Weekly patient-reported symptom monitoring was linked to longer survival in lung cancer

Design

Multicentre, stepped-wedge, cluster-randomised trial; post-hoc survival analysis

Population

446 patients starting treatment for stage I-IV lung cancer, 14 Dutch hospitals

Primary outcome

Overall survival (post-hoc)

Effect

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

SYMPRO-Lung was a stepped-wedge cluster-randomised trial at 14 Dutch hospitals. Each hospital began with usual care and switched at a randomised time to weekly online symptom questionnaires for a year for patients starting treatment for stage I to IV lung cancer. The primary outcome, quality of life, was reported earlier; this post-hoc analysis examined survival.

Of 515 patients enrolled, 446 were included, with a median follow-up of 50 months. Median overall survival was 26 months with monitoring and 21 months with usual care. After adjustment, the hazard ratio for death was 0.80 (95% CI 0.63 to 1.00, p=0.049). Progression-free survival did not differ significantly (HR 0.86).

The survival analysis was not planned in advance and only just reached significance. But it matches earlier trials in advanced cancer, and the intervention is inexpensive. Weekly monitoring appears to help by catching complications early, and the principle can be adapted to phone or messaging follow-up where online forms are impractical.

• Consider structured weekly symptom reporting for patients on lung cancer treatment, online, by phone or by messaging.
• Set clear thresholds that trigger a call back or an earlier review.
• Focus the questions on breathlessness, pain, fatigue, appetite, cough and fever.
• Make sure someone is responsible for reviewing reports every week, or the system adds nothing.

Why it matters

A low-cost change to follow-up was associated with longer survival, not only better quality of life.

Don't overread it

Survival was a post-hoc outcome with a borderline result; it supports adoption, not certainty about the size of benefit.

The statistics, in plain English

A hazard ratio of 0.80 means deaths occurred at about four-fifths the rate with monitoring. The upper limit of the confidence interval is 1.00, so the result is at the edge of chance, and a post-hoc analysis is more prone to chance findings than a planned one.

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.