Laparoscopic and open surgery for colon cancer : studies on costs and health related quality of life

Abstract: Colon cancer is one of the most common forms of cancer in the Western world, and the basis for treatment is surgery. Laparoscopic colon resection (LCR) has been developed since the early 1990s. Short- and longterm outcomes between LCR and open colon resection (OCR) have been compared in randomised controlled trials (RCTs). These studies have demonstrated benefits in short-term clinical outcomes for LCR and equality on the primary long-term outcome measure, i.e. cancer-free survival. If no difference in cancer-free survival can be demonstrated, costs and patient benefits in terms of healthrelated quality of life (HRQL) become important secondary outcome measures. The COLOR trial (COlon cancer Laparoscopic or Open Resection) is the most comprehensive of the RCTs performed. It is a European multi-centre study with 29 centres participating and includes 1082 patients for analysis. The studies presented in this thesis were conducted on the Swedish patients included in the COLOR trial, representing 39% of the entire trial population. In Paper I total costs to society were calculated for the first 12 weeks after surgery. Totally, 2 10 patients (98 LCR and 112 OCR) from 10 centres were included. No significant difference in total cost between groups was found (LCR vs. OCR difference of means E 1846, p=0. 104). However, LCR was more expensive than OCR at operation (E 1 171, p<0.001), for the first admission (E 1556, p=0.015) and for the health care system (E 2 244, p=0.018). In Paper II the method to retrieve data in the cost study was validated. For the first 40 patients to complete the protocol (21 LCR, 19 OCR), data collected by case record forms (CRF), patient diaries and telephone surveys were compared with corresponding data from medical records and social security offices. Complications were underreported, but such items as length of hospital stay, instrument costs, re-operations and time off work, all heavily associated with total costs, were reported with high accuracy. In Paper III HRQL was estimated pre-operatively and 2, 4 and 12 weeks post-operatively by employing two validated HRQL instruments (EORTC QLQ-C30 and EQ-5D). Totally, 285 patients (130 LCR and 155 OCR) from 7 Swedish hospitals were included. In the QLQ-C30 instrument there were statistically significant benefits of LCR in social function at two and four weeks and in role function at two weeks. In Paper IV the aim was to determine whether there was a selection bias of patients into the COLOR trial in Sweden, given the trial criteria of eligibility and exclusion. Further aims were to determine the reasons of exclusion from the trial and to calculate a possible caseload of patients suitable for LCR in Sweden. In the eight Swedish hospitals with the highest inclusion into the COLOR trial 2285 patients underwent surgery during the inclusion period of the trial. When comparing patients included in the COLOR trial (n=391) with eligible patients that were not included ("missed inclusions", n=335), no difference was found on age, gender, body mass index (BMI) and site of tumour. However, the included patients had more severe tumour stage according to TNM (p=0.00 15) and more severe co-morbidity according to ASA (p<0.0001). The caseload for LCR for cancer in Sweden was calculated to 50-60% of all patients. Concerning colon cancer, LCR has been demonstrated in several RCTs to be safe, offer equal cure for cancer and benefits in short-term clinical outcomes relative to OCR. LCR offers HRQL benefits during the first post-operative month and total costs to society have not been demonstrated as being higher. Most Swedish hospitals would have a sufficient caseload of LCR patients, and provided proper training is available, LCR can be recommended in clinical praxis.