Article Text
PDF
Abstract
Objective Acid exposure time (AET) from ambulatory pH studies and reflux oesophagitis are independent measurements used by the Lyon classification to diagnose GORD. This study aimed to validate AET reference ranges and diagnostic thresholds by analysis of 96-hour wireless pH studies from healthy, asymptomatic controls (HCs) and patients with and without oesophagitis.
Design HC and consecutive patients referred for wireless pH studies (off acid suppressants for >7 days) underwent 96-hour pH studies at two tertiary referral centres. Erosive oesophagitis was categorised by the Los Angeles (LA) classification. Linear regression and receiver operating curve (ROC) analysis were performed to define optimal diagnostic cut-offs.
Results Prolonged, 96-hour pH studies were completed in 39 HCs (age 28 (18–53) years, 72% female) and 944 patients (age 46 (16–85) years, 65% female), of whom 136 (14.5%) had reflux oesophagitis. Median AET in HC was 1.3% (upper 95th percentile 4.6%) for any study day and 2.6% (upper 95th percentile 6.9%) for the worst day (24-hour period) during the study. ROC analysis for average AET differentiated HC from patients with moderate-to-severe oesophagitis (LA BCD; sensitivity 87%, specificity 95%, positive predictive value (PPV) 59%, negative predictive value 99% for a cut-off AET of 4.3%; area under the receiver operating curve 0.95). Specificity was higher, but PPV was substantially lower for severe oesophagitis (LA CD). ‘Worst-day’ analysis provided similar results; however, day-to-day variability was high.
Conclusion Diagnostic thresholds for average AET were identified that accurately discriminate between HCs and patients with erosive oesophagitis. The findings provide conditional support for diagnostic criteria for GORD proposed by the Lyon Consensus.
- pH monitoring
- erosive oesophagitis
- gastro-oesophageal reflux disease
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. Available on reasonable request.
Statistics from Altmetric.com
Significance of this study
What is already known about this subject?
Prolonged, ambulatory wireless pH studies improve diagnostic yield and reliability of reflux studies in clinical practice; however, reference ranges and diagnostic thresholds for this methodology have not been established.
The Lyon classification defines acid exposure time (AET) >4% as conditional and >6% as conclusive evidence for the presence of GORD, but these thresholds have not been prospectively validated against independent markers of disease.
What are the new findings?
Diagnostic thresholds for AET were identified that accurately discriminate between healthy controls and patients with moderate to severe erosive oesophagitis. The results provide conditional support for diagnostic criteria proposed by the Lyon Consensus.
How might it impact on clinical practice in the foreseeable future?
The presence of moderate (Los Angeles (LA) grade B) as well as severe oesophagitis (LA grade CD) supports the diagnosis of GORD in patients referred for investigation of reflux symptoms.
The results confirm that average AET <4% represents normal (physiological) acid exposure, whereas AET >7%, 1% higher than the current threshold, provides conclusive evidence of GORD.
Introduction
GORD is present when the reflux of gastric content into the oesophagus causes troublesome symptoms and/or complications.1 Typical symptoms of GORD including heartburn and regurgitation are common, with an estimated worldwide prevalence between 4% and 14% with important variation in different regions,2 and up to 30% of Americans complaining of at least occasional reflux symptoms in the previous week.3
Empirical treatment with proton pump inhibitors (PPIs) is appropriate in patients with typical symptoms; however, the response to PPIs is neither sensitive or specific for GORD diagnosis.4 Endoscopic investigation is performed in patients with persistent reflux symptoms; however, less than half of patients have conclusive evidence of GORD based on the presence of severe erosive oesophagitis or Barrett’s oesophagus.5 6 Ambulatory oesophageal pH studies provide objective evidence of GORD and are particularly indicated in patients in whom the diagnosis of GORD is uncertain and in those under consideration for antireflux surgery.7 The primary outcome measure from this investigation is acid exposure time (AET), defined as the percentage of total recording time that distal oesophageal pH remains less than 4. AET is the best validated and most reproducible measurement acquired by this technique.8 Seminal studies have shown that high levels of AET are associated with increased risk of mucosal damage in the distal oesophagus.9 10 Moreover, a GORD diagnosis based on pathological AET predicts response to medical and surgical therapy.11 12 Wireless pH monitoring is an alternative to investigation performed using a naso-oesophageal pH catheter.7 13 Prolonged measurement with this technique addresses limitations of catheter-based studies related to patient discomfort and reduces the risk of false-negative results that can occur with relatively short, 24-hour measurement period due to high day-to-day variability.14–17 Studies suggest that extending the duration of pH measurements from 24 to 48 hours and up to 96 hours improves reproducibility of GORD diagnoses and provides clinical gains over shorter periods of recording, especially when results from the first 1–2 days are discordant or borderline.16–18
The primary aim of this study was to define reference intervals (normal values) and diagnostic thresholds for AET by analysis of prolonged, 96-hour wireless pH recordings from healthy, asymptomatic controls (HC) and symptomatic patients with and without oesophagitis. Based on this analysis, a prospective validation of diagnostic thresholds for GORD established by the recently published Lyon classification was performed. This system defines AET<4% as normal (physiological) and AET>6% as providing conclusive evidence for the presence of GORD.19 However, these AET thresholds have not been validated against independent markers of this GORD, in particular the presence of reflux oesophagitis.
Methods
Healthy controls
Healthy adults (≥18 years) were recruited by advertisements at the Nottingham University Hospitals NHS Trust, and Guy’s and St Thomas’ NHS Foundation Trust (GSTT), London, UK. Demographic information including age, gender and body mass index (BMI) was collected. Candidates completed validated questionnaires, including the Reflux Disease Questionnaire (RDQ).20 Eligible participants were healthy adults with no more than mild symptoms less than once a week. Subjects with contraindications for wireless pH recording (eg, bleeding diatheses, oesophageal strictures), operations on the upper GI tract, erosive oesophagitis, antireflux treatments either prescribed or over-the-counter, BMI>35 kg/m2, or major comorbidity were excluded.
Symptomatic patients
Consecutive patients (16–85 years) that completed wireless pH studies off antireflux medication at GSTT (March 2008–November 2019) were analysed. All patients had indications for wireless pH studies as set out in published guidelines.19 21 22 Exclusions included endoscopic disease other than erosive oesophagitis (eg, Barrett’s oesophagus, peptic stricture), early detachment or incomplete recording (see Oesophageal acid exposure). PPIs were discontinued at least 7 days before the study. The British Society of Gastroenterology guidelines were followed in patients on antiplatelets or anticoagulants.23 Endoscopy and physiological data were extracted using commercial software and added to a prospectively maintained database at the Oesophageal Physiology Laboratory GSTT.
Study procedures
Standard procedures for the Bravo system (Medtronic, Mansfield, MA, USA) were followed.13 Participants provided written, informed consent and permission for data analysis before enrolment. Endoscopy was performed under conscious sedation. Reflux oesophagitis was classified by the Los Angeles (LA) classification.24 The pH probe was positioned at 6 cm above squamo-columnar junction after maintaining a stable suction pressure for 1 min. Photographic documentation confirmed capsule attachment. Participants were encouraged to engage in their usual activities with unrestricted diet except for alcohol and acidic drinks. Participants kept the recorder within one metre day and night and documented reflux symptoms, supine periods, oral intake using the data logger (atypical symptoms were noted in a paper diary). Prior to the availability of pH recorders capable of recording for 96 hours, subjects attended after 48 hours for interim data upload, capsule position and function check with battery replacement for further 48-hour monitoring.
Oesophageal acid exposure
Data from pH recordings and diaries were analysed by AccuView pH, V.5.2 software. Periods when the recording was lost as a result of a failure of reception or other technical faults were excluded from the analysis. A minimum of 92-hour data was required for inclusion in the analysis. Meal periods were removed to improve the separation of normal and abnormal oesophageal acid exposure.25 Reflux was defined by pH<4 and AET was documented. Data were anonymised and transferred via file encryption service to the statistician.
Statistical analysis
The upper limit of the reference interval (‘normal range’) in healthy controls was defined by the 95th percentile.13 26 Power calculations indicated that if 40 HCs are used to define reference intervals for 96-hour pH studies, then the CI for a diagnostic threshold for GORD of 4.4% is 3.2%–6.1% (CI 2.9%). This calculation assumed a log normal distribution fitted to published data, using a median of 2.0% and 95th percentile of 5.3% time pH<4.13 The study aimed to recruit 80 HCs to allow for exclusions and incomplete data.
Median and 95th percentile values were calculated for non-normally distributed AET over 24, 48, 72 and 96 hours and compared against the single ‘worst-day’ (ie, most pathological 24-hour period). Retrospectively, the same analysis was applied to results from consecutive patients with normal endoscopy and erosive oesophagitis LA A, B, C, D that completed 96-hour wireless pH studies off acid suppressants. Patients with high-grade oesophagitis (LA grades C and D) were considered a single group as small numbers in each group precluded meaningful calculations.24
Linear regression analysis was performed to assess the relationship between endoscopic findings with average AET, adjusted for gender, age and duration (24, 48, 72 and 96 hours). For the worst-day AET, the analysis was adjusted for these factors and the interaction between worst day and duration. Receiver operator curve (ROC) analysis was performed on the 96-hour data to determine thresholds for discriminating between patients with definite endoscopic evidence of GORD and HCs for both average AET and worst-day AET. Youden’s index was used to determine the optimal cut-offs. Upper 95th percentile values from normal subjects and lower 5th percentile values from individuals with proven GORD defined by oesophagitis were used to assess the overlap between health and disease. Statistical analyses were performed using Stata V.15 (StataCorp 2018).
Results
Healthy controls
Eighty HCs (mean age 27, 18–53 years, 65% female) were recruited: 30 in Nottingham (mean age 27, 21–50 years, 60% female) and 50 in London (mean age 27, 18–53 years, 68% female). Participants were symptom-free at screening. Per protocol, 12 HCs were excluded due to oesophagitis on endoscopy (n=10 LA grade A, 2 grade B). Early detachment or reception loss resulting in <92-hour pH data occurred in 22 individuals (65% completion rate). Six reported reflux symptoms during the ambulatory study. One was excluded due to incorrect placement of the pH sensor. Thus, 39 out of 80 HCs (49%, mean age 28 (18–53) years, 72% female, mean BMI 24 kg/m2 (range 19–33.9 kg/m2) had full data for inclusion in the final analysis. No demographic or physiological differences were present between HCs with prolonged pH data in the analysis and those excluded.
No complications occurred with the Bravo system (capsule retention, failure to detach or aspiration, tears in the mucosa, bleeding, perforation). The catheter-free procedures were well tolerated (satisfaction score 4.1/5) by HCs, and 84% (36/43) in education or employment attended courses or went to work during the test.
Reference intervals
Median and 5th and 95th percentiles in HCs for AET on each day of the study and over 24, 48, 72 and 96 hours for both the average of all days recorded (‘average-day analysis’) and the single most pathological day (‘worst-day analysis’) are illustrated in tables 1 and 2. The upper limit of reference of AET for the 96-hour studies was 4.6% for any study day and 6.9% for the worst day. No difference between the study days was found (p=0.52 Kruskal-Wallis test).
Oesophageal acid exposure time (AET) in healthy controls during each 24-hour period
Oesophageal acid exposure time in healthy controls assessed during 24, 48, 72 and 96-hour study period
Clinical study
Between March 2008 and November 2019, 2095 consecutive patients with symptoms of GORD underwent catheter-free pH monitoring. Of these, 306 patients were excluded due to studies lasting <48 hours, invalid or missing data, and non-compliance with the protocol. Endoscopy results were recorded in 1789 patients fulfilling the study criteria (tables 3 and 4). Of these, 944 patients (53%, mean age 46, 16–85 years, 65% female) completed >92-hour study. Eight hundred and eight (85%) had normal endoscopy, and 136 (15%) had erosive oesophagitis categorised by the LA classification having stopped PPI medication at least 1 week prior to investigation.
Endoscopy findings in patients with reflux symptoms referred for ambulatory pH studies
Demographic data, screening, endoscopic findings and capsule attachment
HCs had lower average and worst-day AET than patients with normal endoscopy and those with erosive oesophagitis LA ABCD (p<0.0001 Kruskal-Wallis test). Patients with normal endoscopy had lower AET than erosive oesophagitis LA ABCD (p<0.0001). Further, patients with LA A had lower AET than those with LA B or LA CD (p<0.0001).
An association between median and 95th percentile values for average and worst-day AET over 96 hours and endoscopic findings was present, such that acid exposure increases with oesophagitis grade (table 5 and online supplemental table S1).
Supplemental material
Comparison of average-day AET in healthy controls and patients with reflux symptoms referred for ambulatory pH studies
The ‘dose–response relationship’ between AET and with the presence and the severity of mucosal disease is evident; however, the box and whiskers plot highlights also the large variability in AET for each patient group (figure 1 (average-day AET), online supplemental figure S1 (worst-day AET)).
Box plots demonstrate the overlap of average acid exposure time (AET) in healthy controls with patients classified by the presence and severity of reflux oesophagitis. Dots are outliers. The dashed lines at 4% and 6% AET represent the conditional and conclusive Lyon classification diagnostic thresholds for GORD diagnosis.
Linear regression analysis
Linear regression analysis for the average-day analysis with AET as the dependent variable (table 6) shows a significant difference between HCs, symptomatic patients with normal endoscopy, patients with any grade (LA ABCD), moderate to severe (LA BCD) and severe oesophagitis (LA CD). The AET for severe oesophagitis (LA CD) is different from symptomatic patients with normal endoscopy and LA A oesophagitis (95% CIs of these groups do not overlap), but not LA B oesophagitis (95% CIs of these groups overlap).
Results of linear regression analysis of the 96-hour average AET and endoscopic findings
Linear regression analysis of the worst-day data with total AET as the dependent variable reveals similar findings (online supplemental table S2).
ROC analysis
Receiver operating curve (ROC) analysis estimating the diagnostic accuracy of endoscopic findings for GORD diagnosis with reference to average AET and worst-day AET from 96 hours studies is presented in table 7 and online supplemental table S3 respectively. The optimal thresholds for discriminating between HCs and patients with any grade of oesophagitis (LA ABCD), between HCs and patients with moderate to severe endoscopic evidence of GORD (LA BCD) and between HCs and patients with severe erosive oesophagitis (LA CD) for average AET and worst-day AET are shown in table 7, highlighted in bold text on the ROC curves in figure 2 and online supplemental figure S2, respectively. This analysis shows that 87% of patients with moderate to severe reflux oesophagitis (LA BCD) had average AET>4.3%, and 85% had worst-day AET>6.9%. By comparison, 93% of patients with severe oesophagitis (LA CD) had average AET>4.7%, and 100% had worst-day AET>7.1%.
Summary of results from receiver operating curve (ROC) analysis for the diagnostic accuracy of endoscopic findings for GORD diagnosis based on 96-hour average AET
ROC analysis with tables presenting results for the full range of potential diagnostic thresholds for discriminating between healthy controls and patients with erosive oesophagitis based on average AET from the 96-hour studies. Upper panel: any oesophagitis grade. Middle panel: moderate to severe reflux oesophagitis (LA grades B, C, D). Lower panel: severe oesophagitis (LA grades C, D). In each case, the optimal diagnostic threshold is highlighted in bold text.
Use of AET to discriminate between health and disease
The 5th and 95th percentiles for average and worst-day AET in controls and patients with erosive oesophagitis are shown in table 8 and online supplemental table S4. Histograms of the data, for controls and patients with the 95th centile for controls and the 5th centile for cases marked, reveal a clear division (ie, essentially no overlap) between the ‘upper limit of normal’ AET in HCs and results from individuals with LA CD reflux oesophagitis (figure 3 and online supplemental figure S3). There is limited overlap between HCs and individuals with LA BCD but considerable overlap if any grade of oesophagitis is considered. Overall, the average day data display less overlap than worst-day data (table 8 and figure 3, online supplemental table S4 and figure S3).
Summary of results showing 5th and 95th percentiles in healthy controls and patients with erosive oesophagitis for average day AET from the 96-hour studies
Histogram showing the distribution of 96-hour average AET in healthy controls and patients with erosive oesophagitis. Upper panel: any oesophagitis grade. Middle panel: moderate to severe reflux oesophagitis (LA grade B, C, D). Lower panel: severe oesophagitis (LA grade C, D). The overlap between health and disease is highlighted (upper 95th percentile limit for healthy controls, 5th percentiles for cases).
Validation of Lyon classification
Reflux oesophagitis
Linear regression analysis shows a significant difference between patients with severe oesophagitis (LA CD), moderate to severe erosive oesophagitis (LA BCD) and any grade of oesophagitis (LA ABCD), respectively, and the HC group for average AET and worst-day AET. There is also a highly significant difference between LA grade A, LA grade B and the control group for average AET and worst-day AET. Severe oesophagitis (LA CD) is significantly different from LA A for average AET and worst-day AET; however, this did not reach statistical significance for LA B (online supplemental table S2).
Ambulatory pH studies
The upper limit of reference (‘normal’) of AET for the 96-hour studies calculated using the median and 95th percentile was 4.6% for the average day and 6.9% for the worst-day analysis.
In HCs and each patient group, there is much less variability of the average-day AET data compared with worst-day AET data, whether this is reported as median (5th–95th percentiles), IQR or displayed in box and whiskers plots. Further, histograms show less overlap between normal subjects and individuals with LA BCD and LA ABCD oesophagitis for average-day data when compared with the worst-day data (figure 3 and online supplemental figure S3).
ROC analysis shows that the optimal diagnostic thresholds that discriminate between healthy controls and patients with severe oesophagitis (LA CD), moderate to severe oesophagitis (LA BCD) and any oesophagitis grade (LA ABCD) was 4.7%, 4.3% and 4.3%, respectively, based on average AET, and 7.1 %, 6.9% and 6.9%, respectively, based on worst-day AET.
Discussion
Ambulatory measurement of oesophageal acid exposure provides an objective basis for the diagnosis of GORD. This analysis of endoscopic and physiological data from a large group of HCs and patients referred for investigation of reflux symptoms provides an important validation of diagnostic thresholds for GORD from the Lyon classification.27 This research also provides insight into the optimal analysis of prolonged pH studies.
The average AET is almost identical on all four test days; however, there is important day-to-day variation in AET (table 1).15 18 As a result, the reproducibility of GORD diagnosis based on any 24-hour period is suboptimal.15 18 Increasing the duration of pH measurement from 24 to 96 hours progressively improves the reproducibility of results.18 What remains uncertain is the most appropriate diagnostic threshold and the best method for analysis of pH data for GORD diagnosis. The Lyon Consensus considers AET<4% to be physiological and >6% to be pathological; however, these limits that are based on expert opinion have not been tested prospectively. To validate these thresholds, it is necessary to compare AET results with an independent assessment of GORD severity. The presence of reflux oesophagitis has been used for this purpose since the introduction of ambulatory pH measurement by Johnson and DeMeester.28
This study analysed data from HCs and patients that successfully completed prolonged, 96-hour wireless pH studies with only minimal loss of data. About half the study participants met these stringent criteria. Those included had similar characteristics as those that did not complete 96-hour studies and are likely representative of the full cohort. Patients were instructed to stop PPI therapy at least 7 days before the study. The effects of these medications on gastric acid secretion are no longer present after 1 week; however, recurrence of oesophagitis requires more time with the extension of a PPI-free period from 1 week to indefinitely result in a deterioration of erosive oesophagitis by one grade.24 29 This is one reason why only a minority of patients had mucosal disease on endoscopy. Nevertheless, the large size of the cohort ensured that sufficient patients with reflux oesophagitis (n=136) contributed 96-hour pH data to the analysis.
There is a significant difference between AET measured in HCs and patients without mucosal disease compared with patients with reflux oesophagitis (table 6, online supplemental table S2). A positive, ‘dose-dependent’ relationship between AET with the severity of oesophagitis is present (figure 1, online supplemental figure S1). This confirms that such data can be used for purposes of validation. A key issue concerns the oesophagitis grade that should be considered conclusive for GORD diagnosis. In a large, population-based study, 16% of healthy, asymptomatic subjects had reflux oesophagitis (10% grade A, 4% grade B, 2% severe oesophagitis or Barrett’s oesophagus).30 Similarly, in the current series, 12 (15%) HCs were excluded because oesophagitis was present on endoscopy (n=10 LA grade A, 2 grade B). These findings support the view that grade A oesophagitis is not pathological; however, the average AET in patients with LA grade B is in the conclusively pathological range (>8% AET) and higher than that measured in those with mild, grade A oesophagitis or non-erosive disease (table 5). Only 9 out of 60 (15%) patients with LA B had average day AET<4%. Additionally, the ‘overlap’ of AET results from HCs and from patients with moderate to severe reflux oesophagitis (LA grade BCD) is small (7.8%), just above the traditional 5% threshold used to define the pathological range in clinical diagnostics (figure 3). By contrast, discrimination between health and disease is almost complete for severe, LA grade CD oesophagitis (table 8, figure 3, online supplemental table S4 and figure S3). Taken together, these results highlight that LA grade B oesophagitis provides at least conditional evidence of objective, pathological GORD.31 This assertion is supported by results from the longitudinal ProGERD study that observed frequent, spontaneous resolution of mild oesophagitis on repeated endoscopy but a significant risk of disease progression in patients with moderate to severe oesophagitis.32 Similarly, clinical studies report good outcomes for medical and surgical antireflux treatment in patients with this form of erosive GORD.10 33
The upper limit of oesophageal AET in HCs as determined by the 95th percentile of the normal range is 4.6% for the average day and 6.9% for the worst-day AET analysis (table 2). Thus, although most HCs have a low AET on most days (96 hours average 1.32% AET), it is ‘normal’ for some healthy individuals to have occasional days with 6%–7% AET. These findings are in remarkably close agreement with the results of an ROC analysis that indicates the optimal diagnostic threshold that discriminates between HCs and patients with moderate to severe oesophagitis (LA grade BCD) is 4.3% based on average AET and 6.9% based on the worst-day AET (table 7, online supplemental table S3). The optimal threshold to discriminate HCs and patients with severe oesophagitis (LA grade CD) was somewhat higher, 4.7% based on average AET and 7.1% based on the worst-day AET (table 7, online supplemental table S3). These results provide conditional support for the Lyon classification which defines AET <4% as normal (physiological). By contrast, based on analysis of worst-day AET data, the AET >7% threshold provides more conclusive evidence for GORD diagnosis than the current AET >6% threshold. These results suggest that intermediate values between 4% and 7% AET, a slightly wider range than in version 1.0 of the classification, should be considered inconclusive and require additional supportive evidence to establish GORD diagnosis.19 Thresholds derived from physiological measurement require support from clinical studies. A recent double-blinded clinical trial in patients who underwent 96-hour pH studies demonstrated that the optimal predictor of successful withdrawal of PPIs was an average day AET <4.3% and a worst-day AET <7.4% on reflux monitoring.33 The close agreement with this study confirms that the normal and pathological ranges defined by the present analysis are clinically relevant.
Independent of diagnostic thresholds, it is important to identify the most appropriate analysis of prolonged pH measurements in clinical practice. Worst-day analysis maximises the sensitivity and yield for GORD diagnosis in patients with reflux symptoms (ie, low risk of false-negative result).34 However, this method increases the risk of a false-positive diagnosis due to high day-to-day variation in AET.7 16 18 Graphical illustration of results demonstrate a greater overlap between results from HCs and patients with oesophagitis for the worst day than the average AET analysis (figures 1 and 3, supplemental figure S1 and S3). Average AET analysis has a high specificity for GORD diagnosis because inconclusive (borderline) or discordant data from any one 24-hour measurement period will increasingly become ‘more representative’ of the true result as more information is collected. (Note that diagnosis based on average day AET is closely correlated with that based on ‘predominant AET pattern’ analysis proposed by Hasak et al.17 The latter method tallies the number of days with negative, inconclusive and positive GORD diagnoses with the final result determined by the most frequent (predominant) pattern. This simple approach may increase the proportion of patients with a conclusive diagnosis; however, it cannot be used for measurements <3 days, discards incomplete data (<24 hours) and is inconclusive if the number of positive and negative days are identical). Thus, consistent with recent guidelines,7 current findings indicate that the result of prolonged pH studies should be based on all the data available and not focused on the ‘worst day’ because this is a more stable and statistically robust approach when applied to individual cases.
Strengths and limitations of this study
A group of 39 well-characterised HCs with no evidence of GORD and complete 96-hour pH studies provided 156 days reference data. The numbers of patients with reflux oesophagitis in the clinical cohort (136/944) allowed for a statistically well-powered validation study to be performed. This addresses limitations of previous studies that analysed more heterogeneous groups and/or did not have sufficient numbers to provide conclusive results.13 26 35 The mean age of HCs was lower than that of patients (table 4); however, the increase in AET that can be attributed to the 15–20 year age gap is estimated to be <2%, based on results from the same population.36 37 The study was performed in a tertiary referral unit; however, the Oesophageal Physiology Laboratory in Guy’s Hospital (GSTT) is the main provider for this service in the South-East Thames region. The relatively short, 7-day time between stopping PPI medication and endoscopy can be criticised. Stopping acid suppressant medication for prolonged periods of time in highly symptomatic patients is difficult and not acceptable to the ethics committee. In any case, recurrence of reflux oesophagitis to the grade expected without treatment may take many weeks.32 There may be a selection bias in patients referred for pH studies, as only a minority of patients with severe oesophagitis may be referred for diagnostic purposes unless an operation is planned.
This study does not imply that prolonged pH measurements by a wireless system are superior to prolonged measurements by catheter-based systems, although the latter are less likely to be tolerated by patients.14 Two direct comparisons of wireless and catheter-based recordings have shown good correlation between pH measurements recorded by the two techniques.38 39 One potential limitation of pH only measurements (catheter-based or wireless) compared with pH impedance is the possibility to overestimate AET due to intake of acidic foods.40 In the present study, all subjects were specifically asked to document meal times and avoid acidic drinks.
Conclusions and implications for clinical practice
This study defines the normal reference range for 96-hour ambulatory wireless pH monitoring and identifies pathological thresholds that discriminate between healthy controls and patients with conclusive GORD diagnosis, based on the presence of moderate to severe erosive oesophagitis. The results support the use of prolonged, ideally 4-day, ambulatory pH studies to counter the large day-to-day variation in AET that is seen in patients with reflux symptoms. The study will also inform the decisions of the committee when the next version of the Lyon Consensus is drafted. First, it adds to the evidence that moderate, LA grade B oesophagitis provides at least conditional evidence of GORD. Second, it indicates that the threshold for a conclusive diagnosis of GORD by ambulatory pH measurement should be refined upwards from 6% to 7% AET, with average day AET analysis preferred to worst-day AET analysis when prolonged pH measurements are available. These findings will improve the diagnosis of GORD and facilitate the selection of patients for medical and surgical treatment.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. Available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved at each centre by the NHS Health Research Authority, UK (REC Ref: 09/H0802/104, 18/LO/1467), University of Nottingham and King's College Ethics committees (ClinicalTrials.gov number NCT00481949, NCT03417817). The healthy volunteer component of the study at GSTT was included in the NIHR Clinical Research Network Portfolio (CPMS ID 39211).
Acknowledgments
We are grateful to Dr Jeremy Sanderson, Consultant Gastroenterologist at Guy's and St Thomas' Hospitals, London, UK, and Professor of Gastroenterology at King's College London for his advice and support with funding the healthy control cases; administrative and nursing staff in Endoscopy departments at Nottingham University Hospital NHS Trust and Guy's and St Thomas' NHS Foundation Trust for their contribution to the seamless integration of additional cases into routine clinical service; and to all GI physiologists and assistant practitioners at NIHR Biomedical Research Unit in Nottingham and Guy's Hospital Oesophageal Physiology Laboratory for their assistance in conducting and reporting the procedures.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @SebastianZeki
Correction notice This article has been corrected since it published Online First. The author affiliations have been updated and a duplicated reference removed.
Contributors Study design/conception: R-IR, MRF, TW. Acquisition of data: R-IR, MRF, ET, SZ, JMD, TW. Analysis of data: R-IR, MF, ET, FW, TW. Interpretation of data: R-IR, MRF, ET, FW TW. Drafting of the manuscript: R-IR, TW. Critical revision of the manuscript: R-IR, MRF, SZ, JMD, JJ, FW, TW. Final approval of the manuscript to be published: R-IR, MRF, ET, SZ, JMD, JJ, FW, TW.
Funding This study was funded by institutional research funds from Guy's and St Thomas' Charity, SPF872 and supported by grants and catheter-free pH monitoring equipment from Given Imaging/Covidien AG, now subsidiaries of Medtronic plc. Medtronic was not involved in discussions of the study design and did not contribute to the drafting of the manuscript.
Competing interests MRF has received research and/or educational funding from Medtronic, MMS/Laborie, Sandhill Scientific/Diversatek and Reckitt Benckiser. R-IR, ET, SZ, JMD, JJ, FW, TW: None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.