The effect of comorbid illness and functional status on the expected benefits of intensive glucose control in older patients with type 2 diabetes: a decision analysis

Abstract

Background: Physicians are uncertain about when to pursue intensive glucose control among older patients with diabetes.

Objective: To assess the effect of comorbid illnesses and functional status, mediated through background mortality, on the expected benefits of intensive glucose control.

Design: Decision analysis.

Data sources: Major clinical studies in diabetes and geriatrics.

Target population: Patients 60 to 80 years of age who have type 2 diabetes and varied life expectancies estimated from a mortality index that was validated at the population level.

Time horizon: Patient lifetime.

Perspective: Health care system.

Intervention: Intensive glucose control (hemoglobin A1c [HbA1c] level of 7.0) versus moderate glucose control (HbA1c level of 7.9).

Outcome measures: Lifetime differences in incidence of complications and average quality-adjusted days.

Results of base-case analysis: Healthy older patients of different age groups had expected benefits of intensive glucose control ranging from 51 to 116 quality-adjusted days. Within each age group, the expected benefits of intensive control steadily declined as the level of comorbid illness and functional impairment increased (mortality index score, 1 to 26 points). For patients 60 to 64 years of age with new-onset diabetes, the benefits declined from 106 days at baseline good health (life expectancy, 14.6 years) to 44 days with 3 additional index points (life expectancy, 9.7 years) and 8 days with 7 additional index points (life expectancy, 4.8 years). A similar decline in benefits occurred among patients with prolonged duration of diabetes.

Results of sensitivity analysis: With alternative model assumptions (such as Framingham models), expected benefits of intensive control declined as mortality index scores increased.

Limitations: Diabetes clinical trial data were lacking for frail, older patients. The mortality index was not validated for use in predicting individual-level life expectancies. Adverse effects of intensive control were not taken into account.

Conclusion: Among older diabetic patients, the presence of multiple comorbid illnesses or functional impairments is a more important predictor of limited life expectancy and diminishing expected benefits of intensive glucose control than is age alone.

Figure 1. Model of diabetes complications in older patients

The structure of the decision analytic model is presented in this figure. Hypothetical patients move through the model from left to right for each cycle length (one year). Based on initial patient clinical characteristics, patients are subject to the risk of various complications related to diabetes as well as mortality. Patients who survive a given year repeat the cycle until death.

Figure 2. Expected quality of life benefits of intensive glucose control for 60–64 year old and 75–79 year old patients

**Level of comorbid illness and functional impairment is indicated by additional points on the mortality index score (1–2 points per illness or impairment). Expected benefits for 65–69 year old and 70–74 year old patients are intermediate to those of 60–64 year old and 75–79 year old subgroups (see Technical Appendix).

Figure 3. Expected differences in lifetime incidence of specific complications for 60–64 year old and 75–79 year old patients

**Level of comorbid illness and functional impairment is indicated by additional points on the mortality index score (1–2 points per illness or impairment). The relationships between absolute risk reductions for specific events and the mortality index score are not consistently monotonic because we assessed fairly wide ranges of duration of diabetes (5 years) and the individual complication models vary in their responsiveness to this variable. Expected differences for 65–69 year old and 70–74 year old patients are intermediate to those of 60–64 year old and 75–79 year old subgroups (see Technical Appendix).