Non-parametric statistical tests for independent groups: numerical dataBMJ 2012; 344 doi: https://doi.org/10.1136/bmj.e3354 (Published 16 May 2012) Cite this as: BMJ 2012;344:e3354
- Philip Sedgwick, senior lecturer in medical statistics
- 1Centre for Medical and Healthcare Education, St George’s, University of London, Tooting, London, UK
Researchers investigated recent trends in the use of bariatric surgery in England. In particular, they examined the surgical techniques used and factors influencing postoperative outcomes. A population cohort study was performed. All adult NHS patients with a primary diagnosis of obesity and who had undergone a primary elective bariatric procedure (gastric bypass, gastric banding, or sleeve gastrectomy) in England between April 2000 and March 2008 were included. The main outcome measures included mortality at 30 days and at one year after surgery, unplanned readmission to hospital within 28 days, and duration of stay in hospital.1
In total, 3191 gastric bypass, 3649 gastric band, and 113 sleeve gastrectomy procedures were performed. A significant difference was reported between surgical procedures in the median length of stay (gastric bypass: 5 days (interquartile range 3-7 days); gastric banding: 2 (1-3); sleeve gastrectomy 4 (3-7); P<0.001). A non-parametric test was used to compare groups in the length of hospital stay, as distributional assumptions could not be made.
The researchers reported that the number of bariatric surgical procedures had risen in England in recent years. Gastric banding and gastric bypass were the commonest procedures, while sleeve gastrectomy was first recorded in 2006. Patients selected for gastric banding …