OBJECTIVES : Heavy menstrual bleeding is the main symptom of uterine fibroids (UF), significantly affecting quality of life of women with symptomatic UF, and is measured in most clinical trials involving UF patients. The objective was to develop a predictive utility function based on incremental changes in menstrual blood loss (MBL) to estimate QoL in patients with symptomatic UF for use in health economic models.

METHODS : A parsimonious regression model was developed to promote replicability and transparency while utilizing the most important clinical factors in UF patients such as MBL. An ordinary least squares (OLS) regression was parameterized on patient-level data from clinical trials. The OLS model included MBL volume and age as parameters and EQ-5D index values as outcome. EQ-5D values were generated from Uterine Fibroid Symptom and Quality of Life (UFS-QoL) measures. Within-patient correlation was not modelled, and all patient observations were pooled together to ensure all data was used.

RESULTS : The OLS model was fitted using 1,706 observations. The estimated intercept was 0.69568, the coefficient for MBL volume in mL was -0.0003877 and the coefficient for age was 0.00296. Thus, the model predicts that, all else equal, a 100 mL decrease in MBL volume increases EQ-5D index value by 0.04 and a one-year increase in age corresponds to a 0.003 increase in EQ-5D.

CONCLUSIONS : Improvement in MBL is predicted to be associated with proportional improvements in the overall quality of life of UF patients. The prediction model offers a simple and transparent way to estimate quality of life based on changes in MBL for direct use in health economic modelling.