Dr L Gerson and colleagues used logistic regression analysis on data collected from 517 GERD patients to create a prediction model for Barrett's esophagus.
99 (19%) of the 517 GERD patients had Barrett's esophagus. The research team estimated the predictive value for Barrett's of patient gender, age, and ethnicity.
The importance of the presence of symptoms including heartburn, nocturnal pain, dysphagia, odynophagia, belching, and nausea was also evaluated.
Male gender, heartburn, nocturnal pain, and odynophagia were found to have a positive predictive value for Barrett's esophagus.
|Positive predictors of Barrett's esophagus:|
- Male gender
- Nocturnal pain
|American Journal of Gastroenterology|
Dysphagia had a significant negative predictive value for Barrett's. Other factors were found to be insignificant.
The research team designed a questionnaire to aid the diagnosis of Barrett's using the factors found to have a predictive value for the disease.
The questionnaire consists of seven simple questions. Patients are assigned a score, according to their answers to the questionnaire.
The mean score for patients with Barrett's was 397.4 compared to a mean score of 351.3 for GERD patients without Barrett's.
If screening for Barrett's is performed at a score of 375 or above, the questionnaire would have a specificity of 63% and a sensitivity of 77%.
Dr Gerson concludes, "By asking seven questions about symptom severity, clinicians may be able to predict the presence of Barrett's esophagus."
"This could help to determine the need for endoscopy in GERD patients."