Background  

Following laparoscopic adjustable gastric banding (LAGB), patients usually undergo follow-up (FU) to optimize weight loss and detect complications, with band-volume adjustment performed either under radiological or clinical guidance with no current consensus on what is the best standard of care. We, therefore, analyzed our patient cohort to identify differences between the two, if any, over a 3-year period.

Methods  

We retrospectively reviewed our obesity surgery database to find all LAGB patients and grouped them based on method of FU without weight exclusions. We then selected out 70 consecutive patients from each cohort from an arbitrary time-point to achieve sufficient FU and analyzed the results from the data collated. Patients with postoperative complications were excluded to prevent bias.

Results  

From 2003 to 2007, there were 865 LAGB performed in our unit. We identified 70 consecutive patients from January 2004 from each cohort. After review and exclusions, we were left with 50 patients in the radiology group (RG) and 49 in the clinical group (CG) [median BMI 43.8 and 47.1, respectively; median age 43 years in both]. Routine FU was at weekly, then fortnightly, and monthly intervals, but results were analyzed at 3-month intervals until the first year and every 12 months until the third year. The median percentage of excessive weight loss was 22% and 36% at 6 months; 28% and 43% at 12 months; 27% and 47% at 2 years; and 33% and 46% at 3 years in the RG and CG, respectively. There was no difference in difficulties to band fill in either group as per clinical records, although there was a greater incidence of port damage in the CG.

Conclusions  

Our study suggests that at medium-term follow-up, clinical fill is superior to radiological FU at least in terms of weight loss, with the added benefit of avoiding unnecessary radiation albeit that the difference between the two methods gets smaller with FU beyond 2 years. This topic merits a future randomized control trial to make recommendations without biases inherent to retrospective analysis.