Measuring patient and carer outcomes from genome sequencing: A systematic review of outcome measurement instruments

Background

Genome sequencing is an advanced technology that can be used to predict, screen for and diagnose diseases (including rare diseases, cancers and infectious diseases), and identify drug targets. However, the information uncovered is complex and can impact patients and informal carers across clinical, emotional, cognitive, behavioural and social outcomes. These outcomes have been measured using a variety of methods, including intermediary measures (such as diagnostic yield), direct elicitations of preference‑based utility (such as via discrete choice experiments [DCEs]), generic and genome‑sequencing‑specific outcome measurement instruments, and qualitative methods. Groups such as the Phenotypes and eXposures (PhenX) Toolkit Genomic Medicine Implementation (GMI) Domain Working Group and the Clinical Sequencing Evidence‑Generating Research (CSER) Consortium have sought to harmonise the instruments used in genome sequencing research projects. However, despite these efforts, no studies have sought to systematically identify the instruments used to measure outcomes across the breadth of genome sequencing’s potential applications.

Aims

This study aimed to systematically identify the patient‑ and carer‑reported instruments used to measure outcomes from genome sequencing.

Methods

Primary studies measuring patient‑ and carer‑reported outcomes from genome sequencing and published by 15 December 2023 (initial search) and 19 March 2026 (updated search) were included. Eligibility was unrestricted by setting (i.e., clinical, research, direct‑to‑consumer), testing purpose (i.e., diagnosis, screening), disease area (i.e., rare disease, cancer) and instrument validation status (i.e., validated, unvalidated).

Preliminary results (as of 15 December 2023)

Thirty‑six articles representing 29 studies were included. These studies used a total of 209 patient‑ or carer‑reported outcome measurement instruments (of which 75 were validated). Of these, 63 instruments (of which 49 were validated) met the eligibility criteria and were used 106 times (mean: 1.7; range: 1–9) across the included studies. Each included study used 2.2 included instruments on average (range: 1–12). Only four (13.8%) of the 29 included studies intended to develop and/or validate an instrument, and three reported psychometric results.

Preliminary conclusions

This study highlights the variability in approaches used to measure outcomes from genome sequencing, and the lack of psychometric data available to support instrument‑selection decisions within this context. This is important, as it may limit the ability to generate robust, pooled estimates of outcomes and/or to compare outcomes across studies and populations.