Clinical effectiveness and cost-effectiveness of laparoscopic surgery for colorectal cancer: systematic reviews and economic evaluation.

OBJECTIVE

The aim of this study was to determine the clinical effectiveness and cost-effectiveness of laparoscopic, laparoscopically assisted (hereafter together described as laparoscopic surgery) and hand-assisted laparoscopic surgery (HALS) in comparison with open surgery for the treatment of colorectal cancer.

DATA SOURCES

Electronic databases were searched from 2000 to May 2005. A review of economic evaluations was undertaken by the National Institute for Health and Clinical Excellence in 2001. This review was updated from 2000 until July 2005.

REVIEW METHODS

Data from selected studies were extracted and assessed. Dichotomous outcome data from individual trials were combined using the relative risk method and continuous outcomes were combined using the Mantel-Haenszel weighted mean difference method. Summaries of the results from individual patient data (IPD) meta-analyses were also presented. An economic evaluation was also carried out using a Markov model incorporating the data from the systematic review. The results were first presented as a balance sheet for comparison of the surgical techniques. It was then used to estimate cost-effectiveness measured in terms of incremental cost per life-year gained and incremental cost per quality-adjusted life-year (QALY) for a time horizon up to 25 years.

RESULTS

Forty-six reports on 20 studies [19 randomised controlled trials (RCTs) and one IPD meta-analysis] were included in the review of clinical effectiveness. The RCTs were of generally moderate quality with the number of participants varying between 16 and 1082, with 10 having less than 100 participants. The total numbers of trial participants who underwent laparoscopic or open surgery were 2429 and 2139, respectively. A systematic review of four papers suggested that laparoscopic surgery is more costly than open surgery. However, the data they provided on effectiveness was poorer than the evidence from the review of effectiveness. The estimates from the systematic review of clinical effectiveness were incorporated into a Markov model used to estimate cost-effectiveness for a time horizon of up to 25 years. In terms of incremental cost per life-year, laparoscopic surgery was found to be more costly and no more effective than open surgery. With respect to incremental cost per QALY, few data were available to differentiate between laparoscopic and open surgery. The results of the base-case analysis indicate that there is an approximately 40% chance that laparoscopic surgery is the more cost-effective intervention at a threshold willingness to pay for a QALY of pound 30,000. A second analysis assuming equal mortality and disease-free survival found that there was an approximately 50% likelihood at a similar threshold value. Broadly similar results were found in the sensitivity analyses. A threshold analysis was performed to investigate the magnitude of QALY gain associated with quicker recovery following laparoscopic surgery required to provide an incremental cost per QALY of pound 30,000. The implied number of additional QALYs required would be 0.009-0.010 compared with open surgery.

CONCLUSIONS

Laparoscopic resection is associated with a quicker recovery (shorter time to return to usual activities and length of hospitalisation) and no evidence of a difference in mortality or disease-free survival up to 3 years following surgery. However, operation times are longer and a significant number of procedures initiated laparoscopically may need to be converted to open surgery. The rate of conversion may be dependent on experience in terms of both patient selection and performing the technique. Laparoscopic resection appears more costly to the health service than open resection, with an estimated extra total cost of between pound 250 and pound 300 per patient. In terms of relative cost-effectiveness, laparoscopic resection is associated with a modest additional cost, short-term benefits associated with more rapid recovery and similar long-term outcomes in terms of survival and cure rates up to 3 years. Assuming equivalence of long-term outcomes, a judgement is required as to whether the benefits associated with earlier recovery are worth this extra cost. The long-term follow-up of the RCT cohorts would be very useful further research and ideally these data should be incorporated into a wider IPD meta-analysis. Data on the long-term complications of surgery such as incisional hernias and differences in outcomes such as persisting pain would also be valuable. Once available, further data on both costs and utilities should be included in an updated model. At this point, further consideration should then be given as to whether additional data should be collected within ongoing trials. Few data were available to assess the relative merits of HALS. Ideally, there should be more data from methodologically sound RCTs. Further research is needed on whether the balance of advantages and disadvantages of laparoscopic surgery varies within subgroups based on the different stages and locations of disease. Research relating to the effect of experience on performance is also required.