Impact of Patient Attrition from Bariatric Surgery Practice on Clinical Outcomes

Obesity has become a global epidemic. Bariatric surgery remains the most successful modality for producing sustained weight loss. Attrition rates after bariatric surgery are currently reported between 3 and 63% depending on the type of bariatric operation and the length of follow-up provided by the bariatric surgery team. It is currently unknown if patient attrition from bariatric surgery programs impact clinical outcomes. The availability of the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) and the Mayo Clinic Midwest unified electronic medical record (EMR) provide a unique opportunity to explore this topic.

Raw data was downloaded from MBSAQIP database for all laparoscopic Roux-en-Y gastric bypass (LRYGB) cases between May 1, 2008 and January 8, 2015 (N = 1242). Baseline weight and preexisting comorbidities (type 2 diabetes, hypertension, and hyperlipidemia) were recorded using the MBSAQIP database and the EMR. Current weight and comorbidity data (type 2 diabetes, hypertension, and hyperlipidemia) were subsequently collected at the time closest to the patient's surgical anniversary dates during the following assessment periods: years 1, 2, and 3 after surgery. Mean percentage total weight loss (TWL) was calculated at each time frame for each patient. Data was summarized using descriptive statistics, including counts and percentages for categorical variables by either year or year and location.

The number of patients seen by the bariatric surgery practice (BSP) compared to those seen by other providers within our practice (attrition to bariatric surgery practice (ABSP)) was highest in year 1 and reduced each subsequent year (year 1: BSP N = 740, ABSP N = 166, year 2: BSP N = 425, ABSP N = 309, and year 3: BSP N = 235, ABSP N = 325). The mean TWL in the BSP and ABSP groups at year 1 was 31.84% versus 30.19%, at year 2 was 31.34% versus 29.67%, and at year 3 was 29.01% versus 27.71% respectively. Differences were statistically significant between groups at years 1 and 2 (p < 0.05). A statistically significant difference was not found in year 3 despite a trend towards higher TWL in the BSP group. Among all patients, statistically significant differences (p < 0.0001) were observed between baseline and years 1, 2, and 3 for type 2 diabetes, hypertension, and hyperlipidemia.

Our study confirms the problem of patient attrition to follow up at our BSP. It is also the first to suggest a difference in weight loss outcomes among patients seen in BSP compared to those seen by other providers (ABSP) at 1 and 2 years after Roux-en-Y gastric bypass (LRYGB). The potential implications of this observation on long-term weight maintenance after LRYGB and the impact on metabolic comorbidities remain unknown but warrant further investigation. It also warrants the development of strategies to improve patient retention in BSP and/or engagement of medical providers to achieve that end.