Frequency distribution of weight loss percentage after gastric bypass and adjustable gastric banding.

BACKGROUND

The results of surgical procedures for weight loss are often described in terms of the percentage of excess weight lost. Expressing outcomes using the mean and standard deviation might not adequately describe the clinical experience. This could in part be because the use of the mean +/- standard deviation assumes a normal or random distribution of outcomes. It has been our perception that the weight loss results after gastric bypass are relatively normally and tightly distributed around the mean, making it relatively predictable. However, we have found that the results after adjustable gastric banding are more highly variable. In fact, there appears to be 2 groups of patients after this restrictive operation. One group, that is able to work well and does not struggle much against the restriction, accepts the limits that it imposes, and another group, that does not easily learn to deal with the restriction and hence mal-adapts.

METHODS

To evaluate the validity of our clinical perception, we undertook an analysis of the distribution of weight loss by the percentiles of excess weight lost. All patients with follow-up of > or =1 years after gastric bypass or adjustable banding were evaluated for this analysis. The demographics and percentage of excess weight loss were evaluated. The distribution of the percentage of excess weight loss in 10% increments was evaluated.

RESULTS

Both groups were similar with respect to the mean patient age. However, the patients in the gastric bypass group had had a significantly greater mean preoperative body mass index and were more likely to be women. As expected, the weight loss of the gastric bypass patients fell in a normal single-peak distribution for < or =5 years of follow-up. The data from the adjustable gastric band patients at 1 year demonstrated a normal single-peak distribution, with a longer rightward tail. At 2 and 3 years postoperatively, the data from the band patients had a 2-peaked curve.

CONCLUSION

The initial weight loss results after gastric banding are less predictable than those after gastric bypass. A similar analysis of long-term outcomes might be enlightening and assist in making clinical decisions.