- Email: [email protected]
Su1256 Inter-and Intrarater Reliability of the Chicago Classification in Pediatric High-Resolution Esophageal Manometry Recordings
AGA Abstracts
193 subjects (excluding 10 patients with all four low values) showed that 30% of the patients with low lactase also had low sucrase, while 58% of the patients with low sucrase also had low lactase. Chi-square analysis was performed to examine associations between groups with normal or abnormal disaccharidases and clinical features (epigastric abdominal pain, vomiting and diarrhea). There were no significant associations between the disaccharidase groups and clinical features. Conclusions: A large proportion of patients with CAP have deficiencies in lactase and sucrase. Epigastric abdominal pain, vomiting and diarrhea were no different between groups, suggesting that these clinical features cannot be used to predict deficiencies in lactase or sucrase in children with abdominal pain. Low lactose diet alone in these patients may not be sufficient to alleviate symptoms. 1. Lebenthal E, Rossi TM, Nord KS, Branski D. Pediatrics. 1981;67:828-32. Su1255 Pressure-Flow Signatures Associated With Normal and Disordered Esophageal Motor Patterns: A Pediatric Study Maartje Singendonk, Stamatiki Kritas, Charles Cock, Lara Ferris, Lisa McCall, Nathalie Rommel, Michiel P. van Wijk, Marc A. Benninga, David Moore, Taher Omari Background: The Chicago Classification allows disordered esophageal motor dysfunction to be characterised into four main categories based on esophageal pressure topography (EPT) metrics. As an adjunct to EPT, pressure flow analysis relates impedance-detected bolus movement to pressure-detected bolus propulsion. The aim of this study was to perform pressure-flow analysis in a cohort of Chicago Classified pediatric patients. We hypothesised that patients within the different Chicago Classification categories would exhibit a different pressure-flow signature. Methods: Combined high-resolution impedance-and solid state pressure recordings were performed in 74 pediatric patients referred for diagnostic manometric investigation (32M; 9.1 ± 0.7 years) and 24 healthy young adult controls (7M; 36.1 ± 2.2 years) using the Solar GI acquisition system (MMS The Netherlands). Standardised saline and viscous boluses were tested. EPT metrics were calculated and the Chicago-Classification determined for each patient using MMS analysis software (version 8.23). Pressure-flow analysis of swallows was performed using MATLAB-based software (AIMplot, T. Omari) which calculated the pressure-flow index (PFI), a composite measure of bolus pressurisation relative to flow and the impedance ratio (IR) a measure of the extent of bolus clearance failure. Results: Based on EPT metrics, patients were mostly classified as Normal (40, 54%) or with a Category 4 disorder, usually weak peristalsis (25, 34%). Three (4%) had a Category 3 disorder, five (7%) had a Category 2 disorder (EGJ outflow obstruction) and two (3%) a Category 1 disorder (Achalasia Type II). The control reference range (90th Percentile) for PFI and IR was ≤ 142 and ≤ 0.49 respectively. Patients characterised within these limits were mostly classified with Normal esophageal motility (68%). The majority of patients with pressure-flow characteristics outside these limits also had an abnormal Classification (61%). Patients with a high PFI and a disordered motor pattern all had EJG outflow obstruction. Patients with high IR and a disordered motor pattern were either achalasia, weak peristalsis or frequent failed peristalsis (Figure 1). In the cohort overall larger total peristaltic break size correlated with higher impedance ratios (r=-0.292,p=0.017) and lower PFI values (r= 0.500,p=0.000). Conclusions: Disordered esophageal motor patterns were associated with an altered pressure-flow signature. By defining the degree of over-pressurisation and/or extent of clearance failure, pressure-flow analysis may be a useful adjunct to EPT-based classification of primary esophageal motor disorders. These additional insights may potentially define the optimal treatment strategy for individual dysphagia patients.
Su1256 Inter- and Intrarater Reliability of the Chicago Classification in Pediatric HighResolution Esophageal Manometry Recordings Maartje Singendonk, Marije J. Smits, Ilja Heijting, Michiel P. van Wijk, Samuel Nurko, Rachel L. Rosen, Pim W. Weijenborg, Rammy Abu-Assi, Daniël R. Hoekman, Grace Seiboth, Marc A. Benninga, Taher Omari, Stamatiki Kritas BACKGROUND: The Chicago Classification (CC) of esophageal motility facilitates the interpretation of high resolution esophageal manometry (HRIM) recordings. Application of the CC and related software tools to the pediatric population requires validation. AIM: To assess inter- and intrarater reliability of interactive CC analysis software for diagnosis of pediatric esophageal motility disorders. METHODS: A database of 30 solid state HRIM recordings in children referred for manometry (13M; mean age at study 12.1 SD 5.1 years) was created and 10 liquid swallows per patient were analysed by 10 raters (six experts, four non-experts). Raters were blinded to clinical profile and recordings were provided in a randomized order. Analysis was performed using interactive software (MMS version 8.23) which populated each swallow tracing with analysis landmarks, which observers were required to manually adjust or remove. Swallow metrics (Integrated Relaxation Pressure (IRP4s), Distal Contractile Integral (DCI), Contractile Front Velocity (CFV), Distal Latency (DL) and Break size (BS)) as well as an overall CC diagnosis per study were automatically generated. In addition, all raters provided their CC diagnosis based on personal opinion. To study intrarater reliability, four expert raters performed a second analysis ≥ seven days after their first. To determine intra- and interrater reliability, Cohen's kappa (κ) and Fleiss' kappa (κ) were used for categorical data, and the intraclass correlation coefficient (ICC) for continuous data. Metrics were excluded from analysis if not uniformly retained by all observers. RESULTS:Overall interrater reliability for IRP4, DCI and BS was almost perfect (table 1). Reliability for CFV and DL was moderate and depended on level of experience. Interrater reliability for the software-generated CC diagnosis was substantial (κ =0.65) and moderate (κ =0.53) when based on the raters' personal opinion. Overall, raters changed the software-generated diagnoses in 10%-53% of patient studies. Agreement across all raters on the diagnosis of normal motility, EGJ outflow obstruction (IRP4-dependent) and weak peristalsis with large breaks was substantial (κ=0.77, 0.76, and 0.65 respectively). Agreement for weak peristalsis with small breaks and distal esophageal spasm (DL-dependent) was fair (κ=0.43, 0.41 respectively). Amongst the four experts, intrarater reliability for the CC diagnosis was almost perfect (κ =0.81) when software-generated and substantial (κ =0.70) when based on the raters' personal opinion. CONCLUSION: Inter- and intrarater reliability of CC-based diagnosis of pediatric HRIM recordings was high when software generated, however changes introduced by personal opinion reduced reliability. Moderate agreement for CFV and DL requires caution when diagnosing motility disorders dependent on these metrics in pediatric patients. Figure 1: All, experienced and unexperienced raters' intraclass correlation coefficient values for main Chicago Classification metrics.
(ICC < 0 = no agreement, 0-0.20 = slight, 0.21-0.40= fair, 0.41-0.60 = moderate, 0.610.80 = substantial, 0.81-1 = almost perfect agreement)
Patient Stratification based on average distal esophagus. Pressure Flow Index and Impedance Ratio measured for viscous boluses. A. Scatter plot of PFI vs. IR for individual patients grouped by the category of Chicago Classification. The square demarcates the control range for PFI/IR based on 90th percentile of values from young adult controls. Pressure-flow values outside this box were considered to be of disordered HRM pattern. Graphs B and C show the proportion of patients with findings of normal or disordered esophageal motility within and outside the range of pressure-flow values of controls. B. Patients who had pressureflow characteristics within control limits were mostly classified as having normal motility. C. Patients with abnormal clearance (high IR) and/or abnormal pressurisation (high PFI) were mostly classified with disordered motility. The relative distribution of specific motor pattern categories amongst the two pressure-flow groups was significantly different (Chi squared p<0.001).
AGA Abstracts
Su1257 Combined Multiple Intraluminal Impedance pH Monitoring Is of Great Clinical Usefulness in Children Referred for Gastroesophageal Reflux Disease Paolo Rossi, Saverio Mallardo, Sara Isoldi, Giuseppe Pagliaro, Giovanni Di Nardo, Giulia Biscione, Salvatore Oliva, Salvatore Cucchiara Diagnosis of gastroesophageal reflux (GER) disease (GERD) is usually performed by combined multiple intraluminal impedance pH monitoring (MII-pH) and endoscopy. Although guidelines of Scientific Societies are published, there is a wide variability in the approach to GER patients and often antisecretory therapy is commenced on empirical basis. We wished to
S-416
193 subjects (excluding 10 patients with all four low values) showed that 30% of the patients with low lactase also had low sucrase, while 58% of the patients with low sucrase also had low lactase. Chi-square analysis was performed to examine associations between groups with normal or abnormal disaccharidases and clinical features (epigastric abdominal pain, vomiting and diarrhea). There were no significant associations between the disaccharidase groups and clinical features. Conclusions: A large proportion of patients with CAP have deficiencies in lactase and sucrase. Epigastric abdominal pain, vomiting and diarrhea were no different between groups, suggesting that these clinical features cannot be used to predict deficiencies in lactase or sucrase in children with abdominal pain. Low lactose diet alone in these patients may not be sufficient to alleviate symptoms. 1. Lebenthal E, Rossi TM, Nord KS, Branski D. Pediatrics. 1981;67:828-32. Su1255 Pressure-Flow Signatures Associated With Normal and Disordered Esophageal Motor Patterns: A Pediatric Study Maartje Singendonk, Stamatiki Kritas, Charles Cock, Lara Ferris, Lisa McCall, Nathalie Rommel, Michiel P. van Wijk, Marc A. Benninga, David Moore, Taher Omari Background: The Chicago Classification allows disordered esophageal motor dysfunction to be characterised into four main categories based on esophageal pressure topography (EPT) metrics. As an adjunct to EPT, pressure flow analysis relates impedance-detected bolus movement to pressure-detected bolus propulsion. The aim of this study was to perform pressure-flow analysis in a cohort of Chicago Classified pediatric patients. We hypothesised that patients within the different Chicago Classification categories would exhibit a different pressure-flow signature. Methods: Combined high-resolution impedance-and solid state pressure recordings were performed in 74 pediatric patients referred for diagnostic manometric investigation (32M; 9.1 ± 0.7 years) and 24 healthy young adult controls (7M; 36.1 ± 2.2 years) using the Solar GI acquisition system (MMS The Netherlands). Standardised saline and viscous boluses were tested. EPT metrics were calculated and the Chicago-Classification determined for each patient using MMS analysis software (version 8.23). Pressure-flow analysis of swallows was performed using MATLAB-based software (AIMplot, T. Omari) which calculated the pressure-flow index (PFI), a composite measure of bolus pressurisation relative to flow and the impedance ratio (IR) a measure of the extent of bolus clearance failure. Results: Based on EPT metrics, patients were mostly classified as Normal (40, 54%) or with a Category 4 disorder, usually weak peristalsis (25, 34%). Three (4%) had a Category 3 disorder, five (7%) had a Category 2 disorder (EGJ outflow obstruction) and two (3%) a Category 1 disorder (Achalasia Type II). The control reference range (90th Percentile) for PFI and IR was ≤ 142 and ≤ 0.49 respectively. Patients characterised within these limits were mostly classified with Normal esophageal motility (68%). The majority of patients with pressure-flow characteristics outside these limits also had an abnormal Classification (61%). Patients with a high PFI and a disordered motor pattern all had EJG outflow obstruction. Patients with high IR and a disordered motor pattern were either achalasia, weak peristalsis or frequent failed peristalsis (Figure 1). In the cohort overall larger total peristaltic break size correlated with higher impedance ratios (r=-0.292,p=0.017) and lower PFI values (r= 0.500,p=0.000). Conclusions: Disordered esophageal motor patterns were associated with an altered pressure-flow signature. By defining the degree of over-pressurisation and/or extent of clearance failure, pressure-flow analysis may be a useful adjunct to EPT-based classification of primary esophageal motor disorders. These additional insights may potentially define the optimal treatment strategy for individual dysphagia patients.
Su1256 Inter- and Intrarater Reliability of the Chicago Classification in Pediatric HighResolution Esophageal Manometry Recordings Maartje Singendonk, Marije J. Smits, Ilja Heijting, Michiel P. van Wijk, Samuel Nurko, Rachel L. Rosen, Pim W. Weijenborg, Rammy Abu-Assi, Daniël R. Hoekman, Grace Seiboth, Marc A. Benninga, Taher Omari, Stamatiki Kritas BACKGROUND: The Chicago Classification (CC) of esophageal motility facilitates the interpretation of high resolution esophageal manometry (HRIM) recordings. Application of the CC and related software tools to the pediatric population requires validation. AIM: To assess inter- and intrarater reliability of interactive CC analysis software for diagnosis of pediatric esophageal motility disorders. METHODS: A database of 30 solid state HRIM recordings in children referred for manometry (13M; mean age at study 12.1 SD 5.1 years) was created and 10 liquid swallows per patient were analysed by 10 raters (six experts, four non-experts). Raters were blinded to clinical profile and recordings were provided in a randomized order. Analysis was performed using interactive software (MMS version 8.23) which populated each swallow tracing with analysis landmarks, which observers were required to manually adjust or remove. Swallow metrics (Integrated Relaxation Pressure (IRP4s), Distal Contractile Integral (DCI), Contractile Front Velocity (CFV), Distal Latency (DL) and Break size (BS)) as well as an overall CC diagnosis per study were automatically generated. In addition, all raters provided their CC diagnosis based on personal opinion. To study intrarater reliability, four expert raters performed a second analysis ≥ seven days after their first. To determine intra- and interrater reliability, Cohen's kappa (κ) and Fleiss' kappa (κ) were used for categorical data, and the intraclass correlation coefficient (ICC) for continuous data. Metrics were excluded from analysis if not uniformly retained by all observers. RESULTS:Overall interrater reliability for IRP4, DCI and BS was almost perfect (table 1). Reliability for CFV and DL was moderate and depended on level of experience. Interrater reliability for the software-generated CC diagnosis was substantial (κ =0.65) and moderate (κ =0.53) when based on the raters' personal opinion. Overall, raters changed the software-generated diagnoses in 10%-53% of patient studies. Agreement across all raters on the diagnosis of normal motility, EGJ outflow obstruction (IRP4-dependent) and weak peristalsis with large breaks was substantial (κ=0.77, 0.76, and 0.65 respectively). Agreement for weak peristalsis with small breaks and distal esophageal spasm (DL-dependent) was fair (κ=0.43, 0.41 respectively). Amongst the four experts, intrarater reliability for the CC diagnosis was almost perfect (κ =0.81) when software-generated and substantial (κ =0.70) when based on the raters' personal opinion. CONCLUSION: Inter- and intrarater reliability of CC-based diagnosis of pediatric HRIM recordings was high when software generated, however changes introduced by personal opinion reduced reliability. Moderate agreement for CFV and DL requires caution when diagnosing motility disorders dependent on these metrics in pediatric patients. Figure 1: All, experienced and unexperienced raters' intraclass correlation coefficient values for main Chicago Classification metrics.
(ICC < 0 = no agreement, 0-0.20 = slight, 0.21-0.40= fair, 0.41-0.60 = moderate, 0.610.80 = substantial, 0.81-1 = almost perfect agreement)
Patient Stratification based on average distal esophagus. Pressure Flow Index and Impedance Ratio measured for viscous boluses. A. Scatter plot of PFI vs. IR for individual patients grouped by the category of Chicago Classification. The square demarcates the control range for PFI/IR based on 90th percentile of values from young adult controls. Pressure-flow values outside this box were considered to be of disordered HRM pattern. Graphs B and C show the proportion of patients with findings of normal or disordered esophageal motility within and outside the range of pressure-flow values of controls. B. Patients who had pressureflow characteristics within control limits were mostly classified as having normal motility. C. Patients with abnormal clearance (high IR) and/or abnormal pressurisation (high PFI) were mostly classified with disordered motility. The relative distribution of specific motor pattern categories amongst the two pressure-flow groups was significantly different (Chi squared p<0.001).
AGA Abstracts
Su1257 Combined Multiple Intraluminal Impedance pH Monitoring Is of Great Clinical Usefulness in Children Referred for Gastroesophageal Reflux Disease Paolo Rossi, Saverio Mallardo, Sara Isoldi, Giuseppe Pagliaro, Giovanni Di Nardo, Giulia Biscione, Salvatore Oliva, Salvatore Cucchiara Diagnosis of gastroesophageal reflux (GER) disease (GERD) is usually performed by combined multiple intraluminal impedance pH monitoring (MII-pH) and endoscopy. Although guidelines of Scientific Societies are published, there is a wide variability in the approach to GER patients and often antisecretory therapy is commenced on empirical basis. We wished to
S-416