The bias risk assessment model proposed by the Cochrane shows up well-adjusted when applied to clinical trials of psychological treatments for bulimia and binge?

• Grant number: 16/12703-4
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): September 01, 2016
• Effective date (End): August 31, 2017
• Field of knowledge: Health Sciences - Medicine - Psychiatry
• Principal Investigator: Hugo Cogo Moreira
• Grantee: Juliana Pagotto Trevizo
• Host Institution: Escola Paulista de Medicina (EPM). Universidade Federal de São Paulo (UNIFESP). Campus São Paulo. São Paulo , SP, Brazil

Abstract

Eating disorders encompass a prevalence of 2% of high school students. Despite a relatively low number, these disorders are increasingly concerned since the mortality is 21% in Brazil. These disorders involve, besides the issue of food intake, concern for the image, which generates an alert regarding the health of individuals suffering from this disease. According to the DSM-5, currently, the disorders that draw attention are bulimia nervosa, anorexia nervosa, and "binge eating". Much is already know about the damage that eating disorders can have on the health of their victims, but there is little evidence about its etiology or interventions confirmed as effective. A reputable treatment would be psychotherapy, specifically cognitive behavioral therapy. Intervention studies based on randomized clinical trials allow reviewing of the effectiveness/efficiency of a given intervention for a set of outcome measures. From this context, it is worth mentioning the importance of CONSORT, which came up with the aim of guiding the methodological structure of clinical trials through 25 items, and the Handbook of Cochrane, which deals with the risks of bias in the context of systematic reviews, and assessing the congregant's clinical trials on four criteria. The assessment of the bias of risks as a section of systematic review allows a better understanding of the validity of the analysis, interpretation, and conclusions from the results. Objective: based on the clinical trials on the theme, seeks to: identify which of the four indicators used by Cochrane has a higher load factor; produce a bias risk histogram and qualify it in high or low; observe, by correlating the latent variable "risk of bias" with the observed variable "year of publication", if biases risks were minimized with the implementation of measures over the years. Thus, evaluating the possibility of generating a factor score of the "risk of bias" of these studies (one-dimensional indicator). Methods: Confirmatory factor analysis, is a structural equation modeling, a statistical technique that allows us to understand and model latent phenomena (in this case "risk of bias") underlying a set of variables manifest (in the case "year of publication" of the 48 studies involved in the systematic review "Psychological treatments for bulimia nervosa and binging"). This technical tests model, and if the model proves adequate, allows us to create a measure that creates a risk of bias composite (score). Expected results: Provide evidence of construct validity (factor) on the biases of risk indicators used in clinical trials related to psychological treatments for bulimia and "binge" to then conclude whether these indicators are evaluating well or not trace latent risk of bias in clinical trials. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

TREVIZO, JULIANA PAGOTTO; HAY, PHILLIPA; SWARDFAGER, WALTER; COGO-MOREIRA, HUGO. Risk of bias in randomized controlled trials of psychological treatments for bulimia nervosa and binge eating. ANNALS OF EPIDEMIOLOGY, v. 28, n. 9, p. 625-628, SEP 2018. (16/12703-4)