OBJECTIVES: The Family Reported Outcome Measure (FROM-16) is a generic tool to measure the previously hidden burden of a person’s health condition on the quality of life (QoL) of their partner or family members. The aim of the study was to assign clinical meanings to FROM-16 scores using the anchor-based approach.

METHODS: A prospective cross-sectional online study recruited family members of patients with different health conditions through 58 UK-based patient support groups, research support platforms (Healthwise Wales, Autism Research Centre Cambridge University database, Join Dementia Research) and Welsh social services departments. Family members completed the FROM-16 and a 5-point Likert scale Global Question (GQ) concerning overall impact of their relative's health condition on their QoL. Multiple FROM-16 band sets were devised by mapping mean, median and mode of the GQ scores against each FROM-16 score and ROC-AUC cut off values. The band set with the best agreement with GQ score based on weighted Kappa (WK) was selected.

RESULTS:

A total of 4,413 family members/partners (male=1,533, 34.7%; female=2,858, 64.8%, unknown=16, 0.4%; other=6, 0.14%) of patients (male=1,994, 45.2%; female=2400, 54.4%; unknown=12, 0.3%; other=7, 0.16%) with > 200 health conditions across 27 medical specialities completed the study: mean FROM-16 score=15.0 (range 0-32, SD=8.1), mean GQ score=2.3 (range 0-4, SD=1.1). The proposed FROM-16 score bands are 0-1=no effect on QoL of family member; 2-8=small effect; 9-16= moderate effect; 17-25=very large effect; 26-32=extremely large effect on QoL of family member (WK=0.6).

CONCLUSIONS:

FROM-16 score banding provides new information to clinicians and researchers about how to clinically interpret scores and score changes, allowing better informed treatment decisions for patients and their families. The now meaningful information from the use of FROM-16 can be used to measure and understand more globally the wider burden of disease, providing a novel dimension to the use of “real world data”.