OBJECTIVES:

The United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model (UKPDS OM) and the risk equations that drive it remain widely used in the economic modeling of type 2 diabetes mellitus. Two versions of the UKPDS OM have been developed and published to-date (OM1 and OM2). The UKPDS OM2 was based on substantially longer follow-up, was derived from more events, and captured more complications and covariates than UKPDS OM1. The objective was to evaluate differences in the performance of the OM1 and OM2 risk factor trajectory equations for glycated hemoglobin (HbA1c) and systolic blood pressure (SBP) in economic evaluations.

METHODS:

The UKPDS OM2 risk equations were incorporated into the PRIME Diabetes Model in addition to the existing UKPDS OM1 risk equations for HbA1c and SBP. The performance of the risk equations from the two UKPDS OM versions was evaluated in the UKPDS baseline cohort, with sensitivity analyses being conducted across a range of mean baseline HbA1c and SBP values. The effects of the analysis time horizon were also investigated in sensitivity analyses. Differences in area under the curve, downstream complication event incidence, and economic outcomes were compared. Base case analyses were conducted over a 50-year time horizon, with and without discounting.

RESULTS:

The UKPDS OM2 HbA1c risk equation resulted in a modest increase in undiscounted life expectancy versus the corresponding UKPDS OM1 equation (increasing from 41.179 to 41.260 years), with the difference driven by higher HbA1c earlier in the simulation and lower HbA1c later in the simulation. Differences in complication incidence were also modest; the largest incremental cumulative incidence of any modeled complication was 0.4% (stroke and amputation) over the 50-year time horizon.

CONCLUSIONS:

The analyses showed that using risk equations from the UKPDS OM2 has only a modest effect on modeled clinical and economic outcomes relative to equations from the UKPDS OM1.