Pretreatment quality of life carried prognostic information in head and neck cancer

Design

Systematic review and random-effects meta-analysis of 32 studies (PROSPERO registered)

Population

Patients with head and neck cancer with pretreatment patient-reported quality of life

Primary outcome

Overall survival by quality-of-life domain, per 10-point worsening

Effect

Global QoL HR 1.11 (95% CI 1.08 to 1.14); physical functioning HR 1.16 (1.13 to 1.20)

This meta-analysis pooled 32 independent longitudinal studies (43 estimates across 10 quality-of-life domains) on patient-reported quality of life before treatment and overall survival in head and neck cancer. Hazard ratios were harmonised to a 10-point worsening of each domain.

Worse global quality of life (HR 1.11, 95% CI 1.08 to 1.14), physical functioning (1.16, 1.13 to 1.20), role functioning (1.09, 1.07 to 1.11) and social functioning (1.09, 1.06 to 1.12) were consistently associated with higher mortality and stayed stable when each study was left out in turn. Swallowing (1.15, 1.02 to 1.29) depended on a single study, and dyspnoea (1.16, 1.01 to 1.33) was heterogeneous and unstable.

The authors suggest these domains may help prioritise supportive-care assessment before treatment. They state that the less stable findings should not be treated as targets. The associations were adjusted but remain observational.

• Record patient-reported quality of life, including physical, role and social functioning, before treatment.
• Use low scores to prompt supportive-care assessment such as nutrition and rehabilitation.
• Do not use quality-of-life scores to withhold treatment.
• Treat swallowing and dyspnoea findings as unconfirmed.
• Revisit the same measures during follow-up so change can be tracked.

Why it matters

A questionnaire completed in the waiting room carries prognostic information that complements tumour stage.

Don't overread it

Observational associations; it does not show that improving quality of life before treatment improves survival.

The statistics, in plain English

A hazard ratio of 1.11 per 10-point worsening means about 11% higher risk of death for each ten-point drop in global quality of life, after adjustment. A leave-one-out check removes each study in turn to see if the result survives; the weaker domains did not pass it.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.