Design and Validation of a Preoperative Calculator for "Textbook Outcome" After Bariatric Surgery (BARCINO)

Morbid obesity is considered the great epidemic of our century. It is regarded as the first non-traumatic cause of death in the western population and it is also progressively beginning to affect developing countries. The latest epidemiological studies estimate that in Spain more than 15% of the population has overweight and about 5% has morbid obesity.

Obesity treatment is multidisciplinary, from lifestyle and dietary changes to surgery. Amongst the different available treatments, bariatric surgery is the only cost-effective in patients with morbid obesity. Bariatric surgery provides better results compared to both pharmacological treatments (which are less available and less effective) and lifestyle changes. Furthermore, as well as weight loss, it guarantees a better control of comorbidities, reducing the risk of cardiovascular disease, morbidity and mortality, improving quality of life. There are several currently standardized bariatric surgical techniques. These techniques can be restrictive (such as sleeve gastrectomy or SG), hypoabsortive (such as duodenal switch or DS and single anastomosis duodeno-ileal bypass with sleeve gastrectomy or SADI-S) or mixed (such as gastric bypass or GBP). The indication will depend on factors such as the patient's BMI, their metabolic comorbidities or the experience of the surgical team, among others. All of them have been shown to be safe in experienced teams, with rates of mortality equal to or less than 0.5% and morbidity less than 10%.

However, although postoperative risks are low, they have to be taken into account in each case individually, along with the potential benefits of surgery as accurately as possible, so the decision whether or not to perform surgical procedure in a consensual manner with the patient can be based on objective data.

Recently, a new concept of measuring postoperative results called "Textbook outcome" (TO) has been introduced. It is a multidimensional measure that is obtained from the sum of several traditional surgical measures: the absence of postoperative complications, no prolongation of hospital stay, no mortality, and no readmissions. The TO reflects what is considered the "ideal" postoperative evolution. In bariatric surgery, TO has been defined as hospital stay equal to or less than 2 days, absence of complications in the first 30 days after surgery, no readmission once discharged in the first 30 days after surgery, and no mortality during the stay hospitalization or during the first 30 days after surgery. Additionally, an ordered TO has been created, ordering these quality indicators from worst to best: mortality, severe complications, readmission, mild complications and prolonged hospitalization stay. The fact of ordering the results with this ordered TO allows to find different combinations of these five indicators in the same patient. It does not happen if, as previously, the combination of all the results of the different parameters generate a binomial result. The ordered TO also has the advantage that allows to identify in the results some individual differences for each hospital, thus making it possible to detect variations between hospitals that can help to improve the quality of surgery and of the hospital in particular, and therefore, improve the clinical result of bariatric surgery, which is not possible with the binomial result since it does not provide information to individual hospitals on which aspect is significantly worse than in the other hospitals. The order of parameters mentioned above was based on expert advice and evidence from the literature regarding what patients considered to be a quality improvement assistant care.

Currently, there is not any tool that calculates preoperatively the risk of not presenting a TO in some of its degrees.

The optimal values of the preoperative variables that allow patients to achieve TO have not been studied either. The aim of this study is to create a calculator that allows calculating the risk that each patient has of not achieving TO based on the patient's variables and the selected surgical technique. This tool could provide relevant information for the patient in the process of informed consent for the surgery, as well as assist in the choice of surgical technique in clinical practice in each specific case. The development of the calculator is intended to be done with the help of machine learning. The Machine learning statistics techniques are a branch of artificial intelligence that allow great flexibility to capture complex non-linear relationships especially when using large amounts of data. The situation where we find ourselves due to the rapid expansion of the COVID-19 pandemic has found doctors and the rest of health professionals unprotected to face with the difficulty of giving an adequate and rapid response to the patients.