Timed barium swallow for assessing long-term treatment response in patients with achalasia: Absolute cutoff versus percent change - A cross-sectional analytic study

Background Timed barium swallow (TBS) assesses esophageal emptying before and after therapy in patients with achalasia. Our aim was to compare the accuracy of percent change in barium height with traditional absolute cutoff of <5 cm on post-treatment TBS. Materials and Methods Consecutive patients with treatment naive achalasia treated with either PD, HM, or POEM between 1/2012 and 7/2017 were eligible for inclusion. The accuracy of percent change in pre- and post-treatment barium height at 5 minutes versus an absolute <5 cm cutoff for assessing treatment response was assessed using the receiver operating curve analysis (ROC). Results Eighty-one patients met the inclusion criteria. The median percent change in barium heights at five minutes in patients who did not improve was 6 percent increase (n = 10; mean 10.6) versus 78 percent decrease (n = 71; mean 64) in patients who improved (P = 0.0001). The AUC for percent change in TBS 5 minutes height was 76% (95% CI 48% to 90%), and a 3% decrease from baseline as a cutoff had a sensitivity of 60% and specificity of 99%. The AUC for post-treatment TBS 5 minutes height was 79% (95% CI 53% to 91%), and the 5 cm cutoff had a sensitivity of 70% and specificity of 75%. Conclusions The results show that 3% percent improvement in pre- and post-treatment barium height at 5 minutes rather than absolute cutoff value of <5 cm on post-treatment TBS is a better indicator of treatment success in achalasia patients. These findings indicate the need for reassessment of tools to identify treatment response.

Automated summary

The study found that the percent change in barium height before and after treatment at 5 minutes is a better indicator of treatment success in achalasia patients compared to the traditional absolute cutoff value. Reassessment of tools to identify treatment response is necessary based on these findings.