The Effects of Type 2 Patients Refusing Insulin 

We need to understand the relationship between patients with type 2 diabetes refusing insulin therapy, and worsening blood glucose levels. 

The progression of type 2 diabetes can come with severe consequences if measures to treat make no progress with the disease. Although there are numerous methods to help control or maintain the blood glucose of a person with diabetes, the agents have adverse effects, ceiling effects in lowering the blood glucose or HbA1c, and contraindications, which lessen the value of these options. Insulin, which is usually reserved for uncontrolled diabetes, has fewer contraindications, and is very efficient in lowering blood glucose. However, the initiation of insulin is often postponed and leads to poor glycemic control. Understanding and preventing the causes of these deferred insulin treatments could decrease the prevalence of disease-related complications and promote better and longer-lasting glycemic control. Recent studies have linked the declination of recommended insulin therapy with delayed initiation. The research reported in the “Predictors and Consequences of Declining Insulin Therapy by Individuals with Type 2 Diabetes” serves to examine the refusal of insulin treatment in type 2 diabetes and its correlation with glycemic control.  

Cohorts were established in this retrospective study to ascertain if patients with type 2 diabetes who declined insulin treatment have worse glycemic control than those who initiate insulin when recommended by healthcare providers, and to inspect the characteristics and reasoning behind the increased incidence of refusing insulin treatment in the targeted population. Between January 1, 2000, and December 31, 2014, adult patients with type 2 diabetes were seen at Massachusetts General Hospital and Brigham Women’s Hospital by primary care physicians. The inclusion criteria for this study consisted of being 18 years of age, insulin treatment-naïve, having no prior recommendations of insulin therapy, and an HbA1c of seven percent for a minimum of 12 months before the initial recommendation of insulin during the study. Subjects with type 1 diabetes diagnosis, a baseline GFR<30 mL/min, missing zip codes, missing HbA1c post-insulin recommendation were excluded.  

Informed consent was waived, and demographics along with medication and lab work were acquired from electronic medical records (EMRs). The primary outcome was the time it took to obtain an HbA1c of less than seven. Declines in the treatment had to have been documented in outpatient provider notes. The acceptance of the recommended insulin therapy was defined by a prescription for insulin with no documentation of refusal. The Charlson Co-morbidity Index was calculated, excluding any items related to diabetes, and natural language processing via Canary technology was utilized to extract the desired information from the EMRs at a rapid pace.  

Statistical analysis in this study involved categorical data, including proportions and frequencies, continuous variables, means, medians, and standard deviations. Comparisons were made using long-rank tests to analyze the Kaplan0Meier cumulative incidence curves, which measured the time to control the patients’ HbA1c. To guesstimate the correlation between the previously stated measure and a refusal of insulin treatment, a cox proportional hazards model was applied. A sensitivity analysis using inverse probability weighting was conducted to estimate the propensity scores, which evaluated the probability of treatment group assignment. The construction of a logistic regression model examining multiple variables was used to study forecasters of a decline in insulin treatment. P-values were interpreted with the method of Simes-Hochberg. 

Approximately 42.7% of the 5307 subjects in the study refused insulin therapy, in which it took 50 months to reach an HbA1c of seven. Those who initiated insulin upon recommendation were able to obtain the targeted glycemic control in 12 months less than the other cohort (P < 0.001). The multivariable analysis revealed declining insulin treatment correlated to control of HbA1c hazard ratio of 0.89 (95% CI 0.82 to 0.97; P = 0.008). Incidence of deferring insulin therapy was more common in the older subjects (OR 0.81; 95% CI 0.76 to 0.86; P < 0.001) and in patients taking more diabetes medications (OR 0.78; 95% CI 0.74 to 0.83; P < 0.001). Acceptance was higher amongst patients with diabetes-related complications (OR 1.32; 95% CI 1.13 to 1.53; P < 0.001) and higher HbA1c (OR 1.10; 95% CI 1.07 to 1.13; P < 0.001). 

The findings of this study validated the hypothesis that refusal of recommended insulin treatment by patients with uncontrolled type 2 diabetes is commonly correlated with subpar glycemic control. Future studies could focus on establishing optimal alternative therapies for this patient population to promote obtaining clinical outcomes such as reaching target HbA1c and blood glucose measures.    

Practice Pearls:   

• Recent studies have shown poor glycemic control in type 2 diabetes is associated with delayed insulin treatment.   
• This study revealed those who decline insulin could take 12 months longer to achieve a HbA1c < 7% than patients who initiate insulin therapy right after recommendation by a primary care provider.   
•  Rejection of insulin therapy is more likely in older patients with lower HbA1c and fewer disease-related complications.   

Reference for “The Effects of Type 2 Patients Refusing Insulin”:
Shubina, A. Turchin N. Hosomura H. Zhang S. Malmasi M. “Predictors and consequences   of declining insulin therapy by individuals with type 2 diabetes .” Diabetic Medicine, (2020).   

Mia Flowers, PharmD. Candidate of Florida Agricultural & Mechanical University School of Pharmacy