The Alberta population-based prospective evaluation of the quality of life outcomes and economic impact of bariatric surgery (APPLES) study: background, design and rationale

Lifestyle modification (diet, exercise ± behavioural therapy) and pharmacotherapy for obesity each reduce weight by approximately 3-5% but are limited by poor long-term effectiveness and sub-optimal adherence [12, 13]. In comparison, bariatric surgery leads to substantial weight reduction and has emerged an effective means to reduce weight and improve comorbidity in patients with extreme obesity [14]. Surgery is currently indicated in medically refractory patients with BMI levels of ≥ 40 kg/m² or BMI levels of ≥ 35 kg/m² with a major obesity-related comorbidity (e.g., hypertension, diabetes, sleep apnea) [14].

Bariatric procedures either involve stomach restriction alone or combined restriction plus intestinal diversion. The types of procedures performed have evolved over the past several decades and certain procedures, such as banded gastroplasty, have been abandoned due to poor long-term weight loss results. In Canada and globally, the most common operations performed are adjustable gastric banding (42%), Roux-en-y gastric bypass (40%) and sleeve gastrectomy (5%). Ninety percent of bariatric procedures are performed via laparoscopic (minimally invasive) techniques [15]. In gastric banding, the proximal stomach is encircled with an adjustable band that is progressively inflated to create a small, restrictive gastric pouch which reduces meal portions. In the roux-en-y gastric bypass, a highly restrictive gastric pouch is created and coupled with diversion of the upper small intestine. The sleeve gastrectomy procedure is performed by fashioning the stomach into an elongated tube and resecting the majority of the greater curve of the stomach.

Although no large scale, contemporary randomized controlled trials (RCTs) have examined the impact of surgery on cardiovascular morbidity or overall mortality, compelling data are available from high quality observational studies such as the Swedish Obesity Study (SOS), a matched cohort study of 2010 surgical and 2037 controls [16, 17]. Compared to the poor long-term results of non-surgical therapy, surgery is the only therapy associated with substantial improvements in weight (averaging 33% after 2-3 years[18] and 16% after 10 years[16]); 15-year mortality rates (5.0% versus 6.3% in well-matched controls; HR 0.71; 95% CI 0.54 to 0.92);[19] 11-year incidence rates of first time cancers (HR 0.67; 95% CI 0.53-0.85);[17] and 7-year mortality rates from coronary artery disease (HR 0.41; 95% CI 0.21-0.78) and cancer (HR 0.40; 95% CI 25-0.65) [20]. In terms of other medical comorbidity, surgery increased remission rates for type 2 diabetes (73% versus 13%; p < 0.001; OR 5.5; 95% CI 2.2-14) in a 60-patient RCT;[21] and meta-analyses of primarily observational data has demonstrated that surgery is associated with resolution or improvement of type 2 diabetes, hypertension, dyslipidemia and sleep apnea in 70-86% of cases [18]. Additional studies demonstrate that surgery significantly (p < 0.05) improves psychosocial functioning,[22] quality of life,[23] and physical function [24].

In terms of Canadian data, a retrospective analysis of 1035 bariatric surgery patients from a single practice in Quebec reported 5-year excess weight losses of 61-75% following gastric bypass and banding [25]. In an earlier study, patients from this bariatric program (n = 1035) were also retrospectively compared with 5746 age and sex-matched controls identified using administrative data claims in Quebec (clinical variables such as height and weight were not available for controls) [26]. Mortality rates over 5 years were markedly lower in the bariatric surgery cohort compared to controls (0.68% versus 6.17%; RR 0.11, 95% CI 0.04-0.27) although the design of this study cannot rule out the very real likelihood that surgical selection bias (i.e., healthier and higher socioeconomic status patients with lower likelihood of surgical complications more likely to receive surgery) explains some if not most of these findings.

The complications of surgery can be divided into peri-operative and long-term complications. The totality of data suggests that the benefits of surgery far outweigh these risks [18, 27]. Perioperative death rates are 0.1-0.5% and immediate postoperative complications (e.g., clots, cardiorespiratory events and wound infections) occur in 10% of individuals. Diversionary procedures increase the long-term risk of nutrient deficiency (up to 50% of patients) while gastric bands can slip (6% of patients) or erode (10%) necessitating re-operation [27]. Some long-term consequences of surgery such as micronutrient deficiencies (e.g., vitamin D deficiency and metabolic bone disease) are incompletely understood and require further study.

From the payor perspective and relative to the commonly cited thresholds of acceptability,[28, 29] the long-term (20 years to lifetime) cost-effectiveness of surgery compared with non-surgical management appears attractive. Incremental cost-effectiveness ratios (ICERs) range from $5000 to $35 000 per quality-adjusted-life-year (QALY) [30]. The cost-effectiveness of surgery in patients with type 2 diabetes appears dominant (provides net health benefits and cost savings) compared to non-surgical interventions [31, 32]. A recent Canadian economic evaluation performed by our group from the health care payor's perspective estimated that surgery is associated with ICERs of $8000-10 000/QALY over a lifetime horizon in Canada, with more favourable ICERs in subjects with greater obesity related comorbidity [30]. The only other additional published economic data from the Canadian perspective have been reported from a retrospective cohort study from Quebec. Surgery reduced health care utilization for a variety of disorders and was cost saving after 3.5 years compared to matched controls identified through administrative data claims [33].

However, there are limitations to the above studies. Economic studies based solely upon administrative data inputs do not include home and workforce productivity and patient borne costs. Furthermore, cost effectiveness analyses from the payor perspective do not examine such costs. Therefore, these additional data elements would provide a much more accurate picture of overall costs and benefits from a societal perspective.

Demand for bariatric surgery has increased at an exponential rate. The number of procedures performed globally has increased from 5000 in 1987-9 to 350 000 (63% in US/Canada) in 2009 [15, 34]. In Canada, the annual number of procedures performed in public health care facilities in the past decade has increased nearly 19-fold to ~1500 procedures per year [35, 36].

Despite this dramatic rise in uptake, the number of individuals potentially eligible for surgery greatly exceeds current surgical capacity. Given that the number of Canadians potentially eligible for surgery is 5.8% or 1.5 million (assuming a 2009 adult population of nearly 26 million),[37] and that about 1500 procedures are performed annually in Canada,[36] only 0.1% of potentially eligible patients are accessing surgery in this country. Therefore, actual demand may be orders of magnitude greater than current provision of surgery [38].

In Canada and elsewhere, surgery is available in both publicly and privately funded programs. Private surgery costs approximately $17 000 in Canada and is unaffordable to many [39]. However, wait times for publicly funded bariatric procedures in this country average 5 years,[40] and are similarly protracted in other public health care systems [41]. This clearly indicates a substantial demand-supply gap. A 2005 Ontario Ministry of Health report estimated demand at 3500 surgeries per year in that province, a 7-fold higher number than the 500 surgeries currently performed annually [42]. This necessitated outsourcing of procedures to the US at substantial cost;[42] and resulting in petitions from advocacy groups demanding improved access [43]. In 2009, Ontario announced $75 million funding to increase procedure numbers from 244/year to nearly 1500/year [38].

In summary, the prevalence of extreme obesity has increased dramatically and bariatric surgery is the most effective treatment available. However, access to bariatric care in Canada is severely limited and wait times are lengthy. The ramifications of protracted wait times on health and health care costs have not previously been examined. In addition, data assessing the clinical and cost effectiveness of surgery in Canada are limited and further study is needed. Specifically, a comprehensive, population-based prospective assessment of the economic consequences from a health care payor, public payor (health care + other benefits such as unemployment insurance and other transfer payments), and societal (public payor, out-of-pocket costs, home and work productivity costs) perspective has not been performed to our knowledge in Canada or elsewhere.

Lastly, much of the prospective data evaluating bariatric procedures comes from the SOS study. However, vertical banded gastroplasty, which is now outdated, comprised nearly 70% of the procedures performed in SOS. More recent studies have evaluated gastric banding and gastric bypass and have reported similar results to SOS in terms of weight reduction and improvement in obesity-related comorbidities [44, 45]. However, no prior studies comparing sleeve gastrectomy to medical management and APPLES results will help address this knowledge gap.

In this prospective cohort study, consecutive and consenting patients enrolled in the Weight Wise Regional Obesity Program and without a contraindication to surgery will be enrolled. The minimum enrolment sample size will include 150 surgical patients, 200 patients receiving intensive medical therapy and 150 patients wait-listed to enter the clinic (Figure 1). The number of patients enrolled in the medical arm is larger because we anticipate an increased rate of censoring in this study arm, as some of these patients will ultimately undergo bariatric surgery within the two-year follow-up period.

The Edmonton Weight Wise program is a comprehensive initiative established in 2005 designed to deliver integrated, patient-focused, evidence-based care to the Edmonton Zone of Alberta Health Services (AHS). This region is one of the largest integrated health delivery systems in Canada and includes a catchment population of approximately 1 million residents within greater Edmonton and an additional 600 000 residents in surrounding regions cared for by more than 1000 physicians, with an annual healthcare budget of almost two billion dollars [46].

Weight Wise includes a central, region-wide single-point-of-access referral system; community education and weight management sessions; and adult and pediatric bariatric specialty clinics. Adult specialty services are offered to patients with BMI levels of ≥ 35 kg/m² referred from a medical practitioner. By extrapolating from contemporary Canadian obesity surveillance data (i.e. ~8% of Canadians are moderately [BMI 35.0-39.9 kg/m²] or severely [> 40 kg/m²] obese), we estimate that over 125 000 adult patients within our region's catchment area has a BMI ≥ 35 kg/m² [6]. Community-dwelling patients referred for evaluation in the adult clinic are wait-listed at the time of referral (Figure 2). Currently, over 2000 adult patients are wait-listed for entry into the specialty clinic and their average wait may be up to several years in duration. Wait-listed patients are expected to attend community-based group education sessions prior to clinic entry. Otherwise, they receive no specific intervention.

Within the adult specialty clinic, patients receive approximately 24-36 weeks of intensive lifestyle counselling (diet, exercise, behavioural modification), delivered by a multidisciplinary staff (internists, dieticians, nurses, physiotherapists, and psychologists) according to current recommendations [14]. Patients are seen approximately every 4-8 weeks. Patients interested in bariatric surgery are also evaluated for this procedure by the same multidisciplinary staff. Patients deemed to be appropriate candidates are subsequently evaluated by a bariatric surgeon. Patients with BMI levels ≥ 35-39.9 kg/m² and a major medical comorbidity (e.g., hypertension, type 2 diabetes, sleep apnea) or BMI levels ≥ 40 kg/m² are considered potential candidates for surgery. Absolute contraindications to surgery include pregnancy, uncontrolled psychiatric disease, active substance abuse or smoking (patients are required to quit prior to surgery), an active eating disorder (anorexia or bulimia), and high-risk for surgery medical status (e.g. severe coronary artery disease). Because of limited data documenting the benefits of surgery in patients younger than 18 years of age and evidence for possible harm in patients over 60,[47] procedures are not performed in these age groups. In order to access surgery, patients are also required to demonstrate commitment to attend scheduled appointments and adhere to lifestyle modification and behavioural therapy.

Roux-en-Y gastric bypass, gastric banding and sleeve gastrectomy are all performed. Surgical techniques have been previously detailed [30, 48, 49]. Gastric banding is performed using the Swedish Adjustable Gastric Band (SAGB) Realize I/II(tm) (Johnson & Johnson/Ethicon Endosurgery, Cincinnati, Ohio) and sleeve gastrectomies are created over a 50 Fr bougie using staple line reinforcement throughout. Initially, gastric bypass was performed by hand sewing the gastrojejunostomy over a 34 Fr orogastric tube and positioning of the roux limb was retrocolic. Over the past two years, the technique has been modified and a 21 or 25 mm circular stapler is now used for the gastrojejunostomy pouch, with antecolic positioning of the roux limb. The roux limb is ~100 cm in length. The entero-enterostomy is created using varying techniques, including a combination of linear staplers and sutured closure.

Follow-up visits will be scheduled every 6 months for two years and will be performed in-person or by telephone if necessary. The final two year follow-up visit will be performed in-person.

After informed consent is obtained, baseline data collection for consenting patients will include the following: age, sex, race, marital status, employment status, household income quintile (≤ 20 K, 20-40 K, 40-60 K, 60-80 K, ≥80 K), general medical history and obesity-related comorbidities, smoking status (current, past, never), detailed current and past medications, weight, BMI, waist circumference, blood pressure, fasting lipids, fasting glucose, HbA1c, and liver enzymes.

Body weight will be measured using a validated, calibrated scale to the nearest 0.1 kilogram after the patient has emptied his/her bladder. Subjects will wear light indoor clothing with empty pockets and no shoes. Height will be measured using a wall-mounted stadiometre. A single reading taken using an automated blood pressure monitor and using an appropriately sized blood pressure cuff will be recorded with the subject seated in a chair and after five minutes of rest.

Repeat assessment of blood pressure, body weight, and cardiovascular risk factors (cholesterol profile, glycemic parameters) will be performed at 2 years.

Additional outcomes to be collected every 6 months will include:

Using anonymised and de-identified study identification numbers, we will link our study sample with provincial administrative databases so as to facilitate extended follow-up of health care utilization, clinical events, and costs. We have undertaken several previous studies linking clinical registry and cohort data with these high quality databases, and have previously reported greater than 96% linkage success with follow-up extending 5 years and beyond [55, 67].

APPLES has been approved by the Health Research Ethics Board of the University of Alberta.