Weekly online symptom reporting linked to longer survival in lung cancer

Design

Post hoc survival analysis of a stepped-wedge cluster-randomised trial

Population

446 patients with stage I–IV lung cancer at 14 Dutch hospitals

Primary outcome

Overall survival

Effect

Median 26 vs 21 months; adjusted HR 0.80 (95% CI 0.63 to 1.00)

SYMPRO-Lung was a stepped-wedge, cluster-randomised trial in 14 Dutch hospitals. Patients with stage I–IV lung cancer starting a new treatment completed weekly online symptom questionnaires for a year once their hospital crossed over to the intervention; before that, they received standard care. The primary endpoint, quality of life, was reported earlier. This post hoc analysis followed 446 patients for a median of 50 months.

Median overall survival was 26 months with symptom monitoring and 21 months without. After adjustment, the hazard ratio was 0.80 (95% CI 0.63 to 1.00). Progression-free survival did not differ significantly (HR 0.86).

This extends earlier evidence from advanced cancer to all stages and to routine hospital care. It is post hoc, the confidence interval touches 1.0, and 13% of patients could not be included. But symptom monitoring is cheap, low-risk and already supported on quality-of-life grounds, which makes it reasonable to adopt where it can be run reliably.

• Consider weekly patient-reported symptom monitoring for patients on lung cancer treatment
• Agree in advance which reported symptoms trigger a call from the team
• Make sure someone reviews alerts promptly; monitoring without response adds nothing
• Offer a telephone option where patients cannot use online questionnaires
• Include patients at all stages, not only those with advanced disease

Why it matters

A low-cost process change was associated with months of extra survival, alongside the quality-of-life benefit already shown.

Don't overread it

This was a post hoc analysis with a confidence interval reaching 1.0, so the survival benefit is suggestive rather than established.

The statistics, in plain English

A hazard ratio of 0.80 means about 20% lower risk of death at any given time. The upper limit of exactly 1.00 means the data are also consistent with no benefit, and post hoc analyses are more prone to chance findings.

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.