Article Text
Abstract
Introduction and objectives We aimed to study quantitatively the lung cancer pathway from first symptom to treatment and to explore relationship between symptoms, delay and survival.
Methods Newly diagnosed lung cancer patients, referred to Oncology clinics in Norfolk 2008–2012, completed systematic questionnaires regarding date of onset of each symptom, MRC dyspnoea score and% weight loss. GPs also completed questionnaires. Additional patient, pathway and tumour data were retrieved from hospital records.
The cancer pathway was recorded in five phases: 1) first symptom to GP presentation, 2) to secondary care (SC) referral, 3) to SC appointment, 4) to MDT meeting or date of diagnosis and 5) to treatment commencing.
Results Of 379 patients, mean age was 70.1 years, staging was: I-II (13.7%), III (34.3%) and IV (52%). Cohort survival was 6.1% with minimum follow-up of 39 months.
Mean phase lengths were 221.8, 45.8, 10.7, 21.3, 34.7 days for phases 1–5 respectively. Phase 1 was significantly longer (p < 0.01). Mean phase 1 was shorter in stage III-IV than I-II, 200 and 245 days respectively (p < 0.05), in ex-/never-smokers (191.6 days) than smokers (264.2 days) (p = 0.022) and if first symptom was haemoptysis compared to cough or dyspnoea. 36.9% patients felt they delayed seeing their GP. Commonest reasons were: thinking symptoms were insignificant (35%), anxiety (28.6%) and denial (20%). Good correlation was seen between patient and GP reported dates.
Reported symptoms included (% initial symptoms in parenthesis): cough 71% (62.8%), dyspnoea 62.8% (27.2%, chest pain 37.7% (8.7%), haemoptysis 28% (4.2%).
Symptoms conferring increased hazard of death were defined as B symptoms. These were: grade 4/5 dyspnoea, hoarseness or loss of voice, metastatic pain and systemic symptoms (HRs 1.77, 1.53, 2.21, 1.93 respectively, p < 0.001). Patients with initial cough/mild dyspnoea have means of 127/210 days before B symptoms develop.
Overall no relation was found between phase lengths 1–4 and survival. Survival increased if phase 5 was >31 compared to ≤31 days (HR 0.74, p = 0.006).
Conclusions Phase 1 is longest. There is no relation between phase length and survival except in phase 5. Symptoms are more important to survival than delay. Effective therapy started within 3 months (before B symptoms) could increase survival.