Multidisciplinary surgical management of patients with clinically severe obesity in a publicly funded bariatric surgery service in three public hospitals in Australia

Background: Obesity, a National Health Priority Area, has been a major challenge facing the Australian population and is at a crisis level. It has also become a global epidemic with a disproportionate rise in class III obesity (BMI ≥40 Kg/m2), causing substantial burden in obesity and the healthcare systems. There are approximately one million adults in Australia with clinically severe obesity, defined as class III obesity alone or a BMI ≥35 Kg/m2 with at least one major obesity-related comorbidity. Bariatric surgery is well-established as the most effective treatment for severe and complex obesity that has been unmanageable by other modalities. Despite most Australians relying on the public healthcare system, the vast majority of bariatric surgical procedures are performed in private hospitals owing to scarce resources in the public sector. As a result of this reason and sparse research funding, there is a limited understanding of the management, benefits and safety of publicly funded bariatric surgery in the context of clinically severe obesity and its metabolic consequences, especially in long-term (defined by >5 years). Settings: A multidisciplinary publicly funded bariatric surgery service covering three public hospitals, namely Royal Prince Alfred Hospital, Concord Repatriation General Hospital and Camden Hospital in Sydney, New South Wales (NSW), Australia. Data linkage with the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) was also performed. Aims and research themes: This thesis aims to address key knowledge deficiencies in a broad range of areas encompassing the full spectrum of bariatric surgery in multidisciplinary clinical obesity services, using both retrospective and prospective study designs. These studies were grouped into four major research themes, with each theme representing one chapter in the thesis. The research themes investigated were: (1) long-term health outcomes after bariatric surgery, (2) adherence to multidisciplinary post-operative follow-up care, (3) whether pre-operative weight loss predicts post-surgical weight loss, and (4) prediction of diabetes remission using an algorithm. Main findings: The first research theme (CHAPTER 2) is a retrospective cohort study that examined the long-term effectiveness and safety of bariatric surgery in a highly-complex clinically severe obese population. While the overall weight loss and changes of obesity-related comorbidities [especially type 2 diabetes mellitus (T2DM), hypertension and hyperlipidaemia] were significant following bariatric surgery over 6 years; being super obese (BMI ≥50 Kg/m2) had no multiplicative and detrimental effect in these features compared to the morbidly obese patients (<50 Kg/m2). The prevalence of the obesity-related comorbidities studied [i.e. T2DM, hypertension, osteoarthritis (OA) and/or weight-bearing joint pain (WBJP), sleep-disordered breathing and hyperuricaemia] decreased after bariatric surgery and remained lower than baseline as time progressed; whereas hyperlipidaemia and mental illness surpassed baseline level at post-operative 6 years. For nutrient deficiency, vitamin D deficiency was noted in one-third of patients pre-operatively and decreased significantly after bariatric surgery. Iron deficiency anaemia doubled at year 6 post-operation. Low prevalence of vitamin B12 insufficiency was detected before and after bariatric surgery, with no patient developing a deficiency in years 5 and 6 post-operatively. Of the 34.5% with peri- and post-operative complications, none was life-threatening. The second research theme (CHAPTER 3) is a prospectively conducted study that investigated the reasons for ceasing attendance at clinic reviews after surgery, predictors of adherence to post-bariatric surgery clinic reviews, and the relationship – if any – between adherence to follow-up and weight loss outcomes. The adherence rate to follow-up visits after bariatric surgery (i.e. patients attended follow-up regularly) among the study patients was 63.7% [107 of 168 (63.7%)]; 20 (11.9%) attended irregularly and 41 (24.4%) ceased attending reviews. According to the patients, withdrawal from the publicly funded bariatric surgery service was mainly associated with travel distance. Linear mixed-effects model with random effects revealed no pronounced difference in the mean weight loss between the adherent and nonadherent groups (composite of those who attended irregularly or who ceased attending follow-up) over the years. Logistic regression model shows that older and partnered patients were more likely to adhere to follow-up care after operation. The results could guide care practices for patients needing additional contacts and supports, so they may benefit from additional assistance to experience optimal outcomes. Subsequent studies (CHAPTER 4) explored the relationship between pre-operative weight loss and weight loss post-bariatric surgery, to determine the necessity of the current requirement for weight loss prior to publicly funded bariatric surgery in the future. The result shows insignificant relationship between these two parameters, suggesting that the pre-operative weight management program (WMP) that is currently-mandated prior to the surgery may not be necessary. However, the WMP might still be important as an opportunity to resolve medical problems; to prevent post-operative psychological issues from emerging; to ensure patients understand the implications of bariatric surgery and its necessary lifestyle changes; and to minimise potential surgical risks. Multiple linear regression analysis demonstrated that age at surgery is a reliable predictor of post-operative weight loss across years 1 to 6 post-operation, suggesting that older patients may achieve better outcomes from bariatric surgery. The fourth research theme (CHAPTER 5) identified short- and longer-term post-operative diabetes remission prediction following bariatric surgery using the DiaRem scoring system that computed by the following simple variables – age, HbA1c, glucose lowering treatment other than insulin, and insulin treatment. The DiaRem algorithm was shown to perform well in discriminative capacity, predictive ability and calibration in the study cohort with T2DM. The area under the curve (AUC) of the receiver operator characteristic was 0.869 (95% CI=0.800–0.938) for the standard year 1 post-operative follow-up, with the most optimal cut-off score being ≤12, sensitivity of 94.8% and specificity of 64.2%. The discriminative ability is comparatively high for those with 5-year post-operative diabetes remission [AUC=0.835 (95% CI=0.714–0.956), optimal cut-off score ≤12, sensitivity=89.5% and specificity=52.2%]. Logistic regression models demonstrated that DiaRem algorithm reliably predicted diabetes remission 1 year [OR (95% CI)=0.733 (0.655-0.821), p