In the Journals

Intensive intervention fails to improve insulin persistence in type 2 diabetes

Increasing the focus and intensity of an intervention for a targeted group of adults with poorly controlled type 2 diabetes did not increase the likelihood that they would continue to fill prescriptions for insulin, according to study results.

However, a modest improvement in glycemic control was observed in the targeted group compared an unselected group of patients who received a lower-intensity intervention.

“Given that more intensive interventions tend to result in larger improvements in adherence, focusing only on individuals most likely to benefit may allow more resources to be devoted to fewer individuals without increasing overall costs of an intervention program,” Julie C. Lauffenburger, PharmD, PhD, an instructor in medicine at Harvard Medical School and an associate epidemiologist in the division of pharmacoepidemiology and pharmacoeconomics at Brigham and Women’s Hospital, and colleagues wrote. “A more potent effect for a comparatively small subgroup may also make it more likely to observe population-level effects when appropriately analyzing quality improvement trials using intention-to-treat principles, even though patient acceptance of interventions may be less than 30%.”

Lauffenburger and colleagues analyzed data from 6,000 adults with type 2 diabetes prescribed basal insulin therapy insured by Horizon Blue Cross Blue Shield of New Jersey between July 7, 2016, and Oct. 5, 2017 (mean age, 56 years; 59.8% men). Researchers assigned participants to one of three arms, based on HbA1c level, with arms designed to be equivalently priced to mirror the type of choice a health insurer would make to allocate funds for a quality improvement program (costs were determined by Magellan Rx Management). Arm 1 offered a low-intensity intervention to all patients, with all participants receiving a letter informing them about pharmacist outreach, a reminder postcard and a small pillbox, along with two follow-up phone calls. Arm 2 offered a moderate-intensity intervention to 60% of patients (n = 1,200) based on their predicted risk for insulin nonadherence, and included up to six follow-up calls and two calls with a patient’s primary care physician or pharmacy to clarify treatment issues or receive recommendations for therapeutic changes. Patients were also offered enrollment in a weekly text messaging program focused on medication-taking behaviors.

Arm 3 offered a high-intensity intervention to 40% of patients (n = 800) based on glycemic control and predicted risk for insulin nonadherence, including up to 12 follow-up calls and PCPs or pharmacists called as often as necessary. Patients were also offered text messages delivered weekly, every 3 days or daily.

“Arm 3 was designed to mimic the most intensive type of strategy that a telephone-based disease management program could offer,” the researchers wrote.

Primary outcome was continuing to fill prescriptions for insulin. Secondary outcomes were changes in HbA1c and health care utilization. Researchers evaluated outcomes in arms 2 and 3 vs. arm 1 using claims data, intention-to-treat principles and multiple imputation for missing values during 12 months of follow-up.

Among targeted patients, 24.7% in arm 1, 30.7% in arm 2 and 34.3% in arm 3 completed a telephone consultation with a pharmacist.

Patients with an initial consultation received a mean of 1.7 calls in arm 1, 1.8 calls in arm 2 and 2.1 calls in arm 3. Across arms, most patients who received initial consultations self-reported optimal insulin adherence, according to researchers.

In arm 1, the rate of nonpersistence was 5.4%, according to researchers. Compared with that arm, rates of nonpersistence in arms 2 and 3 were 4.7% (RR = 0.88; 95% CI, 0.75-1.03) and 4.9% (RR = 0.91; 95% CI, 0.77-1.06), respectively. Mean time to insulin nonpersistence was 250 days in arm 1, 255 days in arm 2 and 258 days in arm 3.

Researchers did not observe a between-group difference in arms 1 and 2 for HbA1c; however, patients assigned to arm 3 had a lower HbA1c vs. arm 1 (mean absolute difference, –0.25%; 95% CI, –0.43 to –0.06).

There were no between-group differences for total health care spending; however, patients in arm 2 had a higher likelihood of hospitalizations (OR = 1.22; 95% CI, 1.06-1.41) and ED visits (OR = 1.38; 95% CI, 1.24-1.53), according to researchers. There were no between-group differences in hypoglycemia risk.

“Targeting a different patient population might have increased the effectiveness of the interventions that we tested,” the researchers wrote. “For example, we focused on patients at moderate risk of nonpersistence. It is possible that the intervention could have been more effective if targeted to patients at highest risk rather than those who we hypothesized would be more impactable. In arm 2, we also included patients with good baseline glycemic control, hypothesizing that even they could benefit, although this could have diluted the effect.”

The researchers noted that, without a true control group in the study, they could not test for differences between the untargeted, low-intensity intervention vs. no intervention. The findings also may not be fully generalizable to those with Medicare or Medicaid coverage.

“Our results suggest that targeting patient populations for more intensive interventions based both on predicted risk of nonadherence and level of glycemic control has the potential to be more effective than untargeted approaches.” – by Regina Schaffer

Disclosures: Sanofi supported this study. Lauffenburger reports she has received grants from AstraZeneca and Sanofi. Please see the study for all other authors’ relevant financial disclosures.

Increasing the focus and intensity of an intervention for a targeted group of adults with poorly controlled type 2 diabetes did not increase the likelihood that they would continue to fill prescriptions for insulin, according to study results.

However, a modest improvement in glycemic control was observed in the targeted group compared an unselected group of patients who received a lower-intensity intervention.

“Given that more intensive interventions tend to result in larger improvements in adherence, focusing only on individuals most likely to benefit may allow more resources to be devoted to fewer individuals without increasing overall costs of an intervention program,” Julie C. Lauffenburger, PharmD, PhD, an instructor in medicine at Harvard Medical School and an associate epidemiologist in the division of pharmacoepidemiology and pharmacoeconomics at Brigham and Women’s Hospital, and colleagues wrote. “A more potent effect for a comparatively small subgroup may also make it more likely to observe population-level effects when appropriately analyzing quality improvement trials using intention-to-treat principles, even though patient acceptance of interventions may be less than 30%.”

Lauffenburger and colleagues analyzed data from 6,000 adults with type 2 diabetes prescribed basal insulin therapy insured by Horizon Blue Cross Blue Shield of New Jersey between July 7, 2016, and Oct. 5, 2017 (mean age, 56 years; 59.8% men). Researchers assigned participants to one of three arms, based on HbA1c level, with arms designed to be equivalently priced to mirror the type of choice a health insurer would make to allocate funds for a quality improvement program (costs were determined by Magellan Rx Management). Arm 1 offered a low-intensity intervention to all patients, with all participants receiving a letter informing them about pharmacist outreach, a reminder postcard and a small pillbox, along with two follow-up phone calls. Arm 2 offered a moderate-intensity intervention to 60% of patients (n = 1,200) based on their predicted risk for insulin nonadherence, and included up to six follow-up calls and two calls with a patient’s primary care physician or pharmacy to clarify treatment issues or receive recommendations for therapeutic changes. Patients were also offered enrollment in a weekly text messaging program focused on medication-taking behaviors.

Arm 3 offered a high-intensity intervention to 40% of patients (n = 800) based on glycemic control and predicted risk for insulin nonadherence, including up to 12 follow-up calls and PCPs or pharmacists called as often as necessary. Patients were also offered text messages delivered weekly, every 3 days or daily.

PAGE BREAK

“Arm 3 was designed to mimic the most intensive type of strategy that a telephone-based disease management program could offer,” the researchers wrote.

Primary outcome was continuing to fill prescriptions for insulin. Secondary outcomes were changes in HbA1c and health care utilization. Researchers evaluated outcomes in arms 2 and 3 vs. arm 1 using claims data, intention-to-treat principles and multiple imputation for missing values during 12 months of follow-up.

Among targeted patients, 24.7% in arm 1, 30.7% in arm 2 and 34.3% in arm 3 completed a telephone consultation with a pharmacist.

Patients with an initial consultation received a mean of 1.7 calls in arm 1, 1.8 calls in arm 2 and 2.1 calls in arm 3. Across arms, most patients who received initial consultations self-reported optimal insulin adherence, according to researchers.

In arm 1, the rate of nonpersistence was 5.4%, according to researchers. Compared with that arm, rates of nonpersistence in arms 2 and 3 were 4.7% (RR = 0.88; 95% CI, 0.75-1.03) and 4.9% (RR = 0.91; 95% CI, 0.77-1.06), respectively. Mean time to insulin nonpersistence was 250 days in arm 1, 255 days in arm 2 and 258 days in arm 3.

Researchers did not observe a between-group difference in arms 1 and 2 for HbA1c; however, patients assigned to arm 3 had a lower HbA1c vs. arm 1 (mean absolute difference, –0.25%; 95% CI, –0.43 to –0.06).

There were no between-group differences for total health care spending; however, patients in arm 2 had a higher likelihood of hospitalizations (OR = 1.22; 95% CI, 1.06-1.41) and ED visits (OR = 1.38; 95% CI, 1.24-1.53), according to researchers. There were no between-group differences in hypoglycemia risk.

“Targeting a different patient population might have increased the effectiveness of the interventions that we tested,” the researchers wrote. “For example, we focused on patients at moderate risk of nonpersistence. It is possible that the intervention could have been more effective if targeted to patients at highest risk rather than those who we hypothesized would be more impactable. In arm 2, we also included patients with good baseline glycemic control, hypothesizing that even they could benefit, although this could have diluted the effect.”

The researchers noted that, without a true control group in the study, they could not test for differences between the untargeted, low-intensity intervention vs. no intervention. The findings also may not be fully generalizable to those with Medicare or Medicaid coverage.

PAGE BREAK

“Our results suggest that targeting patient populations for more intensive interventions based both on predicted risk of nonadherence and level of glycemic control has the potential to be more effective than untargeted approaches.” – by Regina Schaffer

Disclosures: Sanofi supported this study. Lauffenburger reports she has received grants from AstraZeneca and Sanofi. Please see the study for all other authors’ relevant financial disclosures.