Currently Viewing:
The American Journal of Managed Care July 2017
The Price May Not Be Right: The Value of Comparison Shopping for Prescription Drugs
Sanjay Arora, MD; Neeraj Sood, PhD; Sophie Terp, MD; and Geoffrey Joyce, PhD
US Internists' Awareness and Use of Overtreatment Guidelines: A National Survey
Kira L. Ryskina, MD, MS; Eric S. Holmboe, MD; Elizabeth Bernabeo, MPH; Rachel M. Werner, MD, PhD; Judy A. Shea, PhD; and Judith A. Long, MD
Cost-Effectiveness of a Patient Navigation Program to Improve Cervical Cancer Screening
Yan Li, PhD; Erin Carlson, DrPH; Roberto Villarreal, MD; Leah Meraz, MA; and José A. Pagán, PhD
The Association Between Insurance Type and Cost-Related Delay in Care: A Survey
Sora Al Rowas, MD, MSc; Michael B. Rothberg, MD, MPH; Benjamin Johnson, MD; Joel Miller, MD, MPH; Mohanad AlMahmoud, MD; Jennifer Friderici, MS; Sarah L. Goff, MD; and Tara Lagu, MD, MPH
Availability and Variation of Publicly Reported Prescription Drug Prices
Jeffrey T. Kullgren, MD, MS, MPH; Joel E. Segel, PhD; Timothy A. Peterson, MD, MBA; A. Mark Fendrick, MD; and Simone Singh, PhD
Twitter Accounts Followed by Congressional Health Staff
David Grande, MD, MPA; Zachary F. Meisel, MD, MS; Raina M. Merchant, MD, MS; Jane Seymour, MPH; and Sarah E. Gollust, PhD
Predicting High-Cost Privately Insured Patients Based on Self-Reported Health and Utilization Data
Peter J. Cunningham, PhD
Adaptation of an Asthma Management Program to a Small Clinic
Kenny Yat-Choi Kwong, MD; Nasser Redjal, MD; Lyne Scott, MD; Marilyn Li, MD; Salima Thobani, MD; and Brian Yang, MS
Currently Reading
Medication Adherence and Improved Outcomes Among Patients With Type 2 Diabetes
Sarah E. Curtis, MPH; Kristina S. Boye, PhD; Maureen J. Lage, PhD; and Luis-Emilio Garcia-Perez, MD, PhD

Medication Adherence and Improved Outcomes Among Patients With Type 2 Diabetes

Sarah E. Curtis, MPH; Kristina S. Boye, PhD; Maureen J. Lage, PhD; and Luis-Emilio Garcia-Perez, MD, PhD
Adherence to glucose-lowering agents was associated with a significant reduction in use of acute care resources without any increased total medical costs.

Objectives: Examine the association between adherence to glucose-lowering agents (GLAs) and patient outcomes in an adult type 2 diabetes (T2D) population.

Study Design: Retrospective analysis.

Methods: Truven’s Commercial Claims and Encounters database supplied data from July 1, 2009, to June 30, 2014. Patients 18 to 64 years with T2D were included if they received a GLA from July 1, 2010, through June 30, 2011. Multivariable analyses examined the relationships among 3-year patient outcomes and adherence, defined as proportion of days covered 80% or more. Outcomes included all-cause medical costs, acute care resource utilization, and acute complications. 

Results: Although there was no statistically significant difference in total costs when comparing adherent and nonadherent patients ($38,633 vs $38,357; P = .0720), acute care costs ($12,153 vs $8233; P <.0001) and outpatient costs ($16,964 vs $15,457; P <.0001) were significantly lower for adherent patients. Adherence was also associated with a lower probability of hospitalization (22.71% vs 17.65%; P <.0001) and emergency department (ED) visits (45.61% vs 38.47%; P <.0001), fewer hospitalizations (0.40 vs 0.27; P <.0001) and ED visits (1.23 vs 0.83; P <.0001), and a shorter hospital length of stay (2.16 vs 1.25 days; P <.0001). Adherent patients were also less likely to be diagnosed with an acute complication in the 3-year post period (12.54% vs 9.64%; P <.0001). 
Conclusions: Compared with nonadherence, adherence to GLAs among patients with T2D was associated with a significant reduction in acute care costs and resource utilization, outcomes that may positively impact the welfare of patients.
Takeaway Points
  • Compared with nonadherent patients, adherent patients had: 
  • Less likelihood of a hospitalization or an emergency department (ED) visit. 
  • Fewer hospitalizations and fewer ED visits. 
  • Reduced acute care costs and outpatient costs. 
  • Increased drug costs. 
  • Better patient outcomes. 
  • No significant difference in total medical costs.
Diabetes is a chronic and highly prevalent condition that may lead to severe complications, such as lower limb amputations, blindness, kidney failure, stroke, heart disease, and early death.1 As of 2012, the estimated number of US adults (20 years or older) diagnosed with diabetes was 28.9 million, or 12.3% of the population.1 For these individuals, the total cost of the disease in 2012 was $245 billion, comprising $176 billion in direct medical costs and $69 billion in decreased productivity.2 The majority of patients with diabetes (90% to 95%) have type 2 diabetes (T2D),1 the prevalence of which is growing so rapidly that about 40% of all Americans are projected to be diagnosed with the disease at some point in their adult lives.3

Given the large and increasing number of patients with T2D and the associated burden, several recent studies have examined factors that may influence T2D treatment outcomes. One such factor is adherence to prescribed glucose-lowering agents (GLAs). Improved adherence to GLAs has been shown to be linked to a reduction in hospitalizations and/or emergency department (ED) visits,4-9 complications,6,10,11 and costs.5,9,12-15 However, limited research has focused on all classes of GLAs, and inclusion of the newer GLA classes is even rarer.5,7-9,11

The goal of the present study was to further the literature examining connections between patient behavior and T2D treatment outcomes. To this end, our retrospective, naturalistic investigation used a US claims database to observe the medical records associated with a large population of Americans aged 18 to 64 years with T2D. This study examined the relationships between GLA adherence and patient outcomes, including acute care resource utilization, acute complications, and total costs.


Truven’s Health Analytics MarketScan Commercial Claims and Encounters database, including data from July 1, 2009, through June 30, 2014, was used for this study. This database consists of the healthcare records of millions of individuals who are covered by fully or partially capitated fee-for-service health plans. As such, the database provides detailed costs, use, and outcomes data for healthcare services performed in both inpatient and outpatient settings. Medical claims are linked to outpatient prescription drug claims and person-level enrollment information. The data are fully deidentified and compliant with the Health Insurance Portability and Accountability Act.

For inclusion in the study, patients were required to have received at least 2 diagnoses of T2D using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes 250.x0 or 250.x216,17 in the time period of July 1, 2010, through June 30, 2011 (ie, the identification window), and to have received at least 1 prescription for a GLA within the identification window; the date of the first such prescription was identified as the index date. Additionally, patients were required to be aged 18 to 64 years as of the index date and to have had continuous insurance coverage from 1 year leading up to the index date (the pre-period) through 3 years following the index date (the post period), as well as valid patient demographic data. Patients were excluded from the analyses if they received a diagnosis of type 1 diabetes (ICD-9-CM codes: 250.x1 or 250.x3) or pregnancy (ICD-9-CM codes: 630.xx-670.xx, V22.xx-V23.xx, V30.xx-V39.xx) at any time from the start of the pre-period through the end of the post period. Figure 1 illustrates how each of these criteria affected sample size.

The analyses focused on the relationship between patient adherence and outcomes, where adherence was proxied by the proportion of days covered (PDC). The PDC is a measure of adherence used by both the Pharmacy Quality Alliance18 and CMS, and it was defined as the percentage of days an individual received at least 1 GLA in the first year of the post period. For all medications except insulin, the PDC was constructed using the “days supplied” field provided in the database. For insulin, this was calculated as the average number of days between fills for an insulin prescription.19 Consistent with Healthcare Effectiveness Data and Information Set guidelines for the use of oral GLAs20 and with previous research, patients were categorized as adherent if they achieved a PDC threshold of at least 80%.6,8,21,22

Outcomes of interest included costs, acute care resource utilization, and acute complications. Costs were subdivided into 4 categories—acute care (hospitalization and ED), outpatient, drug, and total—and all were converted to 2014 prices using the medical component of the Consumer Price Index. In all cases, costs were calculated using gross payments to a provider for a service. Acute care resource utilization was defined as the probability of a hospitalization or an ED visit, the number of hospitalizations or ED visits, and the hospital length of stay (LOS). Acute complications were identified based on receipt of a diagnosis of hyperglycemia (ICD-9-CM code: 790.29), hypoglycemia (ICD-9-CM codes: 250.8, 251.0, 251.1, or 251.2), or diabetic or hypoglycemic coma (ICD-9-CM codes: 250.3, 251.0).

Multivariable analyses were used to examine the relationship between outcomes of interest and patient adherence. All analyses controlled for the individual patient’s characteristics (age, sex, region of residence, and insurance coverage), preperiod general health status, comorbidities, medication use, and providers. General health was proxied by the Charlson Comorbidity Index (CCI), which is scored on a scale of 0 to 33 based on the presence of comorbidities such as cardiovascular diseases, dementia, chronic pulmonary disease, hypertension, liver disease, and malignancies.23,24 The severity of any diabetes complication was proxied using the Diabetes Complications Severity Index (DCSI), which is scored on a scale of 0 to 13 based on the presence of conditions such as retinopathy, neuropathy, nephropathy, and metabolic disorders.25 In addition, the models also included comorbidities that have been shown to be common in patients with diabetes and were not captured in either the CCI or DCSI score. Specifically, anxiety has been shown to be prevalent in patients with diabetes,26 comorbid depression has been shown to be linked to nonadherence,27 and hyperlipidemia has been used in previous studies that examined the relationship between adherence and outcomes.28 Therefore, this study included indicator variables for pre-period diagnoses of anxiety (ICD-9-CM codes: 300.xx except 300.3x and 300.4x), depression (ICD-9-CM codes: 296.2, 296.3, 300.4, or 311.xx), and hyperlipidemia (ICD-9-CM codes: 272.1-272.4).

The analyses also controlled for provider visits/services during the pre-period, with indicator variables for renal dialysis therapy and for visits to a cardiologist, endocrinologist, family medicine practitioner, or internal medicine practitioner. Also included as covariates in the model were the number and type of GLAs prescribed and overall medication use in the first 3 months of the post period. These factors were captured by an indicator variable for insulin use, the number of noninsulin classes of GLAs prescribed, and the number of non-GLA medications prescribed. The classes of noninsulin GLAs included alpha-glucosidase inhibitors, amylin analogs, biguanides, dipeptidyl peptidase-4 inhibitors, dopamine agonists, glucagon-like peptide-1 receptor agonists, meglitinides, sodium-glucose co-transporter-2 inhibitors, sulfonylureas, thiazolidinediones, and oral fixed combinations.

General linear models with a gamma distribution and log link were used to examine all-cause outpatient, drug, and total costs. Two-part models were used to examine acute care costs, in which the first part captured the probability of an all-cause or diabetes-related acute care visit and the second part estimated costs among users of the service. Resource utilization was examined by estimating negative binomial regressions for the number of hospitalizations, the number of ED visits, and hospital LOS. Logistic regressions were used to examine the probability of hospitalization, ED use, or acute complication. Differences in the estimated outcomes were then examined by adherence status. To calculate whether differences in outcomes between adherent and nonadherent patients were statistically significant, t tests were used. All analyses were conducted using SAS version 9.3 (SAS Institute, Inc; Cary, North Carolina). A P value of <.05 was considered to be statistically significant.


Table 1 presents descriptive statistics for the 228,074 individuals included in the study. The majority of patients were male (53.75%), and the mean age was 52 years (standard deviation [SD] = 7.1). Patients most commonly resided in the South (42.08%) or North Central (24.42%) regions of the country, and most were covered by preferred provider organizations (53.95%) or a health maintenance organization (21.33%). More than half the patients visited a family medicine practitioner (54.24%), and over one-third (40.37%) had at least 1 internal medicine visit in the pre-period; less frequently, they visited a cardiologist (17.11%) or endocrinologist (8.23%). Approximately 1 in 8 patients (13.12%) received a prescription for insulin in the first 3 months of the post period, and patients were prescribed an average 1.42 noninsulin classes of GLA, including oral fixed combination drugs as a distinct class, and 4.87 non-GLA medications over this same time period.

In addition to providing characteristics for the entire population, Table 1 also presents descriptive statistics comparing adherent with nonadherent patients. Adherent patients were significantly older (53.47 vs 51.18 years; P <.0001), more likely to be male (56.51% vs 50.80%; P <.0001), and less likely to reside in the South (38.93% vs 45.45%; P <.0001) compared with nonadherent patients. Adherent patients were also in better general health compared with nonadherent patients, as evidenced by lower rates of anxiety (2.79% vs 3.90%; P <.0001) and depression (5.44% vs 7.15%; P <.0001), as well as significantly lower rates of visits to a cardiologist (16.76% vs 17.48%; P <.0001). Adherent patients may have less severe diabetes compared with nonadherent patients, as evidenced by the significantly lower DCSI score (0.50 vs 0.52; P <.0001) and by less use of insulin in the first 3 months of the post period (9.84% vs 16.63%; P <.0001). However, adherent patients received significantly more noninsulin classes of medications in the first 3 months of the post period (1.63 vs 1.20; P <.0001), as well as a higher number of non-GLA medications (5.05 vs 4.68; P <.0001).

Figure 2 focuses on the association between adherence to GLAs and 3-year all-cause medical costs. As shown in this figure, adherence was associated with significant reductions in both acute care costs ($8223 vs $12,153; P <.0001) and outpatient costs ($15,457 vs $16,964; P <.0001) and significant increases in drug costs ($14,816 vs $9390; P <.0001). Consistent with significantly higher drug costs and significantly lower acute care costs and outpatient costs for adherent patients compared with nonadherent patients, there was no statistically significant difference in total costs between these 2 groups ($38,357 vs $38,633; P = .0720).

Copyright AJMC 2006-2020 Clinical Care Targeted Communications Group, LLC. All Rights Reserved.
Welcome the the new and improved, the premier managed market network. Tell us about yourself so that we can serve you better.
Sign Up