Estimation of a preference based single index from the sexual quality of life questionnaire (SQOL) using ordinal data

There is increasing interest in using ordinal methods to estimate cardinal values for health states to calculate quality adjusted life years. This paper reports the estimation of models of rank data and discrete choice experiment (DCE) data to derive a preference-based index from a condition specific measure relating to sexual health and to compare the results to values generated from time trade-off valuation (TTO). The DCE data were analysed using a random effects probit model and the DCE predicted values were rescaled according to the highest and lowest predicted TTO values corresponding to the best and worst SQOL health states respectively. The rank data were analysed using a rank ordered logit model and re-scaled using two alternative methods. Firstly, re-scaling the rank predicted values using identical methods to those employed for DCE and secondly, re-scaling the rank model coefficients by dividing each level coefficient by the coefficient relating to death. The study raises some important issues about the use of ordinal data to produce cardinal health state valuations.