Dataset DOI: 10.5061/dryad.cfxpnvxp2

Description of the data and file structure

Data were collected as part of a prospective clinical trial across 18 different hospitals in which clinical measurements were obtained from adult participants undergoing Barrett Esophagus surveillance endoscopies. 

Files and variables

File: ACID2924sr3.csv

Description: 

The data file ACID2924sr3.csv is a Comma-Saparated Values file which can be opened in Excel and subsequently imported in statistical software such as R. This file contains all data necessary for the various analyses described in our article 'Acetic acid chromoendoscopy for the detection of neoplastic Barrett esophagus: a stepped wedge cluster randomized controlled trial (ACID-trial)'. It contains data from 3050 individuals (rows).

Variables

The CSV file includes the following columns:

• patient: a random number given to each participant. These random identifiers cannot be used to trace the data back to individual participants.
• site: hospital site, n = 18
• diagnosis_barrett: inclusion criterium, an established diagnosis of Barrett esophagus. 1 = yes, 0 = no
• eighteen_age: inclusion criterium, age > 18 years. 1 = yes, 0 = no. 
• exclusion_no1: exclusion criterium1. 1 = yes, 0 = no
• exclusion_no2: exclusion criterium2. 1 = yes, 0 = no 
• exclusion_no3: exclusion criterium3. 1 = yes, 0 = no
• exclusion_no4: exclusion criterium4. 1 = yes, 0 = no
• exclusion_no5: exclusion criterium5. 1 = yes, 0 = no
• exclusion_no6: exclusion criterium6. 1 = yes, 0 = no
• exclusion_no7: exclusion criterium7. 1 = yes, 0 = no
• exclusion_no8: exclusion criterium8. 1 = yes, 0 = no
• exclusion_no9: exclusion criterium9. 1 = yes, 0 = no
• eligible_study: patient eligible for the trial. 1 = yes, 0 = no
• age_cat: patient's age in years, categorized. 1= age =<60, 2= age between 61-70, 3=age >= 71
• gender: patient's gender. 1 = male, 0 = female
• asa: ASA Physical Status Classification. 1 = 1, 2 = 2, 3 = 3, 4 = 4. Missing values are indicated as blank cells. 
• date_procedure: date of the surveillance endoscopy in 'dd-mm-yyyy'
• index_endoscopy: current endoscopy is endoscopy in which Barrett esophagus is newly diagnosed. 1 = yes, 0 = no
• previous_dysplasia: date of previously diagnosis dysplasia in 'dd-mm-yyyy'
• previous_pa: exact diagnosis of prevously diagnosed dysplasia. 2 = indefinite for dysplasia, 3 = low-grade dysplasia
• time_since_index: time since the diagnosis of Barrett endoscopus in months. Missing values are indicated as blank cells. 
• time_since_previous_endoscopy: time since the previous Barrett surveillance endoscopy in months. Missing values are indicated as blank cells. 
• duration_endoscopy: total procedure time in minutes. Missing values are indicated as blank cells. 
• brand_endoscopy: brand of endoscope used during surveillance endoscopy. 1 = Olympus, 2 = Fujinon, 3 = Pentax
• high_resolution: use of high-definiton endoscope during surveillance endoscopy. 1 = yes, 0 = no. Missing values are indicated as '999' 
• sedation_used: type of sedation used during surveillance endoscopy. 1 = no sedation, 2 = midazolam, 3 = propofol, 4 = general anesthesia. 
• hiatal_hernia: size hiatal hernia in centimeters
• prague_classification_c: length of circumferential Barrett extent according to prague classification in centimeters
• prague_classification_m: length of maximum Barrett extent according to prague classification in centimeters
• esophagitis_present: presence of esophageal esophagitis.1 = yes, 0 = no
• esophagitis_grade: grade of esophagitis if present, according to the Los Angeles classification. 1 = grade A, 2 = grade B, 3 = grade C, 4 = grade D, 999 = unknown/not reported. 

Columns filled in for patients receiving surveillance endoscopies without acetic acid chromoendoscopy: 

• scheduled_conventional: patient scheduled for surveillance endoscopy without acetic acid chromoendoscopy. 1 = yes, 0 = no. 

• visible_abnormality: abnormality seen in Barrett epithelium. 1 = yes, 0 = no. 

• location_abnormality: location of abnormality seen in Barrett epithelium in centimeters from the incisor teeth. Missing values are indicated as '999'. 

• clockwise_location: clockwise location of abnormality seen in Barrett epithelium. Missing values are indicated as '999' 

• suspicious_dysplasia: abnormality suspicious for dysplasia or cancer. 1 = obvious cancer (i.e. luminal stenosis), 2 = suspicious for dysplasia or cancer, 3 = not suspicious for dysplasia or cancer

• target_biopsies: targeted biopsies taken from abnormality. 1 = yes, 0 = no

• explain_no_targetbiopsies: explanation if no targeted biopsies are obtained in plain text

• number_biopsies: total number of targeted biopsies taken form abnormality in numbers. 

• pa_subnumber: pathology subnumber (i.e. number of pathology container) in roman numerals

• pa_all_assessable: all biopsies assessable. 1 = yes, 0 = no

• reason_pa_not_assessable: explanation if biopsies are not assessable in plain text

• esophagitis_pa: active esophagitis in pathological evaluation biopsies. 1 = yes, 0 = no

• grade_targetbiopsy_c: degree of dysplasia or cancer in targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• other_pa_c: other pathological result in targeted biopsies in plain text. 

• more_visible_abnormalities: more abnormalities seen in Barrett epithelium. 1 = yes, 0 = no. 

• location_abnormality_2: location of abnormality seen in Barrett epithelium in centimeters from the incisor teeth. Missing values are indicated as '999'. 

• clockwise_location_2: clockwise location of abnormality seen in Barrett epithelium. Missing values are indicated as '999' 

• suspicious_dysplasia_2: abnormality suspicious for dysplasia or cancer. 1 = obvious cancer (i.e. luminal stenosis), 2 = suspicious for dysplasia or cancer, 3 = not suspicious for dysplasia or cancer

• target_biopsies_2: targeted biopsies taken from abnormality. 1 = yes, 0 = no

• explain_no_targetbiopsies_2: explanation if no targeted biopsies are obtained in plain text. 

• number_biopsies_2: total number of targeted biopsies taken form abnormality in numbers. 

• pa_subnumber_2: pathology subnumber (i.e. number of pathology container) in roman numerals

• pa_all_assessable_2: all biopsies assessable. 1 = yes, 0 = no

• reason_pa_not_assessable_2: explanation if biopsies are not assessable in plain text

• esophagitis_pa_2: active esophagitis in pathological evaluation biopsies. 1 = yes, 0 = no

• grade_targetbiopsy_2c: degree of dysplasia or cancer in targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• other_pa_2c: other pathological result in targeted biopsies in plain text. 

• more_visible_abnormalities_2: more abnormalities seen in Barrett epithelium. 1 = yes, 0 = no. 

• location_abnormality_3: location of abnormality seen in Barrett epithelium in centimeters from the incisor teeth. Missing values are indicated as '999'. 

• clockwise_location_3: clockwise location of abnormality seen in Barrett epithelium. Missing values are indicated as '999' 

• suspicious_dysplasia_3: abnormality suspicious for dysplasia or cancer. 1 = obvious cancer (i.e. luminal stenosis), 2 = suspicious for dysplasia or cancer, 3 = not suspicious for dysplasia or cancer

• target_biopsies_3: ·targeted biopsies taken from abnormality. 1 = yes, 0 = no

• explain_no_targetbiopsies_3: explanation if no targeted biopsies are obtained in plain text

• number_biopsies_3: total number of targeted biopsies taken form abnormality in numbers. 

• pa_subnumber_3: pathology subnumber (i.e. number of pathology container) in roman numerals

• pa_all_assessable_3: all biopsies assessable. 1 = yes, 0 = no

• reason_pa_not_assessable_3: explanation if biopsies are not assessable in plain text

• esophagitis_pa_3: active esophagitis in pathological evaluation biopsies. 1 = yes, 0 = no

• grade_targetbiopsy_3c: degree of dysplasia or cancer in targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• other_pa_3c: other pathological result in targeted biopsies in plain text. 

• more_visible_abnormalities_3: visible_abnormalities: more abnormalities seen in Barrett epithelium. 1 = yes, 0 = no. 

• total_target_biopsies: total number of targeted biopsies taken during surveillance endoscopy in numbers. 

• highestgrade_targetbiopsy: highest grade of dysplasia in all targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• target_area_c: total number of target areas biopsied during surveillance endoscopy in numbers

• target_areac_positive: total number of target areas biopsied during surveillance endoscopy positive for dypslasia in numbers

• total_random_biopsies: total number of random biopsies taken during surveillance endoscopy in numbers

• number_containers: total number of pathology containers used in numbers

• total_rpbiopsies: total number of random biopsies positive for dysplasia or esophageal adenocarcinoma in numbers

• highest_degree_rdysplasia: highest degree of dysplasia in random biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• summary_cdysplasia: highest degree of dysplasia in targeted OR random biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• lans_conventional: pathological assessment confirmed/revised. 0 = no revision, 1 = revision by an expert gastrointestinal pathologist, 2 = revision by central Barrett esophageal expert histopathology panel

• parevision_initiated_c: pathological re-assessment/revision initiated by participating center. 1 = yes, 0 = no

• diagnosis_no_biopsies: diagnosis if no biopsies were obtained. 1 = no dysplasia (assumed diagnosis), 5 = esophageal adenocarcinoma (pathological confirmed during follow-up endoscopy in referral center), 6 = no Barrett esophagus (i.e. no Barrett segment seen during surveillance endoscopy)

  Columns filled in for patients receiving surveillance endoscopies with acetic acid chromoendoscopy: 

• scheduled_conventional_a: patient scheduled for surveillance endoscopy with acetic acid chromoendoscopy. 1 = yes, 0 = no. 

• visible_abnormality_a: abnormality seen in Barrett epithelium. 1 = yes, 0 = no. 

• location_abnormality_a: location of abnormality seen in Barrett epithelium in centimeters from the incisor teeth. Missing values are indicated as '999'. 

• clockwise_location_a: clockwise location of abnormality seen in Barrett epithelium. Missing values are indicated as '999' 

• prior_acetic_acid: ·abnormality visible prior to acetic acid chromoendsocopy. 1 = yes, 0 = no. 

• aac_reaction_abnormal: if abnormality was visible prior to acetic acid chromoendoscopy, was imaging with acetic acid also abnormal. 1 = yes, 2 = no, 3 = not specified. 

• specifiaction_aac: if abnormality was not visible prior to acetic acid chromoendoscopy, how was the abnormality detected with acetic acid chromoendoscopy. 1 = irregular mucosal pattern, 2 = early loss of acetowhitening, 3 = both irregular mucosal pattern and early loss of acetowhitening. Missing values are indicated with black spaces. 

• suspicious_dysplasia_a1: abnormality was visible prior to acetic acid. Suspicious for dysplasia or esophageal adenocarcinoma. 1 = yes, 2 = no, 3 = not specified

• suspicious_dysplasia_a: abnormality was not visible prior to acetic acid but acetic acid chromoendoscopy was abnormal. Suspicious for dysplasia or esophageal adenocarcinoma. 1 = yes, 2 = no, 3 = not specified

• target_biopsies_a: targeted biopsies taken from abnormality. 1 = yes, 0 = no

• explain_no_targetbiopsies_4: explanation if no targeted biopsies are obtained in plain text

• number_biopsies_a: total number of targeted biopsies taken from abnormality in numbers. 

• pa_subnumber_a: pathology subnumber (i.e. number of pathology container) in roman numerals

• pa_all_assessable_4: all biopsies assessable. 1 = yes, 0 = no

• reason_pa_not_assessable_4: ·explanation if biopsies are not assessable in plain text

• esophagitis_pa_4: active esophagitis in pathological evaluation biopsies. 1 = yes, 0 = no

• grade_targetbiopsy_a: degree of dysplasia or cancer in targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• other_pa_a: other pathological result in targeted biopsies in plain text. 

• more_visible_abnormalities_a: more abnormalities seen in Barrett epithelium. 1 = yes, 0 = no. 

• location_abnormality_2a: location of abnormality seen in Barrett epithelium in centimeters from the incisor teeth. Missing values are indicated as '999'. 

• clockwise_location_2a: clockwise location of abnormality seen in Barrett epithelium. Missing values are indicated as '999' 

• prior_acetic_acid_2: ·abnormality visible prior to acetic acid chromoendsocopy. 1 = yes, 0 = no. 

• aac_reaction_abnormal_2: if abnormality was visible prior to acetic acid chromoendoscopy, was imaging with acetic acid also abnormal. 1 = yes, 2 = no, 3 = not specified. 

• specifiaction_aac_2: if abnormality was not visible prior to acetic acid chromoendoscopy, how was the abnormality detected with acetic acid chromoendoscopy. 1 = irregular mucosal pattern, 2 = early loss of acetowhitening, 3 = both irregular mucosal pattern and early loss of acetowhitening. Missing values are indicated with black spaces. 

• suspicious_dysplasia_2a: abnormality was visible prior to acetic acid. Suspicious for dysplasia or esophageal adenocarcinoma. 1 = yes, 2 = no, 3 = not specified

• suspicious_dysplasia_a2: abnormality was not visible prior to acetic acid but acetic acid chromoendoscopy was abnormal. Suspicious for dysplasia or esophageal adenocarcinoma. 1 = yes, 2 = no, 3 = not specified

• target_biopsies_2a: targeted biopsies taken from abnormality. 1 = yes, 0 = no

• explain_no_targetbiopsies_5: explanation if no targeted biopsies are obtained in plain text

• number_biopsies_2a: total number of targeted biopsies taken form abnormality in numbers. 

• pa_subnumber_2a: pathology subnumber (i.e. number of pathology container) in roman numerals

• pa_all_assessable_5: all biopsies assessable. 1 = yes, 0 = no

• reason_pa_not_assessable_5: explanation if biopsies are not assessable in plain text

• esophagitis_pa_5: active esophagitis in pathological evaluation biopsies. 1 = yes, 0 = no

• grade_targetbiopsy_2a: degree of dysplasia or cancer in targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• other_pa_2a: other pathological result in targeted biopsies in plain text. 

• more_visible_abnormalities_2a: more abnormalities seen in Barrett epithelium. 1 = yes, 0 = no. 

• location_abnormality_3a: location of abnormality seen in Barrett epithelium in centimeters from the incisor teeth. Missing values are indicated as '999'. 

• clockwise_location_3a: clockwise location of abnormality seen in Barrett epithelium. Missing values are indicated as '999' 

• prior_acetic_acid_3: abnormality visible prior to acetic acid chromoendsocopy. 1 = yes, 0 = no. 

• aac_reaction_abnormal_3: if abnormality was visible prior to acetic acid chromoendoscopy, was imaging with acetic acid also abnormal. 1 = yes, 2 = no, 3 = not specified. 

• specifiaction_aac_3: if abnormality was not visible prior to acetic acid chromoendoscopy, how was the abnormality detected with acetic acid chromoendoscopy. 1 = irregular mucosal pattern, 2 = early loss of acetowhitening, 3 = both irregular mucosal pattern and early loss of acetowhitening. Missing values are indicated with black spaces. 

• suspicious_dysplasia_3a: abnormality was visible prior to acetic acid. Suspicious for dysplasia or esophageal adenocarcinoma. 1 = yes, 2 = no, 3 = not specified

• suspicious_dysplasia_a3: abnormality was not visible prior to acetic acid but acetic acid chromoendoscopy was abnormal. Suspicious for dysplasia or esophageal adenocarcinoma. 1 = yes, 2 = no, 3 = not specified

• target_biopsies_3a: targeted biopsies taken from abnormality. 1 = yes, 0 = no

• explain_no_targetbiopsies_6: explanation if no targeted biopsies are obtained in plain text

• number_biopsies_3a: total number of targeted biopsies taken form abnormality in numbers. 

• pa_subnumber_3a: pathology subnumber (i.e. number of pathology container) in roman numerals

• pa_all_assessable_6: all biopsies assessable. 1 = yes, 0 = no

• reason_pa_not_assessable_6: explanation if biopsies are not assessable in plain text

• esophagitis_pa_6: active esophagitis in pathological evaluation biopsies. 1 = yes, 0 = no

• grade_targetbiopsy_3a: degree of dysplasia or cancer in targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• other_pa_3a: other pathological result in targeted biopsies in plain text. 

• more_visible_abnormalities_3a: more abnormalities seen in Barrett epithelium. 1 = yes, 0 = no. 

• location_abnormality_4a: location of abnormality seen in Barrett epithelium in centimeters from the incisor teeth. Missing values are indicated as '999'. 

• clockwise_location_4a: clockwise location of abnormality seen in Barrett epithelium. Missing values are indicated as '999' 

• prior_acetic_acid_4: abnormality visible prior to acetic acid chromoendsocopy. 1 = yes, 0 = no. 

• aac_reaction_abnormal_4: if abnormality was visible prior to acetic acid chromoendoscopy, was imaging with acetic acid also abnormal. 1 = yes, 2 = no, 3 = not specified. 

• specifiaction_aac_4: if abnormality was not visible prior to acetic acid chromoendoscopy, how was the abnormality detected with acetic acid chromoendoscopy. 1 = irregular mucosal pattern, 2 = early loss of acetowhitening, 3 = both irregular mucosal pattern and early loss of acetowhitening. Missing values are indicated with black spaces. 

• suspicious_dysplasia_4a: abnormality was visible prior to acetic acid. Suspicious for dysplasia or esophageal adenocarcinoma. 1 = yes, 2 = no, 3 = not specified

• suspicious_dysplasia_a4: abnormality was not visible prior to acetic acid but acetic acid chromoendoscopy was abnormal. Suspicious for dysplasia or esophageal adenocarcinoma. 1 = yes, 2 = no, 3 = not specified

• target_biopsies_4a: targeted biopsies taken from abnormality. 1 = yes, 0 = no

• explain_no_targetbiopsies_7: explanation if no targeted biopsies are obtained in plain text

• number_biopsies_4a: total number of targeted biopsies taken form abnormality in numbers. 

• pa_subnumber_4a: pathology subnumber (i.e. number of pathology container) in roman numerals

• pa_all_assessable_7: all biopsies assessable. 1 = yes, 0 = no

• reason_pa_not_assessable_7: explanation if biopsies are not assessable in plain text

• esophagitis_pa_7: active esophagitis in pathological evaluation biopsies. 1 = yes, 0 = no

• grade_targetbiopsy_4a: degree of dysplasia or cancer in targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• other_pa_4a: other pathological result in targeted biopsies in plain text. 

• more_visible_abnormalities_4a: more abnormalities seen in Barrett epithelium. 1 = yes, 0 = no. 

• total_target_biopsies_a: total number of targeted biopsies taken during surveillance endoscopy in numbers. 

• highestgrade_targetbiopsy_2: highest grade of dysplasia in all targeted biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• target_area_a: total number of target areas biopsied during surveillance endoscopy in numbers

• target_areaa_positive: total number of target areas biopsied during surveillance endoscopy positive for dypslasia in numbers

• total_random_biopsies_a: total number of random biopsies taken during surveillance endoscopy in numbers

• pa_containers_aac: ·total number of pathology containers used in numbers

• total_rpbiopsies_2: total number of random biopsies positive for dysplasia or esophageal adenocarcinoma in numbers

• highest_degree_rdysplasia_2: highest degree of dysplasia in random biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• summary_adysplasia_2: highest degree of dysplasia in targeted OR random biopsies. 1 = no dysplasia, 2 = indefinite for dysplasia, 3 = low-grade dysplasia, 4 = high-grade dysplasia, 5 = esophageal adenocarcinoma, 6 = other

• lans_a: pathological assessment confirmed/revised. 0 = no revision, 1 = revision by an expert gastrointestinal pathologist, 2 = revision by central Barrett esophageal expert histopathology panel

• parevision_initiated_a: pathological re-assessment/revision initiated by participating center. 1 = yes, 0 = no

• diagnosis_no_biopsies_a: diagnosis if no biopsies were obtained. 1 = no dysplasia (assumed diagnosis), 5 = esophageal adenocarcinoma (pathological confirmed during follow-up endoscopy in referral center), 6 = no Barrett esophagus (i.e. no Barrett segment seen during surveillance endoscopy)

• secondreading: second review performed by an experienced pathologist in the participating hospital in case of dysplasia or esophageal adenocarcinoma. 1 = yes, 0 = no

File: ACID_full_analysis_BMJ2.R

Description: 

We have uploaded the statistical code used to perform all analyses described in our article 'Acetic acid chromoendoscopy for the detection of neoplastic Barrett esophagus: a stepped wedge cluster randomized controlled trial (ACID-trial)'. All analyses were performed in R.

Code/software

The data file can be opened and analyzed using different statistical tools and software, including R. 

We performed all our analyses in R version 3.5.1 for Windows. The loaded packages that we used to run the analyses can be found in our R-script: 'ACID_full_analysis_BMJ2.R'

The dataset can be loaded using the following code in R.:

> setwd("xxxx") # 'xxxx' location were the file is saved

> df <- read.csv("ACID2924sr3.csv",header6 =RUE, sep = ";" ) 

Human subjects data

The dataset contains coded data derived from human participants. Before deposition in Dryad, all direct identifiers were removed, and each record was assigned a randomly generated identification number. The published dataset is fully anonymized, and the random identifiers cannot be used to trace the data back to individual participants. Data sharing is consistent with the informed consent provided by participants and with approval from the relevant Medical Ethics Review Committee.