy, lack of personal identification, and misunderstanding of the disease state. 21,22 Conclusions For this underserved population of patients with either type 1 or type 2 diabetes, a significant and beneficial relationship was seen in the change in A1c and the change in the number of diabetes-related hospitalizations after 3 visits with a clinical pharmacist when compared with usual care. While the change in ED visits trended in the expected direction, this comparison was not statistically significant. Future research to examine utilization patterns and changes for clinical pharmacy services may be important for expanding CDTM protocols to further improve outcomes for patients with diabetes. Authors NANCY CHUNG, PharmD, BCACP, is Clinical Assistant Professor, University of Houston College of Pharmacy, Houston, Texas. KAREN RASCATI, PhD, is Professor of Health Outcomes and Pharmacy Practice, and DEBRA LOPEZ, PharmD, CDE, BCPS, BCACP, is Clinical Associate Professor, University of Texas College of Pharmacy, Austin, Texas. JASON JOKERST, PharmD, BCPS, is Clinical Pharmacist, and AIDA GARZA, PharmD, CDE, BCACP, is Clinical Pharmacist, CommUnityCare, Austin, Texas. AUTHOR CORRESPONDENCE: Nancy Chung, PharmD, BCACP, Clinical Assistant Professor, University of Houston College of Pharmacy, 1441 Moursund St., Houston, TX Tel.: ; nchung@central.uh.edu. DISCLOSURES The authors report no conflicts of interest regarding this study. All authors collaborated in the concept and design of this study. Chung and Rascati collected the data and were primarily responsible for data interpretation, with assistance from the other authors. The manuscript was written by Chung, and all authors participated in the revision. REFERENCES 1. U.S. Centers for Disease Control and Prevention. National diabetes fact sheet, Available at: pdf. Accessed June 11, American Diabetes Association. Economic costs of diabetes in the U.S. in Diabetes Care. 2013;36(4): American Diabetes Association. Economic costs of diabetes in the U.S. in Diabetes Care. 2008;31(3): Jiang HJ, Stryer D, Friedman B, Andrews R. Multiple hospitalizations for patients with diabetes. Diabetes Care. 2003;26(5): Hammond RW, Schwartz AH, Campbell MJ, et al. Collaborative drug therapy management by pharmacists Pharmacotherapy. 2003;23(9): Chisholm-Burns MA, Graff Zivin JS, Lee JK, et al. Economic effects of pharmacists on health outcomes in the United States: a systematic review. Am J Health Syst Pharm. 2010;67(19): Cranor CW, Bunting BA, Christensen DB. The Asheville Project: longterm clinical and economic outcomes of a community pharmacy diabetes care program. J Am Pharm Assoc (Wash). 2003;43(2): Anaya JP, Rivera JO, Lawson K, Garcia J, Luna J Jr, Ortiz M. Evaluation of pharmacist-managed diabetes mellitus under a collaborative drug therapy agreement. Am J Health Syst Pharm. 2008;65(19): Sease JM, Franklin MA, Gerrald KR. Pharmacist management of patients with diabetes mellitus enrolled in a rural free clinic. Am J Health Syst Pharm. 2013;70(1): Ip EJ, Shah BM, Yu J, Nguyen LT, Bhatt DC. Enhancing diabetes care by adding a pharmacist to the primary care team. Am J Health Syst Pharm. 2013;70(10): Johnson KA, Chen S, Cheng IN, et al. The impact of clinical pharmacy services integrated into medical homes on diabetes-related clinical outcomes. Ann Pharmacother. 2010;44(12): Choe HM, Mitrovich S, Dubay D, Hayward RA, Krein SL, Vijan S. Proactive case management of high-risk patients with type 2 diabetes mellitus by a clinical pharmacist: a randomized controlled trial. Am J Manag Care. 2005;11(4): Coast-Senior EA, Kroner BA, Kelley CL, Trilli LE. Management of patients with type 2 diabetes by pharmacists in primary care clinics. Ann Pharmacother. 1998;32(6): UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes. Lancet. 1998;352(9131): Menzin J, Korn JR, Cohen J, et al. Relationship between glycemic control and diabetes-related hospital costs in patients with type 1 or type 2 diabetes mellitus. J Manag Care Pharm. 2010;16(4): Available at: amcp.org/workarea/downloadasset.aspx?id= Turnbull FM, Abraira C, Anderson RJ, et al. Intensive glucose control and macrovascular outcomes in type 2 diabetes. Diabetologia. 2009;52(11): Kiel PJ, McCord AD. Pharmacist impact on clinical outcomes in a diabetes disease management program via collaborative practice. Ann Pharmacother. 2005;39(11): Scott DM, Boyd ST, Stephan M, Augustine SC, Reardon TP. Outcomes of pharmacist-managed diabetes care services in a community health center. Am J Health Syst Pharm. 2006;63(21): Cook CB, Naylor DB, Hentz JB, et al. Disparities in diabetes-related hospitalizations: relationship of age, sex, and race/ethnicity with hospital discharges, lengths of stay, and direct inpatient charges. Ethn Dis. 2006;16(1): Davis SK, Liu Y, Gibbons GH. Disparities in trends of hospitalization for potentially preventable chronic conditions among African Americans during the 1990s: implications and benchmarks. Am J Public Health. 2003;93(3): Dilworth TJ, Mott D, Young H. Pharmacists communication with Spanish-speaking patients: a review of the literature to establish an agenda for future research. Res Social Adm Pharm. 2009;5(2): Gerber BS, Cano AI, Caceres ML, et al. A pharmacist and health promoter team to improve medication adherence among Latinos with diabetes. Ann Pharmacother. 2010;44(1): Vol. 20, No. 9 September 2014 JMCP Journal of Managed Care & Specialty Pharmacy 919