The Association of Self-Monitoring of Blood Glucose Use with Medication Adherence and Glycemic Control in Patients with Type 2 Diabetes Initiating Non-Insulin Treatment

Abstract

Background:

The value of self-monitoring of blood glucose (SMBG) for persons with type 2 diabetes who do not use insulin remains controversial. This observational study compares the likelihood of medication adherence and change in glycated hemoglobin (A1C) for non–insulin-using patients using SMBG versus those not using SMBG. The study also assessed the association between diabetes medication adherence and SMBG use.

Patients and Methods:

Data were extracted on 5,172 patients who began non-insulin diabetes medication between October 1, 2006, and March 31, 2009. The study assessed change in A1C associated with SMBG use and testing frequency at different categorical levels of baseline A1C. The likelihood of medication adherence for SMBG users was compared with that for non-SMBG users at different categorical levels of baseline A1C. The study further explored the interactions between SMBG use and medication adherence on change in A1C.

Results:

SMBG users had greater reductions in A1C compared with nonusers when the baseline A1C was ≥7%. Increasing SMBG frequency was associated with greater reductions in A1C. The study also examined the associations among SMBG use, medication adherence, and change in A1C. SMBG users had greater decreases in A1C for both medication-adherent and -nonadherent patients. As expected, medication adherence was associated with greater reductions in A1C for both SMBG nonusers and users. It is interesting that medication-nonadherent SMBG users had similar reductions in A1C compared with medication-adherent non-SMBG users.

Conclusions:

Both SMBG use and medication adherence were associated with similar degrees of A1C reduction after controlling for baseline A1C, suggesting that both factors may be important for attaining glycemic control.

Introduction

G lycemic control is important for reducing micro- and macrovascular complications in persons with diabetes, including those who do not take insulin.¹ The literature on the value of self-monitoring of blood glucose (SMBG) as a tool to help non–insulin-using patients with type 2 diabetes achieve optimal glycemic control has been mixed.

Meta-analyses of randomized controlled trials have reported that SMBG, compared with non-SMBG, was associated with a modest, but statistically significant, improvement in glycemic control (glycated hemoglobin [A1C] reduction) in patients with non–insulin-treated type 2 diabetes.^2,3 In general, individual studies that considered SMBG as a direct intervention did not demonstrate benefit.^4
–6 In contrast, studies that used SMBG to facilitate decisions regarding diabetes management showed a significant effect on A1C. Three recent randomized controlled trials suggest that a program of structured SMBG combined with education and medication titration can improve glycemic control over usual care.^7
–9

In addition to improved glycemic control, these studies demonstrated that patients enrolled in a structured SMBG program are more likely to experience weight loss and a decrease in waist circumference.^8,9 One explanation offered for this is that SMBG users in these studies engaged in better lifestyle behavior—such as eating healthier foods and exercising more.^8,9 Furthermore, their healthcare providers were more likely to recommend medication changes to achieve even better control.¹⁰ Karter et al.¹¹ observed that SMBG users in general had better medication adherence than nonusers; greater adherence to a diabetes regimen is associated with better glycemic control.¹²

One gap in prior studies is that they did not address potential mediators, which could explain the association between SMBG use and change in A1C. An important mediator of change in A1C is medication adherence. In order to obtain real-world information, this study was performed using a large U.S. administrative claims database.

The primary objective of this retrospective study was to evaluate the association between SMBG use and change in A1C. This study examines the association between SMBG use and A1C reduction among patients initiating non-insulin diabetes therapy. It also focuses on the effect of one key factor on improving A1C—diabetes medication adherence. The study also assessed the interactions among SMBG use, medication adherence, and baseline A1C level (<7%, 7–8%, 8–9%, 9–10%, and ≥10%) using a factorial analysis.

Research Design and Methods

Study design

This exploratory observational analysis used data from a large U.S. insurance claims database (i3 InVision^™ Data Mart; Ingenix, Inc., Eden Prairie, MN). The database is a de-identified, HIPAA (Health Insurance Portability and Accountability Act of 1996)-compliant database, so no institutional review board approval was necessary. This database primarily represents a commercially insured population and is not representative of Medicare patients (e.g., patients over the age of 65 years). The database includes enrollment dates, patient demographics (age, gender, geographic location), medical claims (place of service, diagnosis, procedures), and pharmacy claims (quantity, strength, days' supply of drug). A unique advantage of this database is that it includes laboratory values coded by Logical Observations and Identifiers Names and Codes (LOINC^®) (Regenstrief Institute, Indianapolis, IN) for a subset of patients, who were used to track A1C changes over time.

Sample identification

Patients with at least one prescription claim for a non-insulin diabetes medication between October 1, 2006, and March 31, 2009, and no evidence of a diabetes-related prescription claim during the prior 365 days were included. Non-insulin diabetes medications were identified using American Hospital Formulary Service (AHFS) codes contained in the database. Index medications were categorized and included metformin, sulfonylureas, thiazolidinediones, dipeptidyl peptidase inhibitors, metformin combination pills, and all other antidiabetes agents (e.g., amylinomimetics, incretin mimetics, and meglitinides). The date of the first prescription for a non-insulin diabetes medication was the patient's index date. Patients younger than 18 years or older than 63 years at the index date were excluded to prevent inclusion of pediatric patients and individuals receiving Medicare.

Patients were required to have at least one valid A1C (laboratory value) measurement (range, 4–20%) recorded both within 90 days before the index date (baseline A1C) and during 91–365 days after the index date (post-index A1C). Patients were excluded if they had records indicating insulin use in the 365 days before or after the index date, had incomplete data on key study variables, or did not remain continuously eligible for services under the benefits plan 365 days pre- and post-index.

Outcomes and analysis variables

The primary outcome was change in A1C. It was calculated as the difference between the post-index A1C and baseline A1C (i.e., post-index A1C minus baseline A1C). The baseline value was A1C prior to the index closest to medication initiation. Post-index A1C was the average of the A1C values during the 91–365 days post-index. The distribution of the number of days for the post-index A1C values was continuous with an average of 223 days. Most patients (56.4% [2,920 of 5,172]) had only one A1C record post-index, and 89.4% (4,623) of the patients had no more than two A1C records post-index.

SMBG use was defined as the presence or absence of at least one claim for any SMBG device or strip during the 365 days post-index. Claims for a SMBG device or strip were identified using a product class code contained in the administrative database (AHFS code M4A). Strip use was further categorized, based on the number of strips available, as a surrogate for SMBG frequency. The total number of strips dispensed during the 365 days post-index was divided by 365, resulting in the following categories: <0.5 test strips/day, 0.5 to <1 test strip/day, and ≥1 test strip/day. The cutoffs for this population were similar to those used by Karter et al.¹¹ and were based on the distribution observed.

For simplicity, baseline A1C was grouped into five categories: <7%, 7% to <8%, 8% to <9%, 9% to <10%, and ≥10%. The average baseline A1C values in each category for SMBG users and SMBG nonusers were similar.

To measure medication adherence, the study used a commonly accepted measure, the medication possession ratio (MPR).¹³ This ratio was calculated for each patient as the sum of days where any non-insulin diabetes medication was available during 365 days post-index, based on the dispensing date and the days supply recorded by the pharmacist(s). The MPR calculation included all diabetes medications the patient was prescribed. Days hospitalized were counted as days with all medications available. Patients with an MPR ≥80% were considered diabetes medication adherent¹⁴ (herein referred to as medication adherent). Those with an MPR < 80% were considered diabetes medication nonadherent (herein referred to as medication nonadherent). An 80% cutoff for MPR is an arbitrary number that is commonly used in adherence calculations.^13,14

In addition to the primary analysis variables, several demographic and treatment characteristics were used to describe the sample. These variables included: age, gender, healthcare professional office visits, hospitalizations, number of chronic medication fills, co-morbidities, presence of a depression diagnosis, and an indicator for the index diabetes medication therapy (metformin, metformin combination, sulfonylurea, dipeptidyl peptidase-4 inhibitor, thiazolidione, other medications). Healthcare professional office visits were divided into pre- and post-index visits and were also grouped by “all visits” and “diabetes-related visits” (identified by the ICD-9 diagnosis code 250.xx). Pre-index chronic medications were measured as the average number of drugs from the following therapy classes (according to AHFS therapeutic categories): antiarrhythmia drugs (240404), antihyperlipidemics (2406xx), antihypertensives (2420xx, 2424xx, 2428xx, and 2432xx), anticonvulsants (281200), and diuretics (4028xx). The presence of co-morbidities was obtained by diagnosis codes and inferred based on prescription drug claims for certain medication classes. The list of co-morbidities includes those used in the Charleson–Quan co-morbidity index.¹⁵ The claims database does not include information regarding blood pressure, weight, height, diet, and exercise.

Statistical analysis

The relationship between SMBG and change in A1C, frequency of strip use and change in A1C, and medication adherence and change in A1C were assessed in three independent analysis of variance (ANOVA) models. In each of the three models, change in A1C (dependent measure) was a function of baseline A1C (one of five categories), either SMBG, strip-count category, or medication adherence (independent variables of interest), and the interactions between the two independent variables (for each model). The least significant difference test was used to assess any differences in mean change in A1C by group in each ANOVA.

The relationship between SMBG use and medication adherence, both categorical variables, was assessed using a Cochran–Mantel–Haenszel test, controlling for baseline A1C category.

The final model evaluated the relationship among SMBG, diabetes medication adherence, and baseline A1C category in a three-way ANOVA model. The model used change in A1C as the dependent measure of interest, with SMBG use, medication adherence, baseline A1C category, and their interactions as independent variables. Follow-up comparisons of means used least significant difference tests.

SAS^® version 9.2 (SAS Institute Inc., Cary, NC) was used for all analyses. All tests were two-tailed and were conducted at the 5% significance level. Given the exploratory nature of the analysis, no adjustment was performed for multiplicity.

Results

In total, 5,172 patients met the analysis criteria, of whom 2,744 (53.0%) used SMBG and 2,428 (46.9%) did not. SMBG users were more likely than SMBG nonusers to be female (49.7% vs. 46.7%, P=0.0331) and less likely to start with metformin therapy as the index medication (66.7% vs. 73.5%, P<0.0001). The groups of SMBG users and nonusers were similar in terms of age and healthcare service utilization overall (see supplementary Table 1 at www.liebertonline.com/dia).

Both groups had similar rates of co-morbidities (listed in Supplementary Table 1) except the frequency of congestive heart failure (SMBG users vs. nonusers, 5.3% vs. 6.8%, respectively; P=0.0191). Other differences between SMBG users and nonusers were noted in the number of times chronic medications were filled pre-index (7.4 vs. 8.2, respectively; P=0.0026) and pre-index diabetes-related office visits (6.4 vs. 8.7, respectively; P<0.0001).

Average baseline A1C was significantly higher among SMBG users compared with SMBG nonusers (8.1% and 7.3%, respectively; P<0.0001).

SMBG use and change in A1C

Overall, SMBG users had a greater decline in A1C compared with nonusers (−1.4% vs. −0.6%, respectively; P<0.0001). The difference in A1C change between SMBG users and nonusers remained, even after adjusting for baseline A1C.

Change in A1C was compared for SMBG users and nonusers at different baseline A1C ranges (Fig. 1). SMBG users had a greater decline in A1C across the A1C ranges, except at baseline A1C at the American Diabetes Association target (<7%). For example, at a baseline of 8% to <9%, the decline in A1C was −1.5% and −0.94% for SMBG users and nonusers, respectively (difference −0.56%, P<0.0001). When baseline A1C was 9% to <10%, SMBG users had a change in A1C of −2.47% versus −1.88% for nonusers (difference −0.59%, P<0.0001).

FIG. 1.

Mean change (±SE) in hemoglobin A1c (A1C) levels (post–pre) in self-monitoring of blood glucose (SMBG) users (n=2,744) and nonusers (n=2,428).

In general, higher SMBG frequency (measured as number of test strips available in the post-index period) correlated with a greater decrease in A1C except as noted in Figure 2. Those who tested less than 0.5 times per day on average had a reduction in A1C of −1.34% and −1.98% from a baseline A1C of 8% to <9% and 9% to <10%, respectively. These reductions increased to −1.5 and −2.59, respectively, for patients who tested at least 0.5 times per day but less than once per day. Those who tested at least once per day had the greatest reductions in A1C: −1.68% and −2.91%, respectively.

FIG. 2.

Mean change (±SE) in hemoglobin A1c (A1C) by self-monitoring of blood glucose (SMBG) category and baseline A1C level. Horizontal lines indicate statistically significant differences between two SMBG categories (P<0.05, analysis of variance (ANOVA)).

SMBG use and diabetes medication adherence

Overall, 44.4% of the patients were medication adherent. SMBG users were more likely to be medication adherent than SMBG nonusers: 49.9% versus 38.2%, respectively (difference 11.6%, P<0.0001). Table 1 summarizes the differences in medication adherence between the two groups. At each level of A1C, SMBG users were more likely to be medication adherent (all differences were statistically significant, P<0.0001). Greater test strip availability (SMBG frequency) was also associated with a greater likelihood of being medication adherent: 41.5% for patients having <0.5 strips/day versus 46.5% for patients having between 0.5 and 1 strip/day (P=0.0281) and 64.1% for patients having ≥1 strip/day (P<0.0001 vs. <0.5 strips/day and vs. 0.5 to ≤1 strip/day). Within each medication group, SMBG users were more likely to be medication adherent than nonusers. A separate analysis of medication-nonadherent patients (MPR <80%) showed that SMBG users had a relatively higher level of medication adherence (i.e., more likely to have an MPR >50%) than nonusers (data not shown).

Table 1.

Numbers and Percentages of Type 2 Diabetes Patients Who Are Medication Adherent by Baseline Hemoglobin A1c Category

	SMBG users (n=2,744)			SMBG nonusers (n=2,428)
Baseline A1C	Group n	Adherent n	%	Group n	Adherent n	%	Difference (%)	P value ^a
<7%	998	463	46.4	1,361	469	34.5	11.9	<0.0001
7% to <8%	676	349	51.6	536	224	41.8	9.8	<0.0001
8% to <9%	325	163	50.2	195	94	48.2	2.0	<0.0001
9% to <10%	236	115	48.7	116	47	40.5	8.2	<0.0001
10%	509	278	54.6	220	94	42.7	11.9	<0.0001
Total	2,744	1,368	49.9	2,428	928	38.2	11.6	<0.0001

Test within categories for association using the Cochran–Mantel–Haenszel test, controlling for pre-period A1C. Overall test for association using χ².

A1C, hemoglobin A1c; SMBG, self-monitoring of blood glucose.

Medication adherence and change in A1C

As expected, medication-adherent patients had a greater decline in A1C compared with nonadherent patients (see supplementary Fig. 1). The difference in change in A1C between medication-adherent and -nonadherent patients increased as baseline A1C increased. For example, when the baseline A1C was 7% to <8%, the difference in A1C change was −0.21% (change in A1C of −0.68% and −0.47% for medication-adherent and -nonadherent patients, respectively; P=0.0006). At a baseline A1C of 9% to <10%, the difference in change in A1C was −0.61% (change in A1C of −2.60% and −1.99% for medication-adherent and -nonadherent patients, respectively; P<0.0001).

SMBG use, medication adherence, and change in A1C

Figure 3 summarizes change in A1C by SMBG use and medication adherence. Medication-nonadherent patients who did not use SMBG (M⁻S⁻) had the smallest decrease in A1C. Medication-adherent SMBG users (M⁺S⁺) had the greatest declines. The differences among the four groups increased as baseline A1C increased. SMBG users had greater declines in A1C for both medication-nonadherent and -adherent groups (M⁻S⁺ and M⁺S⁺, respectively). Change in A1C was similar for M⁻S⁺ (SMBG use only) and M⁺S⁻ (medication adherence only) when the baseline A1C was <10%.

FIG. 3.

Mean change (±SE) in hemoglobin A1c (A1C) (post–pre) categorized by self-monitoring of blood glucose (SMBG) use and medication adherence. Horizontal lines indicate a significant difference between two groups (P<0.05, pairwise comparisons). Significant effects for SMBG use and medication adherence within an A1C bucket were present for A1C 7% to <8% (P=0.001) and A1C 8% to <9%, 9% to <10%, and ≥10% (P<0.001). M⁻S⁻, medication-nonadherent SMBG nonusers (n=1,500); M⁻S⁺, medication-nonadherent SMBG users (n=1,376); M⁺S⁻, medication-adherent SMBG nonusers (n=928); M⁺S⁺, medication-adherent SMBG users (n=1,368).

At a baseline A1C of 7% to <8%, only the differences between the results for the M⁺S⁺ group and those of each the other three groups were statistically significant (P<0.05). A1C reduction ranged from −0.43% to −0.77%.

As baseline A1C rose to 8% to <10%, the differences between M⁻S⁻ versus M⁻S⁺ and M⁻S⁻ versus M⁺S⁻ became statistically significant (P<0.05), and for M⁺S⁺ changes remained significant. Additionally, the difference in change in A1C among the four groups also increased. The ranges for change in A1C from baseline A1C levels of 8% to <9% and 9 to <10% were −0.66% to −1.63% and −1.65% to −2.76%, respectively.

When the baseline A1C was above 10%, the differences for each comparison between each of the four groups were statistically significant (P<0.05). The range of change in A1C was −2.87% to −4.92%.

Discussion

This retrospective analysis of data from a large claims database found that SMBG use was associated with greater reductions in A1C in patients newly treated with non-insulin diabetes medication who had a baseline A1C ≥7% during the first year after initiating treatment. Greater SMBG frequency was associated with larger reductions in A1C. These findings are consistent with other recently published reports that use of SMBG is associated with greater A1C reductions.^11,16
–18

This study extends previous findings by revealing important relationships among SMBG, change in A1C, and medication adherence. As expected, medication-adherent patients had a greater reduction in A1C than nonadherent patients, and SMBG users were more likely to be medication adherent. Both medication-adherent and -nonadherent patients benefitted from SMBG use.

Greater medication adherence among SMBG users was also seen in a prior retrospective study on patients initiating sitagliptin therapy.¹⁹ In this study, those who had greater availability of test strips were also more likely to be medication adherent than those who had less availability of strips. As expected, medication adherence was associated with a greater reduction in A1C. Other studies have shown a link between medication adherence and improved clinical outcomes.^12,20,21

It is interesting that SMBG use among the medication-nonadherent group was associated with a similar level of reduction in A1C as in the medication-adherent group who did not use SMBG. Additionally, SMBG users continued to have larger A1C reductions in both medication-adherent and -nonadherent groups compared with SMBG nonusers. These findings suggest that SMBG use may be associated with additional factors that improve glycemic control beyond medication adherence.

One explanation for this added benefit may be that patients utilize SMBG data to make lifestyle changes, such as improved diet and/or exercising, more often. Weight loss through diet and exercise is known to improve glucose control. Durán et al.⁸ and Franciosi et al.⁹ demonstrated that SMBG combined with education can lead to reductions in weight and waist circumference in persons newly diagnosed with type 2 diabetes. Unfortunately, information regarding weight and waist circumference is not available in this claims database.

Another factor that may help explain the association between SMBG use and improved changes in A1C is the use of diabetes education and nutrition counseling. A preliminary query for patients with coded diabetes education and nutritionist counseling identified only 258 patients. This small number may be due to the lack of availability of nutrition counseling or the lack of coverage for the counseling. No further analysis was performed because the data were too few to be interpreted with any confidence.

SMBG users may have also had more frequent diabetes medication adjustments. SMBG data allow the healthcare professional to monitor the effect that medications have on blood glucose more precisely. Using this information the physician can make appropriate therapy changes more quickly in SMBG users than patients who only have A1C data. Both Franciosi et al.⁹ and Polonsky et al.¹⁰ found that patients who performed structured testing were more likely to receive medication adjustment recommendations compared with patients in the usual care arm. The association between SMBG use and medication adjustments is an interesting topic, which may be explored further in future studies.

Limitations

Limitations of this study include the retrospective, observational design versus a prospective, randomized controlled study. However, this study is exploratory in nature and can help develop hypotheses to test in prospective studies.

Claims data offer no mechanism to verify whether a medication is taken or a SMBG test strip is used. The current study used prescription refills and claims for test strips as surrogates for medication adherence and SMBG use, including average testing frequency.

Claims data also do not provide information regarding diet, exercise, education level, and income. Additionally, the final study did not include cofounders such as age, gender, co-morbidities, healthcare utilization, and number of diabetes medications or chronic medications used pre-index given the exploratory nature of the study. These factors both available and unavailable in claims data may have influenced both medication adherence and glycemic control in the final model presented in this article.

Point-of-care A1C data are not available in claims data. Even laboratory A1C data are only available within a subset of patients for whom the insurer captures electronic laboratory values. Therefore, some A1C values may be missing on patients included in this study.

Results of this study may not be generalizable to the overall population. The study excluded patients without available A1C data. However, these patients had similar characteristics (age, gender, co-morbidities) to the final study population.

There may be a selection bias, especially associated with those patients who use SMBG. These patients, by their very nature, may be more engaged in ensuring they follow clinical advice. Patients more engaged in their care would be more likely to have reductions in A1C compared with those who are more passive (or resistant) to care. On average, A1C decreased in all groups of patients (SMBG users, SMBG nonusers, medication adherent, and medication nonadherent) in each baseline A1C category. This may not be observed in the general population of patients. Therefore, findings from this study should be confirmed in future studies.

In conclusion, patients using SMBG with a baseline A1C of at least 7% had a statistically significant greater decline in A1C compared with SMBG nonusers. The magnitude of difference in A1C change was also clinically significant (difference in change in A1C at least 0.5%) when the baseline A1C was at least 8%. The improvement in A1C appears to occur because SMBG users were also more likely to be adherent to their diabetes medications. Most of the glycemic benefits associated with SMBG use are likely related to greater medication adherence. This study also suggested that SMBG use was likely associated with other factors that contributed to reductions in A1C beyond medication adherence alone. Exploring these other factors could be a topic for future studies. This study adds to the growing evidence that SMBG may yield significant benefits for patients with type 2 diabetes who begin their first non-insulin diabetes medication.

Footnotes

Acknowledgments

Funding of this study was provided by LifeScan, Inc., Milpitas, CA, and Johnson & Johnson, New Brunswick, NJ. The authors received statistical support from Dr. Gang Li, an employee of LifeScan, Inc., and Dr. Gordon Pledger, a consultant. The authors also received editorial and writing support from Excerpta Medica.

Author Disclosure Statement

N.V. and P.R. are employees of LifeScan, Inc. S.N. and M.D. are employees of Johnson & Johnson. C.K. has received consulting fees from LifeScan, Inc. and Johnson & Johnson. All authors were involved in study design, data analysis, and manuscript development. Primary statistical analysis was performed by C.K.

References

American Diabetes Association:. Standards of medical care in diabetes—2011. Diabetes Care, 2011; 34,Suppl 1:S11–S61.

Allemann

, Houriet

, Diem

, Stettler

. Self-monitoring of blood glucose in non-insulin treated patients with type 2 diabetes: a systematic review and meta-analysis. Curr Med Res Opin, 2009; 25:2903–2913.

St John

, Davis

, Price

, Davis

. The value of self-monitoring of blood glucose: a review of recent evidence. J Diabetes Complications, 2010; 24:129–141.

O'Kane

, Bunting

, Copeland

, Coates

. Efficacy of self monitoring of blood glucose in patients with newly diagnosed type 2 diabetes (ESMON study): randomised controlled trial. BMJ, 2008; 336:1174–1177.

Kleefstra

, Hortensius

, Logtenberg

, Slingerland

, Groenier

, Houweling

, Gans

, van Ballegooie

, Bilo

. Self-monitoring of blood glucose in tablet-treated type 2 diabetic patients (ZODIAC) Neth J Med, 2010; 68:311–316.

Farmer

, Wade

, French

, Simon

, Yudkin

, Gray

, Craven

, Goyder

, Holman

, Mant

, Kinmonth

, Neil

. DiGEM Trial Group. Blood glucose self-monitoring in type 2 diabetes: a randomised controlled trial. Health Technol Assess, 2009; 13:iii–ivix–xi, 1–50.

Polonsky

, Fisher

, Schikman

, Hinnen

, Parkin

, Jelsovsky

, Petersen

, Schweitzer

, Wagner

. Structured self-monitoring of blood glucose significantly reduces A1C levels in poorly controlled, noninsulin-treated type 2 diabetes. Results from the Structured Testing Program study. Diabetes Care, 2011; 34:262–267.

Durán

, Martín

, Runkle

, Pérez

, Abad

, Fernández

, Del Valle

, Sanz

, Calle-Pascual

. Benefits of self-monitoring blood glucose in the management of new-onset Type 2 diabetes mellitus: the St. Carlos Study, a prospective randomized clinic-based interventional study with parallel groups. J Diabetes, 2010; 2:203–211.

Franciosi

, Lucisano

, Pellegrini

, Cantarello

, Consoli

, Cucco

, Ghidelli

, Sartore

, Sciangula

, Nicolucci

. ROSES Study Group: ROSES: role of self-monitoring of blood glucose, intensive education in patients with Type 2 diabetes not receiving insulin. A pilot randomized clinical trial. Diabet Med, 2011; 28:789–796.

10.

Polonsky

, Fisher

, Schikman

, Hinnen

, Parkin

, Jelsovsky

, Axel-Schweitzer

, Petersen

, Wagner

. A structured self-monitoring of blood glucose approach in type 2 diabetes encourages more frequent, intensive, and effective physician interventions: results from the STeP study. Diabetes Technol Ther, 2011; 13:797–802.

11.

Karter

, Ackerson

, Darbinian

, D'Agostino

Jr , Ferrara

, Liu

, Selby

. Self-monitoring of blood glucose levels, glycemic control: the Northern California Kaiser Permanente diabetes registry. Am J Med, 2001; 111:1–9.

12.

Krapek

, King

, Warren

, George

, Caputo

, Mihelich

, Holst

, Nichol

, Shi

, Livengood

, Walden

, Lubowski

. Medication adherence and associated hemoglobin A1C in type 2 diabetes. Ann Pharmacother, 2004; 38:1357–1362.

13.

Cramer

, Roy

, Burrell

, Fairchild

, Fuldeore

, Ollendorf

, Wong

. Medication compliance and persistence: terminology and definitions. Value Health, 2008; 11:44–47.

14.

Karve

, Cleves

, Helm

, Hudson

, West

, Martin

. Good and poor adherence: optimal cut-point for adherence measures using administrative claims data. Curr Med Res Opin, 2009; 25:2303–2310.

15.

Quan

, Sundararajan

, Halfon

, Fong

, Burnand

, Luthi

, Saunders

, Beck

, Feasby

, Ghali

. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care, 2005; 43:1130–1139.

16.

Poolsup

, Suksomboon

, Rattanasookchit

. Meta-analysis of the benefits of self-monitoring of blood glucose on glycemic control in type 2 diabetes patients: an update. Diabetes Technol Ther, 2009; 11:775–784.

17.

Polonsky

, Fisher

, Schikman

, Hinnen

, Parkin

, Jelsovsky

, Amstutz

, Schweitzer

, Wagner

. The value of episodic, intensive blood glucose monitoring in non-insulin treated persons with type 2 diabetes: design of the Structured Testing Program (STeP) study, a cluster-randomised, clinical trial [NCT00674986] BMC Fam Pract, 2010; 11:37.

18.

Karter

, Parker

, Moffet

, Spence

, Chan

, Ettner

, Selby

. Longitudinal study of new and prevalent use of self-monitoring of blood glucose. Diabetes Care, 2006; 29:1757–1763.

19.

Nigam

, Virdi

, Daskiran

, Kozma

, Paris

, Dickson

. Association between sitagliptin adherence and self-monitoring of blood glucose [abstract] J Diabetes Sci Technol, 2011; 4:A119.

20.

Schectman

, Nadkarni

, Voss

. The association between diabetes metabolic control and drug adherence in an indigent population. Diabetes Care, 2002; 25:1015–1021.

21.

White

, Kellar

, Prevost

, Kinmonth

, Sutton

, Canny

, Griffin

. Adherence to hypoglycaemic medication among people with type 2 diabetes in primary care. Prim Care Diabetes, 2012; 6:27–33.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB

0.02 MB