Journal of Medical Economics

Article 0118.R1/862538 All rights reserved: reproduction in whole or part not permitted

Bradley Curtis

Global Health Outcomes, Eli Lilly and Company, Indianapolis, IN, USA

Maureen J. Lage

HealthMetrics Outcomes Research, LLC, Bonita Springs, FL, USA

Address for correspondence: Maureen J. Lage, PhD, HealthMetrics Outcomes Research, LLC, 27576 River Reach Dr, Bonita Springs, FL 34134, USA. Tel.: (860) 245-0685; [email protected]

Keywords: Diabetes – Type 2 – Basal insulin – Glycemic control – Electronic medical records database Accepted: 1 November 2013; published online: 21 November 2013 Citation: J Med Econ 2014; 17:21–31

Abstract

,

se oa d

No p Un t yr au fo ig th r ht di or S sp ize a la d le © y, us 20 vi e o ew p r 13 r C o an h d p ibi om In rin ted m fo rm t a . Au e si th rc aU ng or i le is al K co ed D py us is Lim fo ers tr ite rp c i b a er n d so d ut na ow io l u nl n

Original article Glycemic control among patients with type 2 diabetes who initiate basal insulin: a retrospective cohort study

Co

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

1369-6998 doi:10.3111/13696998.2013.862538

Vol. 17, No. 1, 2014, 21–31

Objective: To examine changes in glycemic control for patients with type 2 diabetes mellitus (T2DM) after initiation of basal insulin and factors associated with improved glycemic control. Methods: An analysis of retrospective medical records of patients with T2DM was examined using Humedica’s electronic medical records database (January 2007–August 2012). Patients with T2DM, initiating basal insulin, age 21 years, with a recorded HbA1c test in both the 1 year prior and the 2 years post-initiation were included. A multivariate regression examined factors associated with changes in glycemic control. Logistic regressions examined factors associated with improvements or worsening of glycemic control, compared to relatively unchanged glycemic control. Results: Many (14,457) individuals met the inclusion–exclusion criteria. Multivariate analyses revealed that older age (p50.0001), residence in the Western region of the US (vs South) (p50.0001), Medicare insurance vs Medicaid or being uninsured (p ¼ 0.0138), and higher household income (p ¼ 0.0065) were associated with improved glycemic control. Patients diagnosed with comorbid peripheral vascular disease (p ¼ 0.0072), cancer (p ¼ 0.0019), obesity (p ¼ 0.0002), moderate (p ¼ 0.0103), and severe chronic kidney disease (p50.0001), or end-stage renal disease (p ¼ 0.0075) in the pre-period were found to have significantly improved glycemic control in the post-period. Use of prandial insulin (p ¼ 0.0087), pre-mix insulin (p ¼ 0.0003) in the pre-period, a higher pre-period HbA1c score (p50.0001), and longer duration between pre-period and post-period HbA1c testing (p50.0001) were significantly associated with higher HbA1c levels in the post-period. Limitations: Analyses rely on electronic medical records which cannot capture patient healthcare utilization occurring outside of the data capture system. Analyses do not control for insulin dosage or type of basal insulin prescribed. Conclusions: Among patients with T2DM treated with basal insulin, a number of factors may influence glycemic outcomes. These findings suggest a role for a more personalized approach to the treatment of patients with T2DM.

Abbereviations: CCI, Charlson Comorbidity Index; CHF, Congestive Heart Failure; CKD, Chronic kidney disease; ESRD, End-stage renal disease; HbA1c, hemoglobin A1c; HIPAA, Health Insurance Portability and Accountability Act; T2DM, type 2 diabetes mellitus

! 2014 Informa UK Ltd www.informahealthcare.com/jme

Introduction In the US, type 2 diabetes mellitus (T2DM) affects nearly 26 million individuals and burdens the economy with $176 billion in annual direct medical costs1. Initiation on basal insulin and glycemic control Curtis & Lage

21

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

Journal of Medical Economics

Volume 17, Number 1

January 2014

Much of the humanistic and economic burden of T2DM derives from its complications, which have been shown in randomized clinical trials to be preventable through the use of intensive glucose-lowering therapies2,3. Based on this clinical research, diabetes guidelines have recommended a standardized treatment goal of hemoglobin A1c (HbA1c) 57.0% for T2DM patients4–8. It is recommended that insulin therapy should be added to oral antidiabetic drug (OAD) treatments for patients who are unable to achieve HbA1c 57.0% with OADs alone6.9. An intermediate or long-acting baseline (basal) insulin is typically the first type of insulin treatment, although patients may additionally require mealtime (prandial) therapy with short- or rapid-acting insulin6. Research has shown that insulin therapy improves diabetes symptoms10 and delay of insulin initiation may lead to a large number of diabetes-related complications11. Despite this evidence data have indicated that the majority of patients with T2DM who initiate insulin therapy are unable to achieve the standardized HbA1c target of 57.0%12,13. Further, a majority of patients who intensify insulin treatment by adding a prandial insulin or switched to a prandial or pre-mixed insulin therapy failed to attain levels of HbA1c 57.0%12. Recently, T2DM treatment guidelines have acknowledged that a variety of factors can affect an individual’s ability to reach the standardized treatment goal of HbA1c 57.0%, stating that a higher HbA1c target of 7.5–8.5% may be appropriate for certain patients14,15. These recent guidelines promote a patient-centered or personalized approach to the management of T2DM14,15. Caring for patients in a personalized or patient-centered way has been defined as ‘providing care that is respectful of and responsive to individual patient preferences, needs, and values, and ensuring that patient values guide all clinical decisions’16 (p. 6). For such a personalized approach to be successful, healthcare providers need more data on which specific patient characteristics influence treatment outcomes15,17. Accordingly, this retrospective analysis used data from routine clinical care settings to identify the factors associated with changes in glycemic control after the start of basal insulin therapy and the drivers of better or worse glycemic control in the post-initiation period.

Patients and methods An analysis of retrospective medical records of patients with T2Dm were examined using Humedica’s de-identified Electronic Medical Record (EMR) database. Humedica partners directly with large medical group practices, integrated delivery networks (IDNs), and hospital chains to extract data from their EMRs and various health information technology systems. A sub-set of the data components include laboratory results, radiology and 22

Initiation on basal insulin and glycemic control Curtis & Lage

pathology reports, physician and nurse notes, prescriptions written and dispensed, procedures, diagnoses, and other details of a patient’s office visit and hospital stay. At the time of this study, Humedica’s clinical data warehouse encompassed 12 million patients, all de-identified to comply with Health Insurance Portability and Accountability Act (HIPAA) regulations. The data for this study spanned the period from January 1, 2007– August 30, 2012. To be included in this study, an individual had to have received a prescription for basal insulin during the index period between January 1, 2008–September 1, 2010; the date of the first prescription was identified as the index date. Patients were also required to have 1 year of EMR data available prior to the index date (i.e. in the preperiod) and 2 years of data available after the index date (i.e. in the post-period). Patients were required to have received at least two diagnoses of T2DM (ICD-9-CM 250.x0 or 250.x2) from the start of the pre-period through the end of the post-period and to have received at least one HbA1c test in both the pre- and post-periods. Patients were excluded from the study if they received a diagnosis of type 1 diabetes (ICD-9-CM 250.x1 or 250.x3) or gestational diabetes (ICD-9-CM 648.0) over the study period. Finally, patients were required to be at least 21 years of age on the index date. The study inclusion/exclusion criteria resulted in a sample of 14,457 individuals. Figure 1 illustrates how each inclusion/exclusion criterion affected sample size. The study focused on glycemic control in both the preperiod and post-period, with multivariate analyses used to examine post-period HbA1c level and changes in HbA1c from the pre- to the post-period. For the latter analysis, discretization of the HbA1c change was accomplished by partitioning18 the data so that the patients were grouped into three mutually exclusive categories, based upon their change in HbA1c level. Consistent with non-inferiority margins used in treat-to-target studies, patients were identified as having ‘no change’ in HbA1c using a range from 0.4% to 0.4%19,20. Consistent with this categorization, patients who had a reduction in HbA1c greater in absolute value than 0.4% were considered to have improved glycemic control, while those with an increase greater than 0.4% were considered to have worsened glycemic control. To assess factors that may influence glycemic control, multivariate analyses examined the associations between the post-period HbA1c values or changes in HbA1c, respectively, and patient characteristics, patient general health, comorbid diagnoses, other medication use during the pre-period, renal impairment, initial HbA1c score, and the timing between pre- and post-period HbA1c scores. A patient’s characteristics included age, sex, race, region, and insurance coverage. In addition, the percentage of college graduates and average household income within the 3-level zip code of patient residence were also included www.informahealthcare.com/jme ! 2014 Informa UK Ltd

Journal of Medical Economics

Receipt of a prescription for basal insulin between 1/1/08 and 9/1/10 N=120,395

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

EMR data available 1 year prior to index date and 2 years post index date N=35,970

Receipt of at least 2 diagnoses of T2DM from start of pre-period through end of post-period N=30,727

Receipt of HbA1c test in both preperiod and post-period N=19,207 N=19207

No receipt of diagnosis of T1DM or gestational diabetes N=14,466 ,

Age ≥ 21 Index Date N=14,458

Figure 1. Inclusion–exclusion criteria and sample size.

in the analyses. General health was proxied using the Charlson Comorbidity Index (CCI)21,22. Comorbid diagnoses identified in the pre-period included diabetic retinopathy, diabetic neuropathy, coronary artery disease, peripheral vascular disease, congestive heart failure (CHF), depression, cancer, and obesity. Pre-period medication use included receipt of a prescription for prandial insulin, pre-mixed insulin, an OAD, or a GLP-1 agonist. The degree of renal impairment was proxied by the estimated glomerular filtration rate (eGFR), with patients categorized as having no, mild, moderate, or severe chronic kidney disease (CKD), or end stage renal disease (ESRD). Descriptive statistics (mean and standard deviation for continuous variables and frequency and percentages for categorical variables) were used to characterize the sample. T-tests and Chi-square tests were utilized to assess differences between the cohorts, and p-values were adjusted for multiple comparisons using a Bonferroni correction. Logistic regressions were used to examine the likelihood of having either improved or worsened HbA1c in comparison to having relatively unchanged HbA1c. In addition, an ordinary least squares (OLS) regression was used to examine the association between the factors discussed above and post-period HbA1c score. All analyses ! 2014 Informa UK Ltd www.informahealthcare.com/jme

Volume 17, Number 1

January 2014

were conducted using SAS, version 9.2. A p-value 50.05 was considered to be statistically significant.

Results Table 1 presents the descriptive statistics for the entire cohort. Results revealed that the largest percentage of patients were 55–64 years old (30.5%), Caucasian (63.4%), and residents of the South (45.8%). Patients were approximately equally divided between males (50.2%) and females (49.8%). A majority of patients (71.4%) received an OAD agent prior to initiation on basal insulin, and patients were most commonly diagnosed with comorbid hypertension (69.7%), coronary artery disease (19.7%), and depression (12.0%) in the pre-period. The average HbA1c score prior to initiation on basal insulin was 8.6% (SD ¼ 2.35). In the 2-year post-period, 29.2% of patients achieved HbA1c 57%. While Table 1 shows descriptive characteristics for the entire cohort, Table 2 compares these variables based upon change in HbA1c from the pre-period to the post-period, and reveals significant differences between the categories. For instance, there were statistically significant differences in the age distributions between improved vs unchanged, improved vs worse, and unchanged vs worse glycemic control categories (all p50.05). Those in the improve category, compared to the other two categories, were more frequently 565 years old, while individuals in the unchanged category were more frequently 65 years of age. Similarly, there were statistically significant differences among all three cohorts with regards to race and region of residence. For example, African-Americans and residents of the South were found to be more likely to have worse glycemic control, while Caucasians were more likely to have unchanged glycemic control. Results from Table 2 also reveal that patients with improved glycemic control were significantly different from the unchanged or worse cohorts with regards to insurance type, with improve-category patients more likely to have Commercial insurance. In contrast, those with worse glycemic control, compared to those with improved glycemic control, were more likely to be female, diagnosed with congestive heart failure, and lower in average household income. The multivariate OLS analysis shown in Table 3 presents the factors associated with post-period HbA1c levels, including adjustment for pre-period HbA1c score. The OLS regression results showed an association between higher post-period HbA1c levels and pre-period use of prandial or pre-mixed insulin, African-American race, and higher pre-period HbA1c scores. Specifically, patients who received a prandial insulin had a 0.27 higher (0.0003) HbA1c score compared to patients who did not receive a prandial, while African-Americans had a 0.18 higher Initiation on basal insulin and glycemic control Curtis & Lage

23

Journal of Medical Economics

Volume 17, Number 1

January 2014

Table 1. Descriptive statistics.

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

n or Mean Patient characteristics Age group 21–44 1550 45–54 3035 55–64 4413 65–71 2340 72þ 3120 Sex Female 7197 Male 7261 Race African American 2444 Asian 184 Caucasian 9168 Other/Unknown 2662 Region Midwest 5034 Northeast 604 Other/Unknown 15 South 6622 West 2183 Insurance type Commercial 3170 Medicaid 244 Medicare 2622 Other payor type 192 Uninsured 804 Unknown 7426 Household income (mean; SD)* 42,906.20 % College Educated (mean; SD)* 23.44 Pre-period anti-diabetic therapy use Prandial Insulin 1040 Pre-Mix Insulin 501 Orals 10,328 GLP-1 Agonist 1189 Pre-period comorbidities Diabetic Retinopathy 625 Diabetic Neuropathy 1427 Coronary Artery Disease 2851 Myocardial Infarction 590 Peripheral Vascular Disease 934 Congestive Heart Failure 1241 Hypertension 10,075 Depression 1733 Cancer 824 Obesity 714 Chronic Kidney Disease 862 Pre-period general health CCI (mean; SD) 0.67 Pre-period glycemic control HbA1c (mean; SD) 8.59 Pre-period CKD status (based on eGFR) Normal 7241 Mild 3833 Moderate 1977 Severe 448 ESRD 234 Missing 720

% or SD

10.72 20.99 30.52 16.18 21.58 49.78 50.22 16.90 1.27 63.41 18.41 34.82 4.18 0.10 45.80 15.10 21.93 1.69 18.14 1.33 5.56 51.36 11,229.54 7.71 7.19 3.47 71.43 8.22 4.32 9.87 19.72 4.08 6.46 8.58 69.68 11.99 5.70 4.94 5.96 1.18 2.35 50.08 26.55 13.67 3.10 1.62 4.98

*Based upon 3-digit zip code of residence.

(p50.0001) HbA1c relative to Caucasians, and receipt of prandial insulin, compared to patients who did not receive prandial insulin had 0.14 higher (p ¼ 0.0087) HbA1c scores. In contrast, the findings showed an association between a lower HbA1c level and age445 years; residence 24

Initiation on basal insulin and glycemic control Curtis & Lage

in the West; Medicare insurance; higher household income; or a pre-period diagnosis of obesity, peripheral vascular disease, cancer, moderate CKD, severe CKD, ESRD. For example, compared to patients aged less than 45, those aged 72 or older had 0.75 lower HbA1c scores (p50.0001), while patients diagnosed with CKD in the pre-period had 0.39 lower HbA1c scores (p50.0001). Table 4 presents the results of the logistic regressions, which identified factors independently associated with improved or worsened, as compared to relatively unchanged, HbA1c between the pre- and post-periods. Older age was associated with a lower likelihood of worse glycemic control, as was a diagnosis of myocardial infarction in the pre-period. Specifically, patients aged 55–64, compared to those aged 44 or younger, were 32.6% less likely to have their glycemic control change to the worse category (Odds Ratio [OR] ¼ 0.674; 95% Confidence Interval [CI] ¼ 0.561–0.809], while patients aged 72 or older, compared to those aged 44 or less, were 41.4% less likely to have their glycemic control change to the worse category [OR ¼ 0.526; 95% CI ¼ 0.423–0.653]. In contrast, African-Americans were found to be 18% more likely to have worse glycemic control [OR ¼ 1.183; 95% CI ¼ 1.030–1.358]. When examining factors associated with improved glycemic control compared to unchanged glycemic control, results revealed that older individuals were more likely to have improved glycemic control, as were patients who resided in the West, had a diagnosis of obesity in the pre-period, or had moderate chronic kidney disease, severe chronic kidney disease, or end-stage renal disease. For example, patients aged 72 or older, compared to those aged 44 or younger, were 37.4% (OR ¼ 1.374; 95% CI ¼ 1.106–1.708) more likely to have improved glycemic control, while patients diagnosed with ESRD were 69.4% (OR ¼ 1.694; 95% CI ¼ 1.163–2.469) more likely to have improved glycemic control. Patient sex, insurance coverage type, household income, education level, Charlson score, and pre-period non-basal insulin antidiabetic therapy use were not significantly associated with improvements or degradations in glycemic control.

Discussion This analysis utilizes a large EMR database to examine glycemic control among patients diagnosed with T2DM who initiate therapy with basal insulin. The results indicate that levels of glycemic control differ significantly from more rigorously controlled clinical trials23–25. Given the move towards a ‘payment for outcomes’ as opposed to the current ‘fee for service’ model of healthcare delivery and accountability26, it will be important to generate real world evidence-based models as to what factors are consistently associated with improved outcomes, and in which sub-groups of patients. www.informahealthcare.com/jme ! 2014 Informa UK Ltd

Journal of Medical Economics

Volume 17, Number 1

January 2014

Table 2. Descriptive statistics, by change in HbA1c categories. Improve (n ¼ 7619)

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

n or Mean Patient characteristics Age 21–44 45–54 55–64 65–71 72þ Sex Female Male Race African American Asian Caucasian Other/Unknown Region Midwest Northeast Other/unknown South West Insurance type Commercial Medicaid Medicare Other payor type Uninsured Unknown Household income (mean; SD)* % College Educated (mean; SD)*

% or SD

Unchanged (n ¼ 3077) n or Mean

% or SD

Worse (n ¼ 3761) n or Mean

p-Value

% or SD

A,B,C 816 1690 2405 1167 1541

10.71 22.18 31.57 15.32 20.23

272 508 927 571 799

8.84 16.51 30.13 18.56 25.97

461 837 1081 602 780

12.26 22.25 28.74 16.01 20.74

3727 3892

48.92 51.08

1518 1559

49.33 50.67

1952 1809

51.90 48.10

1278 92 4843 1406

16.77 1.21 63.56 18.45

458 53 2013 553

14.88 1.72 65.42 17.97

707 39 2312 703

18.80 1.04 61.47 18.69

2674 350 6 3327 1272

35.10 4.59 0.08 43.67 16.56

1085 111 5 1429 447

35.26 3.61 0.16 46.44 14.53

1274 143 4 1866 474

33.87 3.80 0.11 49.61 12.60

B,C A,B,C

A,B,C

A,B 1817 127 1350 115 454 3756 43,226.81 23.56

23.85 1.67 17.72 1.51 5.96 49.30 11,237.87 7.71

606 43 588 37 143 53.95 42,859.99 23.39

19.59 1.40 19.11 1.20 4.65 53.95 11,268.79 7.91

747 74 684 40 207 2009 42,293.34 23.25

19.86 1.97 18.19 1.06 5.50 53.42 11,158.40 7.55

Pre-period anti-diabetic therapy use Prandial Insulin Pre-Mix Insulin Orals GLP-1 Agonist

508 248 5710 627

6.67 3.26 74.94 8.23

248 117 2063 254

8.06 3.80 67.05 8.25

284 136 2555 308

7.55 3.62 67.93 8.19

Pre-period comorbidities Diabetic Retinopathy Diabetic Neuropathy Coronary Artery Disease Myocardial Infarction Peripheral Vascular Disease Congestive Heart Failure Hypertension Depression Cancer Obesity Chronic Kidney Disease

331 762 1429 296 485 608 5318 881 446 394 420

4.34 10.00 18.76 3.89 6.37 7.98 69.80 11.56 5.85 5.17 5.51

130 295 682 155 208 258 2155 384 184 128 198

4.22 9.59 22.16 5.04 6.76 8.38 70.04 12.48 5.98 4.16 6.43

164 370 740 139 241 375 2602 468 194 192 244

4.36 9.84 19.68 3.70 6.41 9.97 69.18 12.44 5.16 5.11 6.49

Pre-period general health CCI (mean; SD) Glycemic control Pre-Period HbA1c (mean; SD) Days between pre- and post-period HbA1c (mean; SD) Pre-period eGFR Normal Mild Moderate Severe ESRD Missing

0.66

1.19

0.67

1.19

0.67

1.16

9.62 635.51

2.07 172.77

7.79 653.58

1.51 178.25

7.86 663.03

1.57 175.15

B,C B A A B A

A,C A,C B,C

A B

A,B A,B,C A,C

3947 1971 988 229 113 371

51.80 25.87 12.97 3.01 1.48 4.87

1420 898 450 104 51 154

46.15 29.18 14.62 3.38 1.66 5.00

1874 969 539 115 70 194

49.83 25.76 14.33 3.06 1.86 5.16

A, statistically significant difference (p50.05) between Improved and Unchanged cohorts; B, statistically significant difference (p50.05) between Improved and Worse cohorts; C, statistically significant difference (p50.05) between Unchanged and Worse cohorts. *Based upon 3-digit zip code of residence. ! 2014 Informa UK Ltd www.informahealthcare.com/jme

Initiation on basal insulin and glycemic control Curtis & Lage

25

Journal of Medical Economics

Volume 17, Number 1

January 2014

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

Table 3. Associations with post-period glycemic control.

Intercept Age 45–54a Age 55–64a Age 65–71a Age 72þa Femaleb African-Americanc Other/Unknown Racec Northeastd Midwestd Westd Commercial Insurancee Medicaree Other/Unknown Insurancee Household Income* % College Educated CCI Diabetic Retinopathyf Diabetic Neuropathyf Coronary Artery Diseasef Myocardial Infarctionf Peripheral Vascular Diseasef Congestive Heart Failuref Hypertensionf Depressionf Cancerf Obesityf Prandial Insulin Usef Pre-Mix Insulin Usef Oral usef GLP-1 Agonist Usef Pre-Period HbA1c Months Pre to Post HbA1c Pre-Period Mild CKDg Pre-Period Moderate CKDg Pre-Period Severe CKDg Pre-Period ESRDg Pre-Period Missing CKDg

Coefficient

p-Value

5.6333 0.3196 0.5821 0.6630 0.7544 0.0409 0.1930 0.0180 0.0828 0.0010 0.1948 0.0359 0.0953 0.1297 0.0064 0.0044 0.0297 0.0449 0.0409 0.0204 0.1266 0.1628 0.0408 0.0104 0.0163 0.2446 0.2320 0.1375 0.2671 0.0301 0.0507 0.3333 0.0003 0.0425 0.1179 0.3857 0.2907 0.0387

50.0001 50.0001 50.0001 50.0001 50.0001 0.1372 50.0001 0.8813 0.2997 0.9766 50.0001 0.3068 0.0138 0.1042 0.0065 0.1426 0.1701 0.5244 0.4815 0.6076 0.1080 0.0072 0.4688 0.7373 0.6993 0.0019 0.0002 0.0087 0.0003 0.3321 0.3092 50.0001 50.0001 0.2476 0.0103 50.0001 0.0075 0.5429

*Household income in tens of thousands. a reference category Age 21–45; breference category Males; creference category Caucasian; dreference category South; ereference category Medicaid or Uninsured; freference category no receipt of diagnosis or drug of interest; g reference category no CKD. Italics represent statistically significant variables (p50.05).

Consistent with previous retrospective studies assessing glycemic control after initiation of basal insulin therapy12,13, this analysis of patients with T2DM revealed that only a minority (29.2%) achieved HbA1c 57% during the 2-year follow-up period. This suggests that, given current approaches to care, the standardized blood glucose goal of HbA1c 57%4–8 may be unachievable for many patients with T2DM, even after initiation of basal insulin. While such research is inconsistent with clinical trials such as the APOLLO23, 4T24, or LAPTOP25, with a higher proportion of patients achieving glycemic control, it should be recognized that these clinical trials do not reflect real life clinical care (e.g., narrow inclusion criteria, treat-to-target design, close monitoring, and regular follow-up). 26

Initiation on basal insulin and glycemic control Curtis & Lage

Furthermore, it should be noted that the vast majority of patients did not have their HbA1c tested as frequently as recommended by the ADA. Specifically, the ADA recommends that HbA1c be tested at least twice per year for patients who are meeting treatment goals and quarterly for patients who have changed therapy or who are not meeting glycemic goals27. However, in this study, only 5.89% of individuals in this database received at least 12 HbA1c tests in the 3 years of the study and only 37.04% received at least six HbA1c tests in the same time horizon.

Age The age-related findings from this study support earlier research that ‘different therapeutic approaches may be required to treat hyperglycemia effectively in different age groups’28 (p. 535). In particular, the evidence suggests that younger adults (aged 545) have unique treatment needs, as they were more likely than older individuals to have poorer glycemic control in the post-period, and to have worse control after starting basal insulin therapy. This supports previous studies29,30, which focused specifically on age-related differences that reported ‘Younger age was independently associated with: greater chronic stress and negative life events, higher levels of diabetes-related distress, higher depressed affect, eating healthier foods and exercising less frequently, lower diabetes self-efficacy, and higher HbA1c’30 (p. 154). Another study, which looked at slightly older patients (40–84 years) found that ‘older patients were more likely to be in the target range of glycemic control’31 (p. 276). While the reasons for poorer outcomes among younger adults are unclear and warrant more research, previous survey data have suggested that the youngest individuals (aged 525 years) were more likely than older persons to engage in higher-risk healthrelated behaviors, including ‘smoking cigarettes, binge drinking and heavy drinking’32 (p. 1). Recent literature has suggested that the lower likelihood of improvement in the oldest category of patients as compared to their middle-aged counterparts may relate to metabolic differences among the elderly population28. Specifically, a higher proportion of the overall hyperglycemia of the elderly patients derives from prandial, as opposed to basal, dis-regulation28. Such findings imply that the oldest (aged 65 years) sub-set of T2DM patients constitute a unique treatment cohort.

Baseline glycemic control Patients in the present study with worse glycemic control in the pre-period were more likely to improve after starting on basal insulin. Whereas prospective, clinical research in Germany has indicated that patients with good baseline glycemic control have minimal deterioration of glycemic www.informahealthcare.com/jme ! 2014 Informa UK Ltd

Journal of Medical Economics

Volume 17, Number 1

January 2014

Table 4. Predictors of changes in glycemic control.

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

Worse vs Unchanged

Age 45–54a Age 55–64a Age 65–71a Age 72þa Femaleb African-American Blackc Other/Unknown Racec Northeastd Midwestd Westd Commercial Insurancee Medicaree Other/Unknown Insurancee Household Income* % College Educated CCI Diabetic Retinopathyf Diabetic Neuropathyf Coronary Artery Diseasef Myocardial Infarctionf Peripheral Vascular Diseasef Congestive Heart Failuref Hypertensionf Depressionf Cancerf Obesityf Prandial Insulin Usef Pre-Mix Insulin Usef Oral usef GLP-1 Agonist Usef Pre-Period HbA1c Months Pre to Post HbA1c Pre-Period Mild CKDg Pre-Period Moderate CKDg Pre-Period Severe CKDg Pre-Period ESRDg Pre-Period Missing CKDg

Improved vs Unchanged

Odds ratio

95% CI

Odds ratio

95% CI

0.958 0.674 0.582 0.526 1.086 1.183 0.640 1.133 0.939 0.865 0.913 1.040 1.023 0.993 1.008 1.050 1.059 0.981 0.965 0.681 0.987 1.331 1.000 0.948 0.861 1.112 0.936 0.948 1.059 0.922 0.994 1.000 0.989 1.107 0.941 1.137 1.012

0.793–1.158 0.561–0.809 0.470–0.720 0.423–0.653 0.982–1.200 1.030–1.358 0.417–0.981 0.835–1.537 0.831–1.062 0.740–1.011 0.802–1.041 0.906–1.192 0.757–1.381 0.985–1.002 0.997–1.019 0.969–1.138 0.820–1.369 0.794–1.212 0.838–1.110 0.515–0.899 0.792–1.229 1.089–1.627 0.894–1.119 0.815–1.102 0.647–1.147 0.878–1.407 0.781–1.123 0.732–1.226 0.950–1.180 0.770–1.104 0.961–1.028 1.000–1.001 0.865–1.129 0.941–1.303 0.705–1.258 0.778–1.662 0.806–1.272

1.301 1.321 1.287 1.374 1.061 0.932 0.652 1.292 0.946 1.202 1.034 1.163 1.205 1.004 0.999 1.021 0.992 1.027 0.969 0.821 1.109 1.148 1.021 0.920 1.146 1.288 0.858 0.794 1.211 0.950 1.974 1.000 0.993 1.186 1.479 1.694 1.172

1.070–1.583 1.096–1.592 1.040–1.593 1.106–1.708 0.963–1.169 0.810–1.073 0.443–0.959 0.972–1.717 0.838–1.068 1.041–1.389 0.912–1.171 1.019–1.327 0.897–1.620 0.996–1.012 0.989–1.010 0.949–1.099 0.775–1.269 0.842–1.253 0.848–1.108 0.631–1.068 0.901–1.365 0.943–1.397 0.915–1.138 0.794–1.065 0.877–1.497 1.021–1.624 0.719–1.025 0.615–1.024 1.087–1.351 0.801–1.127 1.904–2.047 1.000–1.000 0.875–1.128 1.013–1.388 1.124–1.946 1.163–2.469 0.937–1.466

*Household income in tens of thousands. a reference category Age 21–45; breference category Males; creference category Caucasian; dreference category South; e reference category Medicaid or Uninsured; freference category no receipt of diagnosis or drug of interest; greference category no CKD. Italics represent statistically significant variables (p50.05).

control over multi-year follow-up periods33,34, other research has shown that patients with higher initial HbA1c levels were more likely to have significant HbA1c improvement over a 1-year time horizon35. While comparisons across studies should be made with caution given the considerable differences among patient populations and anti-diabetic treatments, there may be several reasons why patients with worse glycemic control at baseline are more likely to improve after treatment. Firstly, they have a clear incentive to do so; as stated by the Centers for Disease Control and Prevention: ‘Although the most appropriate target A1c levels to achieve optimal health impact might vary among persons, an A1c level of 9% constitutes a clearly modifiable, high level of risk that few, if any, persons with diabetes should be exposed to’36 (p. 32). Accordingly, the Healthy People ! 2014 Informa UK Ltd www.informahealthcare.com/jme

2020 objectives. . . ‘include a 10% reduction in the proportion of the diabetes population that has poor glycemic control (A1c 49%) as a target’36 (p. 32). In addition, patients with better blood-glucose control at baseline may be more likely to lose control after starting basal insulin therapy, as tight control is challenging to maintain37. Indeed, a previous large-scale prospective trial reported that an HbA1c level of 7% was the lowest that could reasonably be maintained among diabetic patients, even when their diabetes was treated intensively31. Notably, patients in our present study who had tighter glycemic control at baseline were more likely to have better glycemic regulation in the post-period, a finding consistent with the results of the large-scale (n ¼ 12,537) outcome reduction with an Initial Glargine Intervention (ORIGIN) trial38. These results suggest that the treatment Initiation on basal insulin and glycemic control Curtis & Lage

27

Journal of Medical Economics

Volume 17, Number 1

January 2014

of patients with glycemic control close to the target range of HbA1c 57% is complex and could benefit from more a patient-centered approach, as promoted in the recent literature14,15.

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

Region Relative to those residing in the South, patients in the West were more likely to have tighter post-period glycemic control and improved glycemic control after starting basal insulin. The South of the US has the highest rate of diagnosed diabetes in the US, while the West has the lowest39. This regional difference has previously been attributed to differences in lifestyle between the two regions, with the South shown to have a higher rate of sedentariness and obesity relative to other regions40, and with Western states reported to have lower obesity rates41 and better overall health ratings based on a variety of measures42,43. Sedentariness and other lifestyle factors are accepted risk factors for developing T2DM38,44. Similarly, previous research has shown a clear relationship between ‘absence of whole-body movement’ and ‘obesity, abnormal glucose metabolism, and the metabolic syndrome’, as well as a link between physical activity and better treatment outcomes for patients with T2DM45,46.

Race Consistent with previous research focusing on various anti-diabetic treatment regimens47–49, AfricanAmericans patients in our study tended to have poorer glycemic control in the post-period relative to the Caucasians, and they were more likely than the Caucasians to deteriorate in glycemic control after starting basal insulin therapy. Although the reasons for these results are unclear, one previous analysis conducted among veterans in the Southwest showed that all minorities, and particularly African-Americans, were given lower doses of insulin relative to non-Hispanic whites49. More recent literature has reported a link between tighter adherence and increased risk of death among AfricanAmericans, stating that: ‘In adherent medication users, HbA1c 57% predicted higher mortality in NHB [nonHispanic blacks] (HR, 1.18; 95% CI, 1.07–1.31), whereas HbA1c greater than 9.0% predicted higher mortality risk across all race/ethnic groups’50 (p. 74). This evidence clearly indicates an area for further research and better treatment.

Comorbid diagnoses Only one pre-period diagnosis, CHF, was found in this study to be associated with deteriorated glycemic control after initiation on basal insulin. Previous research has 28

Initiation on basal insulin and glycemic control Curtis & Lage

shown the relationship between heart failure, insulin use, and diabetes to be complex. Improvement in glycemic control has been established as an independent risk factor for the development of CHF among patients with T2DM51, and insulin therapy has been shown to increase the odds of developing, or being hospitalized due to CHF51–53. Large-scale randomized trials have shown that tight glycemic control has no cardiovascular benefit relative to standard control for patients with longstanding T2DM54–56, and the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial was ended early due to an increased cardiovascular death rate among patients in the intensive glycemic control group53. Although posthoc analysis has indicated that insulin dose was not an independent driver of the increased cardiovascular mortality in the ACCORD trial57, a preliminary study has recently indicated that an aggressive approach to glycemic control is not beneficial and may be injurious to diabetic patients undergoing elective cardiac surgery58. Just as the American Diabetes Association has called for ‘Studies specifically designed to estimate the potential effects of antidiabetic agents on CHF incidence’59 (p. 1883), it is clear that more research is needed to examine the effects of insulin use among T2DM patients who have a diagnosis of comorbid CHF. In contrast to CHF, a number of pre-period comorbid diagnoses were associated with better post-period glycemic control, a result which may imply a link between additional diagnoses and better medical care. In particular, patients with a pre-period diagnosis of obesity had a higher likelihood of improving in glycemic control after initiating on basal insulin, and they also tended to have better post-period glycemic control relative to patients without that diagnosis. These findings may indicate that patients diagnosed with obesity are more likely than those without the diagnosis to lose weight, since weight loss has been shown to be an independent driver of improved glycemic control among patients with T2DM60.

Prandial insulin use T2DM treatment guidelines state that basal insulin levels should be controlled before prandial or pre-mixed insulin is prescribed6. In support of this guideline, the present study demonstrated that those who were prescribed prandial or pre-mixed insulin in the pre-period had poorer glycemic control and were less likely to improve than those without such use. Lack of compliance with this treatment guideline may indicate a lack of expertise on the part of the healthcare provider, a condition which might independently contribute to the poorer glycemic outcomes and which warrants further research. www.informahealthcare.com/jme ! 2014 Informa UK Ltd

Journal of Medical Economics

Relative affluence

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

Patients who were more affluent, as evidenced by a higher household income or Medicare insurance (relative to Medicaid or uninsured status), were more likely to have tighter glycemic control in the post-period. This finding is consistent with previous examinations showing lower socioeconomic status61 or food insecurity62,63 to be associated with poorer glycemic control. These results may also reflect the benefits of Medicare’s extensive diabetes prevention, screening, and treatment programs39.

Volume 17, Number 1

January 2014

responsibility for the analyses and drafting of the manuscript. Both authors reviewed and approved the final manuscript. Declaration of financial/other relationships BC is an employee and stockholder of Eli Lilly and Company. MJL is Managing Member of HealthMetrics Outcomes Research, LLC, and was a contractor to Eli Lilly and Company on this project. Acknowledgments The authors thank Patricia Platt, a freelance medical writer, for assisting in the writing of this manuscript. These services were funded by Eli Lully and Company.

Limitations and conclusions As with any research, the findings presented here should be interpreted within the context of limitations of the study design. First, this analysis was conducted using an EMR database and the results do not capture any healthcare utilization which occurred outside of the data capture system. Second, retrospective analyses must rely upon diagnostic codes to identify patients, assess patient general health, and determine other values, whereas formal diagnostic assessments may be more reliable. For example, in this database, 25% of patients who received at least two diagnoses of T2DM also received at least one diagnosis of type 1 diabetes over the study period. Third, this study did not control for differences in basal insulin dosage, although changes in insulin dosage have been shown to affect the ability of patients to achieve target HbA1c64. Further, as an intent-to-treat analysis, the research does not account for how adherence to a basal insulin regimen may affect outcomes. Fourth, the database did not contain information about other factors, such as duration since first diagnosis, that may affect outcomes. Finally, the present research did not control for type of basal insulin, which has been shown significantly to influence patient outcomes other than glycemic control (e.g., body weight), although no previous research has indicated the superiority of one insulin type over another in treating to HbA1c targets65,66. In conclusion, the results of this study indicate that a number of factors may influence glycemic outcomes among T2DM patients who start basal insulin therapy in the US. These factors included age, baseline glycemic control, region of residence, race, certain comorbid diagnoses, and prior prandial insulin use. The results of this study suggest a need for a more personalized approach to the treatment of patients with T2DM, an approach which has been advocated in recent treatment guidelines14,15.

Transparency Declaration of funding Eli Lilly and Company was the sponsor of this study. BC had primary responsibility for the study design, while MJL had primary

! 2014 Informa UK Ltd www.informahealthcare.com/jme

References 1. American Diabetes Association. Economic costs of diabetes in the U.S. in 2012. Diabetes Care 2013;36:1033-46 2. Haffner SM, Lehto S, Ro¨nnemaa T, et al. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med 1998;339:229-34 3. 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 (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet 1998;352:837-53 4. Bergenstal RM, Bailey CJ, Kendall DM. Type 2 diabetes: assessing the relative risks and benefits of glucose-lowering medications. Am J Med 2010;123:374.e 9-18 5. Home P, Mant J, Diaz J, et al; Guideline Development Group. Management of type 2 diabetes: summary of updated NICE guidance. BMJ 2008;336:1306-8 6. Nathan DM, Buse JB, Davidson MB, et al. Medical management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy: a consensus statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care 2009;32:193-203 7. National Institute for Clinical Excellence: NICE. Management of Type 2 Diabetes. London: NICE, 2002 8. Rodbard HW, Jellinger PS, Davidson JA, et al. Statement by an American Association of Clinical Endocrinologists/American College of Endocrinology consensus panel on type 2 diabetes mellitus: an algorithm for glycemic control. Endocr Pract 2009;15:540-59 9. Holt RIG. Insulin initiation in Type 2 diabetes: the implications of the 4-T study. Diabet Med 2010;27:1-3 10. Hajos TRS, Pouwer F, de Grooth R, et al. Initiation of insulin glargine in patients with Type 2 diabetes in suboptimal glycaemic control positively impacts health-related quality of life. A prospective cohort study in primary care. Diabet Med 2011;28:1096-102 11. Goodall G, Sarpong EM, Hayes C, et al. The consequences of delaying insulin initiation in UK type 2 diabetes patients failing oral hyperglycaemic agents: a modelling study. BMC Endocr Disord 2009;9:19 12. Blak BT, Smith HT, Hards M, et al. Optimization of insulin therapy in patients with type 2 diabetes mellitus: beyond basal insulin. Diabet Med 2012;29:e13-20 13. Pollock RF, Erny-Albrecht KM, Kalsekar A, et al. Long-acting insulin analogs: a review of ‘‘real-world’’ effectiveness in patients with type 2 diabetes. Curr Diabetes Rev 2011;7:61-74 14. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of hyperglycemia in Type 2 Diabetes: a patient-centered approach position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Dia Care 2012;35:1364-79 15. Raz I, Riddle MC, Rosenstock J, et al. Personalized management of hyperglycemia in type 2 diabetes: reflections from a diabetes care editors’ expert forum. Diabetes Care 2013;36:1779-88

Initiation on basal insulin and glycemic control Curtis & Lage

29

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

Journal of Medical Economics

Volume 17, Number 1

January 2014

16. Institute of Medicine (IOM). Crossing the quality chasm: a new health system for the 21st Century. Washington, DC: The National Academy Press, 2001 17. Smith RJ, Nathan DM, Arslanian SA, et al. Individualizing therapies in type 2 diabetes mellitus based on patient characteristics: what we know and what we need to know. J Clin Endocrinol Metab 2010;95:1566-74 18. Clarke EJ, Barton BA. Entropy and MDL discretization of continuous variables for Bayesian belief networks. Int J Intel Syst 2000;15:61-92 19. Fogelfeld L, Dharmalingam M, Robling K, et al. A randomized, treat-to-target trial comparing insulin lispro protamine suspension and insulin detemir in insulin-naive patients with Type 2 diabetes. Diabet Med 2010;27:181-8 20. Philis-Tsimikas A, Charpentier G, Clauson P, et al. Comparison of once-daily insulin detemir with NPH insulin added to a regimen of oral antidiabetic drugs in poorly controlled type 2 diabetes. Clin Ther 2006;28:1569-81 21. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613-9 22. D’Hoore W, Sicotte C, Tilquin C. Risk adjustment in outcome assessment: the Charlson comorbidity index. Methods Inf Med 1993;32:382-7 23. Bretzel RG, Eckhard M, Landgraf W, et al. Initiating insulin therapy in type 2 diabetic patients failing on oral hypoglycemic agents. Diabetes Care 2009;32(2 Suppl):S260-5 24. Riddle MC, Rosenstock J, Gerich J; Insulin Glargine 4002 Study Investigators. The treat-to-target trial: randomized addition of glargine or human NPH insulin to oral therapy of type 2 diabeteic patients. Diabetes Care 2003;26:3080-6 25. Janka HU, Plewe G, Riddle MC, et al. Comparison of basal insulin added to oral agents versus twice-daily premixed insulin as initial therapy for type 2 diabetes. Diabetes Care 2005;28:254-9 26. Healthcare.gov. Key Features of the Affordable Care Act by Year. US Department of Health & Human Services. http://www.hhs.gov/healthcare/ facts/timeline/timeline-text.html 27. Executive summary: Standards of medical care in diabetes-2013. Diabetes Care 2013;36(1 Suppl):S4–S10 28. Munshi MN, Pandya N, Umpierrez GE, et al. Contributions of basal and prandial hyperglycemia to total hyperglycemia in older and younger adults with type 2 diabetes mellitus. J Am Geriatr Soc 2013;61:535-41 29. Egan BM, Shaftman SR, Wagner CS, et al. Demographic differences in the treatment and control of glucose in type 2 diabetic patients: implications for health care practice. Ethn Dis 2012;22:29-37 30. Hessler DM, Fisher L, Mullan JT, et al. Patient age: a neglected factor when considering disease management in adults with type 2 diabetes. Patient Educ Couns 2011;85:154-9 31. Tirosh A, Stern Z, Mazar M, et al. The influence of age on the management of patients with diabetes in the Israeli population. Popul Health Manag 2013;16:276-82 32. Ahluwalia IB, Mack KA, Murphy W, et al. State-specific prevalence of selected chronic disease-related characteristics–Behavioral Risk Factor Surveillance System, 2001. MMWR Surveill Summ 2003;52:1-80 33. Best JD, Drury PL, Davis TME, et al; Fenofibrate Intervention and Event Lowering in Diabetes Study Investigators. Glycemic control over 5 years in 4900 people with type 2 diabetes: real-world diabetes therapy in a clinical trial cohort. Diabetes Care 2012;35:1165-70 34. Ott P, Benke I, Stelzer J, et al. [‘‘Diabetes in Germany’’(DIG) study. A prospective 4-year-follow-up study on the quality of treatment for type 2 diabetes in daily practice]. Dtsch Med Wochenschr 2009;134:291-7 35. O’Connor PJ, Asche SE, Craine AL, et al. Is patient readiness to change a predictor of improved glycemic control? Diabetes Care 2004;27: 2325-9 36. Ali MK, McKeever Bullard K, Imperature Barker L, et al. Characteristics associated with poor glycemic control among adults with self-reported diagnosed diabetes – National Health and Nutrition Examination Survey, United States, 2007-2010. MMWR Morb Mortal Wkly Rep 2012;61: Suppl:32-37 37. American Diabetes Association. Living with diabetes: tight diabetes control. Alexandria, VA: American Diabetes Association (ADA). July 16, 2003. http:// www.diabetes.org/living-with-diabetes/treatment-and-care/blood-glucosecontrol/tight-diabetes-control.html. Accessed 13 Oct 2013

30

Initiation on basal insulin and glycemic control Curtis & Lage

38. The ORIGIN Trial Investigators. Characteristics associated with maintenance of mean A1C56.5% in people with dysglycemia in the ORIGIN Trial. Diabetes Care 2013;10:2915-22 39. Centers for Disease Control and Prevention. Diabetes Report Card 2012. Atlanta, GA: Centers for Disease Control and Prevention, 2012. p 1-14 40. Barker LE, Kirtland KA, Gregg EW, et al. Geographic Distribution of Diagnosed Diabetes in the U.S. Am J Prev Med 2011;40:434-9 41. Mokdad AH, Bowman BA, Ford ES, et al. The continuing epidemics of obesity and diabetes in the United States. JAMA 2001;286:1195-200 42. Mendes E. Wellbeing: Hawaii Tops Utah for Nation’s Best. 2010. Washington DC: Gallup. www.gallup.com/poll/125849/hawaii-tops-utah-nation-best.aspx. Accessed 13 Oct 2013 43. Page S. Western cities fare best in well-being index. USA Today, 2010 McLean, VA. usatoday30.usatoday.com/news/nation/2010-02-15-cities_N. htm. Accessed February 15, 2010 44. Mayer-Davis EJ, Costacou T. Obesity and sedentary lifestyle: modifiable risk factors for prevention of type 2 diabetes. Curr Diab Rep 2001;1:170-6 45. Healy GN, Dunstan DW, Salmon J, et al. Breaks in sedentary time beneficial associations with metabolic risk. Diabetes Care 2008;31:661-6 46. Ivy DJL. Role of exercise training in the prevention and treatment of insulin resistance and non-insulin-dependent diabetes mellitus. Sports Med 1997;24:321-36 47. Egede LE, Gebregziabher M, Hunt KJ, et al. Regional, geographic, and racial/ ethnic variation in glycemic control in a national sample of veterans with diabetes. Diabetes Care 2011;34:938-43 48. Cummings DM, Lutes LD, Littlewood K, et al. EMPOWER: a randomized trial using community health workers to deliver a lifestyle intervention program in African American women with Type 2 diabetes: design, rationale, and baseline characteristics. Contemp Clin Trials 2013;36:147-53 49. Wendel CS, Shah JH, Duckworth WC, et al. Racial and ethnic disparities in the control of cardiovascular disease risk factors in Southwest American veterans with type 2 diabetes: the Diabetes Outcomes in Veterans Study. BMC Health Serv Res 2006;6:58 50. Hunt KJ, Gebregziabher M, Lynch CP, et al. Impact of diabetes control on mortality by race in a national cohort of veterans. Ann Epidemiol 2013;23:74-9 51. Nichols GA, Hillier TA, Erbey JR, et al. Congestive heart failure in type 2 diabetes prevalence, incidence, and risk factors. Diabetes Care 2001;24:1614-9 52. Cosmi F, Cosmi D, Savino K, et al. [Insulin therapy may hasten congestive heart failure in cardiac patients: case series and review of the literature]. G Ital Cardiol (Rome) 2008;9:509-12 53. Rajagopalan R, Rosenson RS, Fernandes AW, et al. Association between congestive heart failure and hospitalization in patients with type 2 diabetes mellitus receiving treatment with insulin or pioglitazone: a retrospective data analysis. Clin Ther 2004;26:1400-10 54. Action to Control Cardiovascular Risk in Diabetes Study Group, Gerstein HC, Miller ME, et al. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med 2008;358:2545-59 55. ADVANCE Collaborative Group, Patel A, MacMahon S, et al. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med 2008;358:2560-72 56. Duckworth W, Abraira C, Moritz T, et al. Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med 2009; 360:129-39 57. Siraj E, Rubin D, Miller M. The relationship between insulin exposure and cardiovascular mortality in the ACCORD trial. Presented at the American Diabetes Assocation, 73rd Scientific Sessions, Chicago, IL: 2013, Abstract:386–OR. http://www.abstractsonline.com/Plan/ViewAbstract.aspx? sKey¼cc61c947-010f-409e-98f9-fdfd0339ffa7&cKey¼53683283-141d47f7-921d-3c56faab197f&mKey¼{89918D6D-3018-4EA9-9D4F-711F98 A7AE5D}#. Accessed 13 Oct 2013 58. Wood S. Aggressive glucose control no benefit in CV surgery patients with T2D. theheart.org, 2013. New York. http://www.staging.iad1.theheart.org/ article/1556399.do. Accessed June 26, 2013

www.informahealthcare.com/jme ! 2014 Informa UK Ltd

Journal of Medical Economics

January 2014

63. Seligman HK, Jacobs EA, Lo´pez A, et al. Food insecurity and glycemic control among low-income patients with type 2 diabetes. Diabetes Care 2012;35:233-8 64. Barnett A. Dosing of insulin glargine in the treatment of type 2 diabetes. Clin Ther 2007;29:987-99 65. Garber AJ. Treat-to-target trials: uses, interpretation and review of concepts. Diabetes Obes Metab 2013. doi:10.1111/dom12129 66. National Collaborating Centre for Chronic Conditions. Type 2 Diabetes: National clinical guideline for management in primary and secondary care (update). London (UK): Royal College of Physicians. 2008. http://www. nice.org/uk/nicemedia/pdf/CG66diabetesfullguideline.pdf. Accessed 13 Oct 2013

Journal of Medical Economics Downloaded from informahealthcare.com by York University Libraries on 12/09/14 For personal use only.

59. Nichols GA, Gullion CM, Koro CE, et al. The incidence of congestive heart failure in type 2 diabetes an update. Diab Care 2004;27: 1879-84 60. Fabricatore AN, Wadden TA. Treatment of obesity: an overview. Clin Diabetes 2003;21:67-72 61. Shani M, Taylor TR, Vinker S, et al. Characteristics of diabetics with poor glycemic control who achieve good control. J Am Board Fam Med 2008;21:490-6 62. Berkowitz SA, Baggett TP, Wexler DJ, et al. Food insecurity and metabolic control among U.S. Adults With Diabetes. Diabetes Care 2013;36: 3093-9

Volume 17, Number 1

! 2014 Informa UK Ltd www.informahealthcare.com/jme

Initiation on basal insulin and glycemic control Curtis & Lage

31