Practical Focus on American Diabetes Association/European Association for the Study of Diabetes Consensus Algorithm in Patients with Type 2 Diabetes Mellitus: Timely Insulin Initiation and Titration (Iran-AFECT)

Study design

This was a 24-week nation-wide, prospective, multicenter, observational study of people with T2DM who had been initiated or switched to insulin glargine in order to evaluate its clinical safety and effectiveness in routine clinical practice. The study was carried out in 70 centers in Iran. Participants were recruited between March 2011 and March 2012. The physician took the independent decision to initiate (or switch) insulin glargine as basal insulin therapy as part of the routine clinical care. Changes to OGLDs at the time of starting insulin glargine, or thereafter, were entirely at the discretion of the advising physician. Verbal and written instructions were provided to the patient on how to perform self-monitoring of blood glucose levels, how to self-titrate insulin glargine, and how to fill in their individual diary.

According to the American Diabetes Association/European Association for the Study of Diabetes Consensus Algorithm, and as part of usual clinical care, the patient had an assessment of HbA1c every 3 months until the target of <7% was reached. Thus, each patient included in this study had an HbA1c determination at baseline (at most 2 months before inclusion), 12 weeks after insulin initiation and at the end of the study (week 24). HbA1c determination was performed in the same laboratory for a given patient. Data were collected from the physicians' clinical notes and participants' recall and self-monitoring diary/meter at each visit, as available. This information was transferred to a standard case report forms.

Participants

Seven hundred and twenty-five people with T2DM, >18 years of age, for whom the physician had decided to initiate or to switch to insulin glargine were included in the study. Subjects had HbA1c >7.0% but <10%. Eligible participants had the diagnosis of T2DM for at least 1 year.

Subjects were excluded for any of the following criteria: pregnant or lactating women, patients already treated with insulin other than insulin NPH, patients treated with OGLDs other than Metformin, Sulfonylurea or Pioglitazone, patients not willing, or not able, to perform self-monitoring blood glucose, patients not willing, or not able, to self-titrate insulin glargine under physician's guidance, patients with any serious underlying illness, including hepatic failure (aspartate aminotransferase test >3 upper limit of normal range), renal failure (creatinine >1.5 mg/dL in males and >1.4 mg/dL in females), psychiatric patients, patients with history of drug or alcohol abuse and patients on chronic treatment with systemic corticosteroids, hospitalized patients, patient who were involved in other clinical trials, patients who were not suitable for close follow-up, and known hypersensitivity to insulin glargine. Ethics Committee Approval was obtained from Iran University of Medical Sciences. All participants signed informed consent from. In addition, they were free to withdraw at any time.

Assessments and outcome measures

The primary objective of this study was to measure the percentage of patients achieving HbA1c <7% by the end of the study (week 24). The efficacy was also assessed as follows: the percentage of patients achieving HbA1c <7%, the changes in HbA1c at week 24 compared to baseline HbA1c, the changes of basal insulin doses at week 24 (average dosage per patient and timing), and the changes in body weight and lipid profile from baseline.

Safety assessment were the changes in the number of hypoglycemic episodes (blood glucose levels of less than 60 mg/dL [3.3 mmol/L]) occurred in preceding 4 weeks prior to enrollment, and during the study, the changes in the number of nocturnal hypoglycemic events and the number of adverse drug reactions (ADRs) from baseline to final visit.

Major hypoglycemic events were defined as events with symptoms of hypoglycemia for which subjects required the assistance of another person, and were associated with a blood glucose level below 60 mg/dL (3.3 mmol/L) or prompt recovery after oral carbohydrate consumption, intravenous glucose, or glucagon administration. Minor hypoglycemia was any event, where symptoms consistent with hypoglycemia were experienced and either the subjects responded to ingestion of carbohydrate/snack or the episode was associated with a glycemia below 60 mg/dL (3.3 mmol/L).

Statistical methods

IBM SPSS version 19 (IBM Co., Armonk, NY, USA) was used for the statistical analysis. The data were analyzed anonymously. Based on the assumption that 30% of subjects would have an HbA1c <7% at week 24, it was planned to include a total sample of 659 evaluable patients to measure this rate with a 95% confidence interval of ±3.5%. Assuming a 10% drop-out rate, the total number of subjects to be recruited in the study was at least 725.

Analyses were performed on all patients who finished the study according to prior insulin use. Continuous variables were summarized using number of patients, mean±standard deviation, and discrete variables were summarized using number of patients and percentage. Hypoglycemia were summarized as number of events per patient-year, All statistical tests were two-sided, and a P<0.05 was considered statistically significant. For hypoglycemia change from baseline, the percentage of people reporting at least one event was analyzed using McNemar's test. Change from baseline HbA1c, fasting plasma glucose (FPG), and blood lipids was analyzed using paired sample t-test. The percentage of patients having HbA1c <7.0% at 24 weeks was reported using descriptive statistics.

Study participants

Seven hundred and twenty-three subjects were included in this study. Four hundred and fifty-three (63%) were female. The mean age was 54.2±11.2 years, and the mean duration of diabetes was 9.5±6.3 years. Table 1 shows baseline characteristics of insulin naïve, prior insulin users, and entire participants. Use of OGLDs prior to initiating insulin glargine is also demonstrated in Table 1. Regarding medication, 89.8% of patients were being treated with OGLDs alone, 6.9% were receiving OGLD+insulin therapy, and 3.3% were on insulin only.

Table 1

Baseline characteristics for the entire cohort and by pre-study therapy

With regards to prior insulin treatment, insulin-treated people were the same age as insulin naïve patients and they had similar BMI (27.99 kg/m² vs. 27.84 kg/m²). Six hundred and thirty-nine patients completed the study. Baseline HbA1c was lower (8.7% vs. 8.9%) in prior insulin users, while diabetes duration was longer (11.2 years vs. 9.4 years).

Insulin dose and metabolic outcomes

Blood glucose improved significantly by week 24. The HbA1c dropped from 8.88%±0.93% at baseline to 7.62%±1.17% in the whole participants. This improvement was also observed in both the insulin naïve and prior insulin users (Table 2). By the end of the study 27% of the entire participants reached to HbA1c target of less than 7% and 52% had HbA1c ≤7.5%.

Table 2

Glucose control and body weight for the entire participants and by pre-study therapy at baseline and after 24 weeks of insulin glargine therapy

Total daily insulin dose had been titrated up to 20.32±10.31 U/day at week 24 in people who were insulin naïve. In prior insulin users, the starting insulin dose was 22.98±12.02 U/day titrating up to 30.24±13.65 U/day at week 24.

After initiating insulin in insulin naïve people, metformin was continued in around 86% of people. While 41.4% of people starting insulin glargine continued using sulfonylureas and 64% discontinued thiazolidinedione.

In prior insulin users, metformin was discontinued in about 6% of people, 68.7% continued sulfonylurea while thiazolidinedione was continued in all prior insulin users. The changes in mean body weight were neither statistically nor clinically significant over 24 weeks for the total participants, the insulin-naïve population, and prior insulin users (Table 2).

Total cholesterol level reduced in the whole participants from a mean of 189.96 to 174.98 mg/dL after 24 weeks (P<0.001) (Table 2). A significant reduction was seen in low density lipoprotein cholesterol level in the entire cohort and insulin-naïve people.

Triglyceride levels also dropped from a mean of 187.94 to 160.21 mg/dL in total participants (P<0.001). Results were similar for the insulin-naïve people and prior insulin users. Furthermore, there was a small increase in high density lipo-protein cholesterol levels during the study in the entire population and insulin-naïve people.

Hypoglycemia

The reported rate of overall hypoglycemic episodes did not differ significantly between baseline and week 24 in the entire cohort and prior insulin users while it was statistically different in the insulin naïve people (Table 2). In the whole participants reported rates of overall hypoglycemia increased from 0.08 to 0.12 events/person-year, with no statistical difference in the proportion of people having an event. In the prior insulin users, the reported rate of hypoglycemia decreased from 0.16 to 0.13 events/person-year by the end of the study, but it was not statistically significant (P=0.99). In Insulin naïve patients, this number increased from 0.06 to 0.09 events/person-year (P=0.03). The rates of minor, nocturnal, and major hypoglycemic events remained clinically and statistically unchanged in the entire participants (Table 2).

Adverse drug reactions and adverse events

During the study, 99 ADRs were reported. Of these, three people suffered a serious ADR due to major hypoglycemic event; the rest was due to minor and nocturnal hypoglycemic events.