This is nationwide registry-based randomised clinical multicenter trial in which patients will be randomised to gastric bypass (RYGB) or sleeve gastrectomy (SG). The co-primary endpoint are weight control over 5 years and the amount of severe adverse events. Additionally the investigators have predefined a number of secondary endpoints, and the trial has a sufficient number of patients to allow comparisons across subgroups.

Follow-up of the patients will use the routines respectively for regular follow-up on the Scandinavian Obesity Surgery Registry (SOReg) after 6v, 1 year, 2 years, and after 5 years.

In order to evaluate if SG has advantages compared to the previous standard, the investigators want to examine whether SG operations are equivalent (non-inferiority) for weight loss and weight stability five years after surgery in comparison to RYGB, and if SG is associated with fewer long-term complications (superiority). The primary outcome measure emanates from assessment of long term weight management and the frequency of serious complications.

The unforeseen global Covid-19 pandemic resulted in that almost all elective benign surgery in Scandinavia was cancelled from March 2020. Thus, the pandemic had severe consequences on the recruitment to the BEST trial during 2020-2021.

During autumn of 2021 the BEST steering committee decided to perform an additional analysis of the power for primary endpoints.

Additional information (Courcoulas et al, JAMA Surg 2020 March; Howard et al, JAMA Surg 2021 Dec) revealed that the risk of any of the predefined substantial adverse events after bariatric surgery is higher than previously anticipated in the revised power calculation, i.e. >25% instead of 13%. These figures were confirmed in an analysis of real-world data from the bariatric national quality register SOReg in Sweden which registered all patients undergoing sleeve or bypass in Sweden since 2007.

An independent statistician performed the analysis based on information above, but also on 2-year data in BEST. In conclusion, it was stated:

Two post hoc power analyses were conducted based on the data from February 2022:

1. Weight reduction. In the protocol the following is stated "This sample size will also have >95% power to evaluate non-inferiority of 5% weight loss difference over 5 years between the two groups, assuming 15 kg standard deviation in weight loss over follow-up with two-sided 2.5% significance level.". The post hoc power calculation is based on the two-year follow up data where an average weight loss for all patients (both groups, N=1031) of 29.3 kg was attained with a Sd=21.6. Given a loss to follow up of 20% from 2 year follow up to the 5-year follow up the sample size is assumed to be (1031*0.80)/2=413 patients per group. With a non-inferiority level of 5 kg weight loss the power is found to be 90% if n=393 per group, and 95% if n=486 per group (https://www.sealedenvelope.com/power/continuous-noninferior/).
2. Substantial adverse events. This post hoc power calculation is for a superiority test where the rate is assumed to be 25% for the gastric bypass at 5 years and sleeve would have a 35% lower level, i e 25%*0.65=16.25%. Given a sample of n=413 per group the post hoc power will then be 87.5%.

Taking information above into account and in the interest of not prolonging inclusion period unnecessarily the trial steering committee took a decision to stop inclusion in BEST during spring 2022 (final date 31st of March). At termination of inclusion the number of participants that had been included and operated in BEST were 1752. The trial Data Safety and Monitoring Committee reviewed and supported the decision before termination of recruitment to the BEST trial.

Inclusion Criteria:

• BMI 35-50 kg/m2
• Ability to understand and decide on the merits of the study participation
• Accepted for bariatric surgery
• Must understand the information, and be able to make a decisions about participation in the study

Exclusion Criteria:

• Previous bariatric surgery, anti reflux surgery or other gastric surgery
• Moderate to severe reflux disease, Barretts oesophagus or known hiatus hernia >4 cm
• Unstable mental illness or other known contraindication to bariatric surgery.
• Planned significant surgery at the same time
• Inflammatory bowel disease
• ongoing drug or substance abuse
• not appropriate to randomise the patient, according to surgeon