We thought the answer was YES.
The effectiveness outcomes (from RCTs and UKPDS) were measured in surrogate outcomes (mmHg, mg cholesterol, glycaemic levels). There is evidence that these correlate well with clinical outcomes.
In order to be incorporated into the model (i.e. how do transition probabilities change), these results need to be converted into a risk reduction. This is described for glycaemic control on p2543, 2nd column, 4th paragraph. For hypertension, the authors make an assumption on risk reduction based on UKPDS data (p2543, 3rd column, 3rd paragraph). For cholesterol, the authors modelled the risk reduction based on two trials (p2545, 1st column, 3rd paragraph).
Little information is given on how the QALYs were obtained but there are references to the sources.
Years of life gained are also used. Utility values are given for various health states (e.g. 0.69 for blindness). The references for these may be incomplete: does ref 25 provide the utilities for blindness, ERD, and lower extremity amputation or just for amputation? It is important to know the validity of these as the ascribed utility is an important parameter when calculating the incremental cost/QALY.
Data on distribution of patients at diagnosis and transition probabilities were taken from the UKPDS study, previous referenced models and studies contained within the technical report. Utility data used to estimate QALYs were referenced (but was this complete?)
The authors assumed that data on progression for type 1 diabetes could be used for type 2 diabetes (p 2543, 2nd column, 4th paragraph)