Lung cancer on Haugalandet, Norway. A long term follow-up study

Abstract

Background: Little was known on long term survival, the use of hospital days, symptoms both initially and terminally, and the quality of lung cancer care for the population of the patients with lung cancer in Haugalandet, Norway. Material and methods: In a retrospective study we followed a cohort of all incident lung cancer patients from 01.01.1990 to 31.12.1996 in the hospital area of Haugesund hospital (Haugalandet). To study the predictors for long term survival we followed the patients either to death or to the last follow up to 31.12.2008. All hospital admissions and hospitalization days were recorded for all patients up to 01.12.2003. We studied the terminal symptoms in the last eight weeks of patients’ lives who died before that date. Finally, we compared for the same time period four quality indicators in the patient cohort from the local hospital-based lung cancer registry (LCR) with the patient cohort in the Cancer Registry of Norway (CRN) both from the same geographical area (Haugalandet), and a cohort of all lung cancer patients from the rest of Norway. Results: A total of 271 patients were diagnosed with lung cancer in these seven years. The long term survival was poor, with one-year survival of 29.2% and five- and ten-year survival of 8.5% and 5.5%, respectively. The median (IQR) survival time was 5.7 (1.9,14.1) months. No weight loss, young age, limited stage, good performance status and surgical treatment were predictors for long survival, which were also not influenced by the diagnostic delay time. Furthermore, we found that all 271 patients had a median number (inter quartile range, IQR) of 3 (2,5) admissions and 35 (18,58) hospitalization days. Those who did not survive spent 19% of their remaining life time in institutions. Young age, limited disease and good performance status were associated with high number of hospital days, but these effects were not significant when adjusted for treatment. Information on symptoms in the terminal 8 weeks was obtained in 247 of the patients who died before 1st December 2003. Pain was observed in 85%, psychological symptoms (anxiety, insomnia and/or depression) in 71%, dyspnea in 54%, neurologic symptoms in 28%, cough in 24%, nausea in 21%, and hemoptysis in 9%. Young age and small cell cancer (SCLC) were risk the factors for psychological symptoms, and initial stage III disease was a risk factor for terminal dyspnea. Terminal cough was associated with NSCLC and nausea with SCLC. The average minimal difference of clinical importance judged by 26 physicians for four lung cancer quality indicators (histological/cytological verification, staging, surgery and one year survival) varied from 18% to 23% from the national average of the indicators. The level of the four quality indicators studied was in agreement with the patient cohort from Haugalandet and the cohort from the rest of Norway. However, the sample sizes necessary to detect a 20% difference from the national average (power 0.80, p<0.05) varied from 435 to 2826 cases depending on the prevalence of the indicators. Conclusions: The long term survival of lung cancer is poor, with only less than 10% of the patients surviving after five years. The patients who died spent one fifth of their remaining time after diagnosis in health care institutions. Those with limited disease and young age had most days in the hospital. All patients had one or more symptoms requiring treatment in the terminal stage of the disease, and pain was the most frequent symptom in 85% of the patients. The quality of lung cancer care is difficult to evaluate in small management units. The small decentralized units should thus be merged to larger cooperative units with standardized routines. A national quality registry on lung cancer would then be a powerful tool contributing to improved quality of local lung cancer care.