David P. Nau, PhD

Adherence to therapy in patients with type 2 diabetes mellitus is contingent upon a number of variables, including variables specific to the patient, to the provider, and to the treatment. While treatment selection will involve consideration to maximize effectiveness and minimize side effects, the physician must also take into account the priorities and preferences of each individual patient. For some patients, the risk of weight gain may exert a significant influence on adherence, while for others the risk of hypoglycemia or the cost of medications may be more important factors. It is incumbent upon physicians to discuss these issues with patients and to develop a patient-centric treatment plan to achieve optimal adherence and therapeutic outcomes. The nature of the clinical setting can also influence the likelihood of patient adherence to treatment. A multidisciplinary team approach to diabetes management has been shown to improve outcomes and to have a neutral or beneficial effect on costs. The treatment plan itself plays an additional role in the likelihood of a patient adhering to treatment. Less complex treatment regimens with fewer pills are associated with higher rates of adherence, as are fixed-dose combinations for those patients requiring combination therapy. Frequency and timing of dosing are also important aspects of adherence, as once-daily dosing is associated with higher rates of adherence than twice-daily dosing for anti-hyperglycemic medications.

The effective management of type 2 diabetes mellitus (T2DM) is a multifaceted effort involving lifestyle interventions (eg, diet and exercise) and, for most patients, drug therapy. Regardless of the intervention made, adherence is critical to achieving treatment goals. In fact, a connection between adherence to medications and glycated hemoglobin (A1C) has been shown: when adherence increases, A1C decreases. Higher adherence likely leads to better treatment effectiveness.

 Yet despite the importance of adherence to achieving treatment goals, there is room for improvement with regard to the current level of adherence to treatments among patients with T2DM.

 Focusing on drug therapy, in patients taking oral medications for glycemic control, the rate of adherence has been reported to be as low as 65%.

The challenges to patient adherence to prescribed treatment for T2DM are multiple and varied, may be mutually reinforcing, and may be difficult to overcome. Among the many reasons for nonadherence in T2DM, the most common ones relate to fear of treatment side effects (particularly weight gain and hypoglycemia), needle anxiety (for injectable therapies), inconvenience or complexity of a prescribed treatment regimen, inaccessibility to treatment arising from cost or formulary issues, and poor levels of patient knowledge about the use and importance of therapies for glycemic control.

Oral and injectable non-insulin therapies represent a broad array of treatments and drug classes which are commonly used before insulin is initiated. Strategies to improve adherence to oral and injectable non-insulin therapies must address each of the areas of potential adherence deficit, and these can range from the relatively straightforward—eg, optimizing therapies to avoid or minimize side effects, prescribing medications with easier-to-use formulations and simpler dosing schedules&mdash;to more complex approaches that may require changes in the way T2DM is treated by clinicians as well as cooperation on the part of third-party payers. The potential benefits of achieving higher rates of treatment adherence in T2DM include improved patient outcomes and cost savings.

 This article will focus both on the potential causes of poor adherence in T2DM and strategies to improve adherence to oral and injectable non-insulin therapies. Potential causes for poor adherence to insulin therapy in patients with T2DM along with strategies for improving adherence will be discussed in the third article in this supplement.

Impact of Patient-Related Factors on Adherence to Drug Therapy

A key therapeutic challenge in managing T2DM—and one with particular relevance to adherence&mdash;is the need to balance therapeutic efficacy with tolerability. The dual goals of achieving A1C target and avoiding or minimizing side effects that may impact tolerability and promote poor adherence need to be taken into consideration when formulating a treatment plan for patients with T2DM. The situation is complicated by the fact that unlike certain other diseases, in which disease management and therapeutic intervention may occur during a discrete period of time of illness, the task for patients with T2DM of achieving glycemic control is usually one that extends for the duration of a patient’s life. Ultimately, the patient is responsible for the day-to-day management of T2DM, including adherence to lifestyle changes as well as medications. Psychosocial factors can weigh heavily on patients and have a meaningful impact on the likelihood of treatment adherence. An international study of patients with diabetes and their providers found that between two-thirds and three-fourths of providers reported that their patients had psychosocial barriers that interfered with their adherence to therapy.

Among the barriers to adherence in T2DM, patient fear of side effects such as weight gain, increased risk of cardiovascular events, and hypoglycemia are among the most influential factors. The frequency of these side effects, and their impact on adherence, is illustrated by a survey of 2074 patients with T2DM receiving at least 1 oral antidiabetic drug (OAD) and no insulin therapy.

 Nearly 72% of respondents had experienced a tolerability issue in the 2 weeks prior to the survey and half had experienced more than 2 tolerability issues in that time. Symptoms related to hypoglycemia were the most common, affecting 57% of patients with T2DM, while other common issues included constipation and/or diarrhea (28%), headaches (26%), weight gain (23%), and fluid retention (21%). A limited but statistically significant correlation was observed between the number of tolerability issues experienced by a patient and their likelihood of nonadherence to treatment (

 <.01). Moreover, for every additional tolerability issue experienced by a patient, the risk of medication nonadherence increased by 28% (

The confluence of efficacy, side effects, and adherence was explored in the Real-Life Effectiveness and Care Patterns of Diabetes Management (RECAP-DM) study, an observational study conducted in 6 European countries. RECAP-DM included patients who had failed metformin monotherapy and subsequently added a sulfonylurea or thiazolidinedione to their existing metformin therapy.

 The study sought to evaluate the relationships between the occurrence of hypoglycemia, patient satisfaction with treatment, and poor adherence. Data from 1709 patients with T2DM participating in the study showed that within the previous 12 months, 38% of patients receiving combination therapy had experienced hypoglycemic symptoms. Moreover, only 28% of patients were found to be below the International Diabetes Federation A1C goal of 6.5%. Compared with patients who had not experienced hypoglycemic symptoms, those who had experienced them rated the efficacy, convenience, and satisfaction with treatment significantly lower, and also gave a lower rating to the side-effect profile of their regimen (all 

 <.0001). Patients with hypoglycemic symptoms were also significantly more likely to report barriers to treatment adherence (eg, being bothered by medication side effects). The impact of hypoglycemia on glycemic management is further reflected in the fact that patients with more severe hypoglycemic events were less likely to be at A1C goal (

). The authors suggest that reducing the incidence and severity of hypoglycemic events might help improve treatment adherence and glycemic control.

In reality, the challenge of balancing efficacy and tolerability in T2DM treatment is even more complex since successful disease management should also take into account the perceptions and preferences of the patients themselves. In an ideal world, patients would prefer those treatments that were the most measurably effective and tolerable. However, patients may have different priorities and may also be influenced by factors that they themselves underestimate as being influential on their adherence behavior. For example, in a 2009 survey patients with T2DM in the United States and the United Kingdom receiving OADs were asked a series of questions, the purpose of which was to determine the influence of treatment effectiveness and side effects on adherence to therapy.

 Subgroup analyses revealed several key differences in priorities. For example, weight gain was twice as important to women compared with men, while for those without a college education, cardiovascular risk was the most important factor.

Newer medications may help address medication-related weight gain. Glucagon-like peptide-1 (GLP-1)-receptor agonists represent a newer generation of medications for glycemic control that have the advantage of being associated with moderate weight loss, which might be expected to lower the risk of nonadherence.

 However, it is not clear how patients will balance the lower risk of weight gain against fears of self-injection or the costs of this class of medications. Nonetheless, it may be an appropriate option for patients who are concerned about weight gain.

In summary, the calculation necessary to achieve maximal treatment efficacy and adherence while minimizing side effects is a nuanced one that requires evaluating the benefits and limitations of particular treatments along with an understanding of the preferences and inclinations of the patients who must take them.

Practitioner-Related Factors in Treatment Adherence

If patients are to adhere to treatment, and treatment outcomes are to be optimized, it is important that treatment be customized to individual patients and modified to meet their evolving clinical requirements. There are currently about 10 different classes of antihyperglycemic agents on the market (depending on the classification scheme used), including agents administered orally or parenterally (including injectables); each class has its own advantages and disadvantages.

 The emergence of a diverse armamentarium of treatment options for diabetes provides additional tools for clinicians; however, the diversity of treatment options can make the management of T2DM more complex and difficult, particularly in the primary care setting where time is short and the clinical agenda is long. Additionally, gaps in knowledge of diabetes management among some primary care physicians contribute to the clinical challenges of managing T2DM.

 Improvements both in continuing medical education for physicians who manage patients with T2DM and in educational initiatives for patients who have T2DM can help address these gaps in knowledge.

Other factors beyond knowledge gaps influence the success of diabetes treatment. The Diabetes Outcome Study, which examined factors that were associated with poorer A1C control in primary care clinics, found that single-specialty practices were more successful than multispecialty practices, solo practices, and academic settings in achieving glycemic targets. The same study also found that practices that included nurse practitioners or physician assistants were more successful in helping patients meet their A1C targets.

It may, ultimately, be the case that the best solution for achieving clinical outcomes for patients with diabetes involves the use of multidisciplinary treatment teams. Multidisciplinary teams have been repeatedly shown to improve outcomes and have typically demonstrated beneficial or neutral effects on costs.

 Scanlon et al compared the impact of multidisciplinary teams on diabetes outcomes and costs in a Medicaid- and Medicare-enrolled patient population. Significant improvements were observed for A1C, body mass index (BMI), and blood pressure among patients with higher baseline levels for each of those variables, while no significant differences were seen in treatment costs compared with patients not receiving treatment from a multidisciplinary team.

 An Australian study by Zwar et al found that the implementation of multidisciplinary care in patients with T2DM resulted in significant reductions in A1C (from 8.3% to 7.8%; 

 <.001) for those patients who had A1C levels greater than 7%, and also significantly increased adherence to outcomes guidelines for A1C, blood pressure, and total cholesterol (all 

A different approach was taken in a French study which evaluated the utility of a multidisciplinary team intervention by 1) providing a group of 73 volunteer general practitioners who treated T2DM with an 8-page guideline booklet for the appropriate treatment of T2DM and 2) conducting 2 subsequent audits of outcomes. The first audit, conducted in the year following the provision of guidelines, found generally poor guideline adherence and subpar patient outcomes, including inadequate screening for complications, underestimation of the role of education, and insufficient control of hypertension.

 After the first audit, a multidisciplinary team was given the task of evaluating the clinical deficits it observed and making recommendations for their improvement. Recommended interventions for improvement included the implementation of educational programs and the creation of protocols for screening and treating retinopathy, nephropathy, and foot complications. The following year, a second audit was conducted which demonstrated that the interventions had resulted in significant improvements, including decreases in A1C levels, reductions in foot complications, increases in prescription of antihypertensive medications, and reductions in severe complications.

It should be noted that physicians themselves recognize the need for a multidisciplinary team approach to the treatment of diabetes, and also recognize the challenges that a single practitioner faces when treating diabetes.

 Thus, success in optimizing outcomes would appear to be at least partially contingent on the structure of the clinical setting in which T2DM is treated and on the diversity of clinical expertise involved in the provision of care.

In addition to the strategies already discussed, programs that incentivize providers for quality improvements (eg, pay-for-performance [P4P] programs) also have the potential to impact patient adherence. Given the importance of adherence to achieving treatment goals and the current levels of adherence to treatment for T2DM, financial rewards or penalties tied to clinical outcomes may provide additional motivation for providers to help improve patient adherence to treatment for T2DM. Providers may help encourage adherence by several means, including educating patients regarding the consequences of hyperglycemia and the importance of adherence to lifestyle modifications as well as medications, and selecting medications that have favorable side effect profiles and convenient dosage and administration requirements.

Foels and Hewner evaluated the effect of a quality improvement initiative combining P4P with physician education on the percentage of patients with diabetes meeting specific process and outcome measures (eg, date and value of A1C) and physicians’ scores on composite adherence to guidelines (scored on a 10-point scale). Patient records and physician adherence to guidelines were evaluated in 6-month cycles; the results were discussed with physicians to help identify opportunities for improvement. The physician education component of the program focused on helping physicians formulate a systematic approach to the care of patients with T2DM. From the first cycle to the ninth cycle, the proportion of patients achieving an A1C less than 7% increased from 39.0% to 53.5%, and physicians’ scores on composite adherence to guidelines increased from 4.3 to 6.2.

To investigate the effect of a P4P program on how frequently patients with diabetes received certain quality-of-care interventions, and the relationship between receipt of these interventions and the likelihood of hospitalization, Chen et al conducted a longitudinal study of patients enrolled in a preferred provider organization between January 1, 1999, and December 31, 2006. Patients who had a claim for at least 2 A1C tests and 1 low-density lipoprotein cholesterol (LDL-C) test during 1 year were considered to have received quality care. Patients who saw physicians participating in the P4P program were significantly more likely to receive quality care compared with patients who saw physicians not participating in the P4P program (odds ratio, 1.16; 95% confidence interval [CI], 1.11-1.22; 

 <.001). Patients who received quality care were significantly less likely to be hospitalized in the following year compared with patients who did not receive quality care (incident rate ratio, 0.80; 95% CI, 0.78-0.83; 

Impact of Medication Formulations and Administration on Adherence

Factors related to the medications themselves can also influence adherence. A greater number of daily doses of medication or a more complex regimen involving more than 1 medication can decrease adherence.

 A large-scale study based on data from a pharmacy claims database for a large MCO compared rates of adherence in nearly 17,000 patients with T2DM whose disease was poorly controlled on metformin or on a thiazolidinedione, and who switched to oral rosiglitazone-metformin combination therapy either as a fixed-dose combination (FDC) or loose-pill combination (LPC) regimen.

 Patients switching to the FDC were found to have significantly better improvements in treatment adherence measured by medication possession ratio (MPR) compared with those who switched to LPC (—12.4% vs –4.6%; difference in MPR improvement = 7.8%; 

 <.001). Similarly, a retrospective analysis of data from 2 very large health plan databases found improved adherence in patients switching from sulfonylurea or rosiglitazone monotherapy to rosiglitazone/glimepiride FDC compared with a switch to sulfonylurea/rosiglitazone LPC therapy.

A 2011 systematic literature review analyzed studies of oral FDC therapies and LPC therapies in patients with T2DM to determine the relationship between particular formulations and adherence, patient satisfaction, and treatment costs.

 Based on an analysis of 7 studies, the authors observed that among patients initiating combination therapy, FDC therapies were associated with adherence rates 10% to 13% higher than those seen with LPC therapies. Switching from monotherapy to LPC therapy resulted in a 10% drop in adherence, compared with a 1.5% drop in adherence for those switching from monotherapy to FDC therapy (

 <.001). Switching from LPC to FDC therapy was associated with adherence increases ranging from 3.5% to 12.4%. The authors further found that 4 out of 5 randomized controlled trials showed greater patient satisfaction with FDC therapy, while limited data suggested that FDC therapies were associated with cost savings.

The number of pills, however, may exert a smaller influence on adherence than the frequency of dosing. A study comparing once-daily dosing of glipizide oral therapy with twice-daily dosing found a 60% adherence rate for the former compared with 52% for the latter (

 <.027) despite the fact that the once-daily regimen required the administration of more pills at the time of dosing.

 The influence of frequency of dosing was also the subject of a retrospective observational study from the United Kingdom that compared adherence to immediate release (IR) versus sustained release (SR) metformin in patients with T2DM. IR metformin is dosed at least twice daily, while SR metformin is dosed once daily. The investigators found that switching from IR to SR metformin was associated with an increase in adherence from 62% to 81% (

 One adherence-related benefit of the metformin SR formulation is its potential to reduce the gastrointestinal side effects that impact many patients who take metformin, since gastrointestinal side effects result in high rates of discontinuation.

Currently available GLP-1 agonists include liraglutide, which requires once daily injection at any time of the day regardless of timing with meals, and exenatide, which requires twice-daily injection within the 60 minutes preceding the morning and evening meals.

 Longer-acting GLP-1 agonists, requiring only 1 injection per week, are currently in development, and a form of exenatide requiring only 1 injection per week was approved by the FDA in January 2012.

 The impact of weekly versus daily injections on adherence remains to be determined.

Adherence to therapy in T2DM is influenced by many variables, including treatment side effects and medication formulations. Practitioner-related factors such as the structure of the clinical setting and the particular personnel employed for disease management also impact outcomes. Adherence is also influenced by patients’ preferences and priorities. Formulation of a highly effective treatment regimen to which the patient can be adherent should include discussion with the patient regarding treatment effectiveness, side effects, dosing regimen, and costs upon treatment initiation. As patients typically will require treatment intensification, it is important that clinicians also discuss effectiveness, side effects, and costs of therapy when adjusting the treatment regimen. It is, furthermore, advisable for physicians to educate their patients with T2DM about the consequences of hyperglycemia and key strategies for effective disease management, in addition to apprising patients of expected patterns of therapy escalation, the importance of blood glucose monitoring, and the necessity of maintaining regular appointments in order to achieve glycemic control.

Because of the complexity of diabetes and its comorbidities, a multidisciplinary team approach to managing diabetes is likely to be the most effective one and has been shown to be cost effective or cost neutral. Achieving adherence in order to optimize outcomes within the strictures of cost controls requires cooperation between patients, providers, and third-party payers so that proven treatment strategies can be fully and properly implemented. 

 Pharmacy Quality Alliance, Fairfax Station, VA.

 Funding for the development of this supplement was provided by Novo Nordisk.

 Dr Nau reports no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this supplement.

 Concept and design; drafting of the manuscript; critical revision of the manuscript for important intellectual content; supervision.

 David P. Nau, PhD, Pharmacy Quality Alliance, 9687 South Run Oaks Dr, Fairfax Station, VA, 22039. E-mail: 

Lawrence DB, Ragucci KR, Long LB, Parris BS, Helfer LA

. Relationship of oral antihyperglycemic (sulfonylurea or metformin) medication adherence and hemoglobin A1c goal attainment for HMO patients enrolled in a diabetes disease management program. 

. The relationship between diabetes-related attitudes and patients’ self-reported adherence. 

. Adherence to pharmacologic therapy in patients with type 2 diabetes mellitus. 

Nam S, Chesla C, Stotts NA, Kroon L, Janson SL

. Barriers to diabetes management: patient and provider factors. 

. Third-party reimbursement for diabetes care, self-management education, and supplies. 

. Health systems, patients factors, and quality of care for diabetes: a synthesis of findings from the TRIAD study. 

Hauber AB, Mohamed AF, Johnson FR, Falvey H

. Treatment preferences and medication adherence of people with Type 2 diabetes using oral glucose-lowering agents. 

. Patient perceptions of injection pain and anxiety: a comparison of NovoFine 32-gauge tip 6mm and Micro Fine Plus 31-gauge 5mm needles. 

Alvarez Guisasola F, Tofé Povedano S, Krishnarajah G, Lyu R, Mavros P, Yin D

. Hypoglycaemic symptoms, treatment satisfaction, adherence and their associations with glycaemic goal in patients with type 2 diabetes mellitus: findings from the Real- Life Effectiveness and Care Patterns of Diabetes Management (RECAP-DM) Study. 

Vanderpoel DR, Hussein MA, Watson-Heidari T, Perry A

. Adherence to a fixed-dose combination of rosiglitazone maleate/ metformin hydrochloride in subjects with type 2 diabetes mellitus: a retrospective database analysis. 

Hansen RA, Farley JF, Droege M, Maciejewski ML

. A retrospective cohort study of economic outcomes and adherence to monotherapy with metformin, pioglitazone, or a sulfonylurea among patients with type 2 diabetes mellitus in the United States from 2003 to 2005. 

. Effect of medication nonadherence on hospitalization and mortality among patients with diabetes mellitus. 

. Recommendations for improving adherence to type 2 diabetes mellitus therapy—focus on optimizing insulin-based therapy. 

Peyrot M, Rubin RR, Lauritzen T, Snoek FJ, Matthews DR, Skovlund SE

. Psychosocial problems and barriers to improved diabetes management: results of the Cross-National Diabetes Attitudes, Wishes and Needs (DAWN) Study. 

Pollack MF, Purayidathil FW, Bolge SC, Williams SA

. Patient-reported tolerability issues with oral antidiabetic agents: associations with adherence: treatment satisfaction and health-related quality of life. 

. A meta-analysis of placebo-controlled clinical trials assessing the efficacy and safety of incretin-based medications in patients with type 2 diabetes. 

. Diabetes mellitus. In: Longo DL, Fauci AS, Kasper DL, Hauser SL, Jameson JL, Loscalzo J, eds. 

Harrison’s Principles of Internal Medicine

. 18th ed. New York, NY: McGraw-Hill; 2012.

Brown JB, Harris SB, Webster-Bogaert S, Wetmore S, Faulds C, Stewart M

. The role of patient, physician and systemic factors in the management of type 2 diabetes mellitus. 

. Management of type 2 diabetes in the primary care setting: a practice-based research network study. 

Scanlon DP, Hollenbeak CS, Beich J, Dyer AM, Gabbay RA, Milstein A

. Financial and clinical impact of team-based treatment for medicaid enrollees with diabetes in a federally qualified health center. 

O’Reilly D, Hopkins R, Blackhouse G, et al

. Long-term costutility analysis of a multidisciplinary primary care diabetes management program in Ontario. 

Zwar N, Hasan I, Hermiz O, Vagholkar S, Comino E, Harris M

. Multidisciplinary care plans and diabetes--benefits for patients with poor glycaemic control. 

Zwar NA, Hermiz O, Comino EJ, Shortus T, Burns J, Harris M

. Do multidisciplinary care plans result in better care for patients with type 2 diabetes? 

Varroud-Vial M, Mechaly P, Joannidis S, et al

. Cooperation between general practitioners and diabetologists and clinical audit improve the management of type 2 diabetic patients. 

Ritholz MD, Beverly EA, Abrahamson MJ, Brooks KM, Hultgren BA, Weinger K

. Physicians’ perceptions of the type 2 diabetes multi-disciplinary treatment team: a qualitative study. 

. Integrating pay for performance with educational strategies to improve diabetes care. 

. The effect of a PPO pay-for-performance program on patients with diabetes. 

. Impact of dosage frequency on patient compliance. 

. Patient compliance and persistence with antihyperglycemic drug regimens: evaluation of a Medicaid patient population with type 2 diabetes mellitus. 

Thayer S, Arondekar B, Harley C, Darkow TE

. Adherence to a fixed-dose combination of rosiglitazone/glimepiride in subjects switching from monotherapy or dual therapy with a thiazolidinedione and/or a sulfonylurea. 

Hutchins V, Zhang B, Fleurence RL, Krishnarajah G, Graham J

. A systematic review of adherence, treatment satisfaction and costs, in fixed-dose combination regimens in type 2 diabetes. 

. Effects of once-daily and twice-daily dosing on adherence with prescribed glipizide oral therapy for type 2 diabetes. 

. Adherence in patients transferred from immediate release metformin to a sustained release formulation: a population-based study. 

		Byetta [package insert]. Amylin Pharmaceuticals, Inc; 2010.

		Victoza [package insert]. Princeton, NJ: Novo Nordisk, Inc; 2011.

http://www.novonordisk.com/science/pipeline/rd_pipeline.asp?showid=12

		FDA approves Bydureon—the first and only once-weekly treatment for type 2 diabetes [press release]. Amylin Pharmaceuticals website. 

http://investors.amylin.com/phoenix.zhtml?c=101911&p=irol-newsArticle_pf&ID=1653756&highlight=

Articles

Recommendations for Improving Adherence to Type 2 Diabetes Mellitus Therapy-Focus on Optimizing Oral and Non-Insulin Therapies

To determine whether there were sex-related differencesin the management of dyslipidemia among managed careenrollees with diabetes mellitus.

Retrospective analyses were conductedusing medical and pharmacy claims data from a healthmaintenance organization during 2000 and 2001. Patients withtype 2 diabetes mellitus were identified through a validated algorithmof medication and diagnosis codes. Chi-squared analysis wasused to determine if women were less likely than men to receive alipid test or a lipid-modifying drug. Logistic regression models wereconstructed to compare sex-related differences while controllingfor age, cardiovascular disease diagnosis, and 2 proxies of illnessseverity (hospitalization in 2000 and the intensity of diabetes mellitusdrug therapy).

During 2000 and 2001, 79.4% of women received alipid test compared with 84.2% of men (&#967;

= .01). Also,33.2% of women received a lipid-modifying drug compared with45.5% of men (&#967;

< .01). Logistic regression analysisrevealed that men were more likely than women to receive a lipidtest when controlling for age, cardiovascular disease diagnosis, andillness severity (odds ratio [OR], 1.45; 95% confidence interval,1.13-1.81). Men were also more likely than women to receive alipid-modifying drug when controlling for age, cardiovascular diseasediagnosis, illness severity, and lipid testing (OR, 1.51; 95%confidence interval, 1.22-1.86).

Women with type 2 diabetes mellitus were lesslikely than men with type 2 diabetes mellitus to receive lipid testsor lipid-modifying drugs.

The worldwide prevalence of diabetes mellitus isincreasing, as are the demand for and the cost ofmedical care.

In 2002, there were 18.2 millionpeople (6.3% of the population) in the United States whohad diabetes mellitus.

The direct medical expendituresfor diabetes mellitus totaled nearly $92 billion in 2002,with cardiovascular complications comprising a substantialportion of the costs.

The prevalence of cardiovasculardisease (CVD) and its associated mortalityare substantially higher among persons with diabetesmellitus than among those without diabetes mellitus.

have examined the sex gap in CVD morbidityand mortality among persons with diabetes mellitus.A key epidemiological study

of heart disease inpersons with diabetes mellitus identified that mortalityfrom coronary heart disease has been declining in menbut rising in women. Three meta-analyses

found thatthe odds ratio for CVD-related events in women withdiabetes mellitus is higher than that in men with diabetesmellitus, although the most recent meta-analysis

showed that the sex differences become negligible whenadjusted for CVD risk factors (eg, age, smoking, andcholesterol levels). Nonetheless, a recent study

amongthe participants of the Framingham Heart Study foundthat the relative risk for coronary heart disease mortalityin women with diabetes mellitus is substantiallyhigher than that in men with diabetes mellitus, evenwhen controlling for the primary risk factors for coronaryheart disease.

The sex differences in CVD-related mortality may bedue to a greater number of risk factors among womenwith diabetes mellitus or a greater effect of CVD risk factorson women with diabetes mellitus.

An additionalreason may be that CVD risk factors are less aggressivelymanaged in women with diabetes mellitus comparedwith their male counterparts.

Researchers have identifiedsex disparities in the use of drug therapy for dyslipidemia;however, it is not clear if these disparitiesrepresent an actual sex bias in prescribing.

Guidelines from the American Diabetes Associationrecommend annual screenings for dyslipidemia in allpersons with diabetes mellitus and aggressive managementof dyslipidemia, with drug therapy initiationthresholds that are similar for men and women.

Theproportion of managed care enrollees with type 2 diabetesmellitus who are receiving drug therapy for dyslipidemiahas increased substantially in recent years.

However, it is not clear whether women with diabetesmellitus are receiving similar treatment for dyslipidemiacompared with men with diabetes mellitus. Theobjectives of this study were 2-fold: (1) to determine ifmen with type 2 diabetes mellitus were more likely tobe screened or monitored for dyslipidemia than womenwith type 2 diabetes mellitus and (2) to determine ifmen with type 2 diabetes mellitus were more likely toreceive drug therapy for dyslipidemia than women withtype 2 diabetes mellitus.

This study used a retrospective cross-sectionaldesign examining medical and pharmacy claims datafrom the commercially insured population of a university-affiliated health maintenance organization in themidwestern United States. The study was approved bythe University of Michigan Institutional Review Board.Eligible subjects were at least 18 years of age, were continuouslyenrolled in the health plan during the entirestudy period, had more than 1 medical claim with a diabetesmellitus diagnosis, and had more than 1 claim foran oral diabetes mellitus drug. Patients using insulinwere excluded from the analyses to increase the homogeneityof the sample.

International Classification of Diseases, Ninth Revision,

Drug use data were extracted from the pharmacyclaims, while laboratory tests were identified from

codes within the medicalclaims. All the variables were recoded into categoricaldata for the purpose of statistical analysis. Adichotomous variable for lipid testing was coded as 1 ifthe patient had a medical claim containing a 

code for any lipid test (eg,total cholesterol, low-density lipoprotein, or lipid panel)and as 0 if the patient did not have any claims for thesetests. The lipid drug variable was coded as 1 if thepatient had at least 1 claim for any prescription lipid-modifyingdrug and as 0 if there were no claims forthese drugs. Sex was coded as 1 for male and as 0 forfemale. Age in 2000 was stratified into 4 age categories(18-44, 45-54, 55-64, and &#8805;65 years) for comparisonacross the age groups. A dichotomous variable was constructedto indicate whether the patient had a CVDdiagnosis during 2000 or 2001. The variable was codedas 1 if the patient had at least 1 medical claim with an

code for selected CVDs(an appendix listing these codes is available from theauthor). Finally, 2 proxies for severity of illness werecreated: (1) a hospitalization event in 2000 (yes or no)and (2) multiple-drug therapy versus single-drug therapyfor diabetes mellitus.

Descriptive analyses were conducted to assess thedistribution of each variable. Bivariate comparisons oflipid testing and lipid-modifying drug use across sex categorieswere conducted using cross-tabulations and &#967;

tests. Multivariate modeling of data was done usinglogistic regression models to assess the relationship ofsex with lipid testing and lipid-modifying drug use whileadjusting for covariates such as age, CVD diagnosis, hospitalization,and multiple-drug therapy. All statisticalanalyses were performed using SPSS version 11.0 (SPSSInc, Chicago, Ill).

There were 1729 patients identified in 2000 and2001 who met the inclusion criteria. Of these, 50.1%were women and 49.9% were men. The mean age forwomen was 49 years, while the mean age for men was52 years. Among all patients, 1429 (82.6%) received alipid test in 2000 or 2001. In 2000 and 2001, 675patients (39.0%) received a lipid-modifying drug.Descriptive data for the entire sample are found in

Descriptive data regarding sex differences in the useof lipid tests and lipid-modifying drugs were evaluated.During 2000 and 2001, 79.4% of women received a lipidtest compared with 84.2% of men (&#967;

= .01).Also, 33.2% of women received a lipid-modifying drugcompared with 45.5% of men (&#967;

Logistic regression was used to control for severalvariables that could confound the results. For model 1(dependent variable, lipid test in 2000-2001), the controlvariables included the subject's age, hospitalizationin 2000, multiple-drug therapy, and a CVD diagnosis(

). When controlling for these variables, menwere 1.45 times (95% confidence interval [CI], 1.13-1.81 times) more likely than women to receive a lipidtest during 2000 or 2001.

For model 2 (dependent variable, lipid-modifyingdrug use in 2000-2001), the control variables includedthe subject's age, hospitalization in 2000, multiple-drug therapy, a CVD diagnosis, and a lipid testin 2000 or 2001 (

). When controlling forthese variables, men were 1.51 times (95% CI,1.22-1.86 times) more likely than women toreceive a lipid-modifying drug during 2000 or2001.

Women with type 2 diabetes mellitus were lesslikely than men with type 2 diabetes mellitus toreceive lipid tests or lipid-modifying drugs whencontrolling for age, CVD diagnosis, and 2 proxiesfor severity of illness (hospitalization and diabetesmellitus drug intensity). The reasons forthese differences could not be determined fromclaims data alone, and the sex disparity in lipidtesting and treatment may reflect clinical differencesin the patients. However, prior researchhas shown that women with diabetes mellitus areat a higher risk of cardiovascular complicationsthan men with diabetes mellitus; thus, the disparitiesin testing and treatment warrant furtherinvestigation.

revealed that diabeticpatients with coronary artery disease are lesslikely to receive lipid testing or lipid-modifyingtherapy than nondiabetic persons with coronaryartery disease. Additional studies have identifiedthat women are less likely to receive diagnosticprocedures

for CVD.Although women with diabetes mellitus appear tohave an elevated risk of cardiovascular morbidityand mortality compared with men with diabetesmellitus,

the present study reveals that womenare also less likely to receive lipid tests or lipid-modifyingdrugs. Therefore, women with diabetesmellitus may be at high risk for undertreatmentof CVD.

The reasons for the sex disparity in testingand treatment for CVD in patients with diabetesmellitus are not clear. It is possible that thelower rates of treatment for dyslipidemia inwomen with diabetes mellitus occur for severalreasons. Women, in general, may not be highly concernedabout heart disease and may not raise the issuewith their physicians.

Women may also experiencesymptoms of CVD that are different from those of men,and these "atypical"symptoms may not be discussedin a clinical encounter or may not prompt further evaluationof cardiovascular causes of the symptoms.

Middle-aged women with diabetes mellitus, or their clinicians,might also place higher priority on treatingthe symptoms of other health problems (eg, menopause).If women place higher priorities on noncardiovascularissues in discussions with their clinicians, it ispossible that cardiovascular concerns may not receiveadequate attention.

The aforementioned explanations for sex disparitymay be intertwined with the general problem of clinicalinertia (ie, failure to prescribe or intensify therapywhen indicated).

Therefore, efforts to combat clinicalinertia (eg, physician feedback and reminder systems)could help to alleviate the sex disparities in monitoringor treating dyslipidemia in women with diabetes mellitus.

Specifically, reminders that prompt physicians toorder lipid tests for patients with diabetes mellitus maybe particularly helpful for women if sex disparities haveoccurred because of a greater number of competing prioritiesin clinical encounters with women. Flagging elevatedlow-density lipoprotein levels may also help tostimulate prescribing of lipid-modifying therapy whenappropriate.

have suggested that sex disparitiesare an important concern in quality of care formanaged care organizations. They suggest that healthplans consider sex-stratified reporting of performancemeasures and feedback to physicians. The Women'sHealth Measurement Advisory Panel of the NationalCommittee for Quality Assurance recommends thatresearch be done to examine sex differences in healthplan performance on conditions that are highlyprevalent among men and women (eg, diabetesmellitus).

Further education of physicians onthe high risk of cardiovascular mortality inwomen with diabetes mellitus may be warranted.

It is important to note that the data used in thisstudy were derived from only the commerciallyinsured population of one health plan. Althoughthe health plan represents a large, diverse groupof patients, caution should be used in generalizingthe findings beyond the managed care population.In addition, the claims data set did not includelaboratory test results (eg, low-density lipoproteinlevels), so it is difficult to evaluate the appropriatenessof the prescribing patterns for lipid-modifyingdrugs. However, the analyses controlled forthe presence of medical claims with CVD diagnoses,age, and 2 proxies for severity of illness.

Women with type 2 diabetes mellitus were lesslikely than men with type 2 diabetes mellitus toreceive lipid tests or lipid-modifying drugs. Giventhe higher rate of CVD mortality among womenwith diabetes mellitus compared with men withdiabetes mellitus, the lower rate of lipid testingand lipid-modifying drug therapy for women warrantsfurther attention.

We appreciate the comments of William Herman, MD, and ThomasSpafford, MS, on early drafts of the manuscript, as well as the cooperationof M-CARE, University of Michigan, in providing the data used in this study.

From the Center for Medication Use, Policy & Economics, Department of Social andAdministrative Sciences, University of Michigan College of Pharmacy, Ann Arbor.

This work was supported by the University of Michigan Health System.Preliminary analyses from this study were presented at the 9th Annual InternationalMeeting of the Society for Pharmacoeconomics and Outcomes Research; May 18, 2004;Arlington, Va.

Address correspondence to: David P. Nau, PhD, Center for Medication Use, Policy& Economics, Department of Social and Administrative Sciences, University of MichiganCollege of Pharmacy, 428 Church Street, Ann Arbor, MI 48109-1065. E-mail:dnau@umich.edu.

1. Centers for Disease Control and Prevention Web site. 

Data & trends: diabetessurveillance system: prevalence of diabetes: number of persons with diagnosed diabetes,United States, 1980-2002. Available at: http://www.cdc.gov/diabetes/statistics/prev/national/figpersons.htm. Accessed July 5, 2004.

2. Hogan P, Dall T, Nikolov P; American Diabetes Association. 

Economic costs ofdiabetes in the US in 2002. 

Estimates of direct medical costs for microvascularand macrovascular complications resulting from type 2 diabetes mellitus in theUnited States in 2000. 

4. Haffner SM, Lehto S, Ronnemaa T, Pyorala K, Laakso M. 

Mortality from coronaryheart disease in subjects with type 2 diabetes and in non-diabetic subjectswith and without prior myocardial infarction. 

David P. Nau, PhD

Articles

Recommendations for Improving Adherence to Type 2 Diabetes Mellitus Therapy-Focus on Optimizing Oral and Non-Insulin Therapies

Sex Disparity in the Management of Dyslipidemia Among Patients With Type 2 Diabetes Mellitus in a Managed Care Organization