OBJECTIVES: A recently conducted systematic review including 76 type-2-diabetes (T2D) cost-effectiveness studies utilizing 10 different simulation models, reported a consistent relationship between initial HbA1c intervention effect-size (HbA1c-IES) and simulated, undiscounted health-benefits (0.642 and 0.371 increase in life-expectancy (LE) and quality-adjusted-life-expectancy (QALE) per 1%-point HbA1c decrease, respectively). Beyond HbA1c-IES, other aspects of the model configuration such as employed cardiovascular-risk-equations (CV-RE), consideration of a long-term HbA1c-treatment-effect-decay (HbA1c-TED) and study population represent important determinants of this relationship. The objective of our study was to illustrate the HbA1c/outcome relationship across a series of possible model configurations.

METHODS: The QuintilesIMS-CORE-diabetes-model (CDM, version 9.0) was used to project undiscounted lifetime gain in LE and QALE per 1%-point HbA1c reduction considering the following configurations: Patient characteristics, obtained from NHANES patient-level-data, were stratified across three age groups (Young (<=45yrs), intermediate (46-64yrs), old (>=65yrs)); three alternative sets of CV-RE were applied (UKPDS-Outcomes-Model-1 (OM1), UKPDS-Outcomes-Model-2 (OM2), Swedish-National-Diabetes-Registry (SNDR)); HbA1c time trajectories were assessed via a random-effects-panel-equation from UKPDS (considering HbA1c-TED over 5-10 years) or a steady increase of 0.15%-points per year (considering no HbA1c-TED).

RESULTS: Across all configurations explored, the CDM predicted a mean increase in LE and QALE of 0.56 and 0.49, respectively. Benefits were higher when no HbA1C-TED was considered (mean increase in LE and QALE of 0.98 and 0.83 (no- HbA1c-TED) vs. 0.15 and 0.15 (included- HbA1c-TED)). LE and QALE benefits also declined with increasing age (0.88, 0.59 and 0.22 LE gain and 0.79, 0.5 and 0.19 QALE gain for young, intermediate and old patients, respectively). Differences for alternative CV-RE were projected at 0.64, 0.54 and 0.51 LE gain and 0.57, 0.46 and 0.45 QALE gain for OM1, OM2 and SNDR, respectively.

CONCLUSIONS: Patient age, HbA1c-TED, and CV-RE represent important components (beyond HbA1c-IES) for an appropriate characterization of the HbA1c/outcome relationship in T2D-simulation-models.