This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.

Bariatric surgery for the treatment of morbid obesity was compared across high-volume (more than 100 cases/year), medium-volume (50 - 100 cases/year) and low-volume (fewer than 50 cases/year) hospitals.

Treatment.

Cost-effectiveness analysis.

The study population comprised patients who had undergone Roux-en-Y gastric bypass (bariatric surgery) for the treatment of morbid obesity.

The setting was secondary care. The economic study was carried out in the USA.

The effectiveness data were collected between 1 January 1999 and 31 December 2002. The price year was unclear.

The effectiveness data were obtained from the University Health System Consortium Clinical (UHC) database.

The effectiveness data and cost data were directly linked on the UHC database.

No power calculations were reported. The study sample was selected using data from the UHC database. Data were collected on all patients who underwent Roux-en-Y gastric bypass for the treatment of morbid obesity. These patients were identified using the International Classification of Diseases 9th edition (ICD9). To estimate the number of procedures performed laparoscopically, all discharges that included a code for diagnostic laparoscopy, laparoscopic lysis of adhesions, or laparoscopic cholecystectomy were used. The total number of cases was 13,180 in the high-volume group, 7,634 in the medium-volume group and 2,722 in the low-volume group.

This was a multi-centre, historical, cross-sectional study. Data from 93 academic centres across the USA were examined.

The basis of the analysis was treatment completers only. The primary health outcomes used were perioperative outcomes (mean length of stay, complications and readmissions) and the in-hospital mortality rates. There were a number of significant differences in the groups of patients. Compared with low-volume hospitals, the high-volume group contained a lower number of laparoscopic cases and a lower number in the severity classes major and catastrophic. There were also higher numbers of white and "other" races and of low severity cases in the high-volume group (in comparison with low-volume group).

The mean length of stay was 3.8 (+/- 2.9) days in the high-volume group, (p<0.01 compared with low volume), 4.4 (+/- 3.2) days in the medium-volume group, and 5.1 (+/- 4.0) days in the low-volume group.

The percentage of overall complications was 10.2% in the high-volume group, (p<0.01 compared with low volume), 12.3% in the medium-volume group, and 14.5% in the low-volume group.

The 30-day readmission rate was 0.3% in the high-volume group, (p<0.05 compared with low volume), 0.3% in the medium-volume group, and 0.6% in the low-volume group.

In-hospital mortality was 0.3% in the high-volume group compared with 1.2% in the low-volume group, (p<0.01).

There was no significant difference between the high- and medium-volume groups, or between the medium- and low-volume groups. There were significant differences between the high- and low-volume groups for the mean length of stay, overall complications, 30-day readmission rates and in-hospital mortality.

The health outcomes were left disaggregated and no summary benefit was used. In effect, a cost-consequences analysis was conducted.

It would appear that the only cost used in the study was that of the operation. There were no details of what this cost covered. The source of the costs was the UHC database. Discounting was not relevant. The price year was unclear.

No statistical analysis was undertaken.

The indirect costs were not included in the study.

US dollars ($).

No sensitivity analysis was undertaken.

See the 'Effectiveness Results' section.

The mean cost of surgery was $10,292 (+/- 6,680) in the high-volume group, (p<0.01 compared with low volume), $11,619 (+/- 7,899) in the medium-volume group, and $13,908 (+/- 9,573) in the low-volume group.

Not relevant as a cost-consequences analysis was carried out.

In-hospital mortality was lower in hospitals that performed more than 100 bariatric surgical cases per year. High-volume hospitals also had lower readmissions, lower preoperative morbidity and decreased costs.

The rationale for the choice of the comparator was clear. The purpose of the study was to examine the outcomes of surgery, depending on the volume of cases undertaken. The method for categorising hospitals was clearly explained.

The analysis of effectiveness was based on a historical cross-sectional study, which appears to have been appropriate for the study question. The study sample was representative of the study population, but there were differences in the study groups. Most notably, there were significantly fewer patients with major or catastrophic class of surgery in the high-volume group.

No summary measure of benefit was used in the economic analysis. The analysis was therefore categorised as a cost-consequence study. The reader is thus referred to the comments in the 'Validity of estimate of measure of effectiveness' field (above).

The costing was conducted from the perspective of the health service. The source of the costs was not explicitly stated, but it appears to have been the UHC database. There were no details as to what costs were included, other than the cost of surgery. The price year was unclear.

The findings of the study were not compared with those from other studies. The issue of the generalisability of the study results to other settings was discussed in that the results were limited to academic centres and may not be generalisable to non-academic institutions. No sensitivity analysis was conducted, which tends to limit the external validity of the analysis.

The authors did not make any recommendations.

None stated.