Ajg_41084.tex

ORIGINAL CONTRIBUTIONS
Reproducibility of Multichannel Intraluminal ElectricalImpedance Monitoring of Gastroesophageal RefluxAlbert J. Bredenoord, M.D., Bas L.A.M. Weusten, M.D., Ph.D., Robin Timmer, M.D., Ph.D.,and Andre J.P.M. Smout, M.D., Ph.D.
Department of Gastroenterology, Sint Antonius Hospital, Nieuwegein; and Gastrointestinal Research Unit,University Medical Center, Utrecht, The Netherlands Esophageal impedance measurement is a novel method for gastroesophageal reflux monitoring.
Reproducibility is an important aspect of every biomedical test. The aim of this study was thereforeto asses the reproducibility of gastroesophageal reflux monitoring using impedance measurements.
Impedance and pH signals were recorded in 20 healthy volunteers during 90-min postprandialperiods on two separate days. Hourly rates of gas, liquid, and mixed gas-liquid reflux episodes weremeasured in each recording period as well as percentage of time with pH < 4 and rate of acidreflux episodes. As a quantitative description of inter- and intraindividual variation for each variable,the mean percentage of covariation (100 × SD/mean: %COV) was calculated. As a second measurefor reproducibility, Kendall’s coefficients of concordance (W values) were calculated.
For all variables, interindividual %COV was at least 50% higher than intraindividual %COV.
Statistically significant concordances were found for gas reflux (W = 0.81, p = 0.04) and mixedreflux (W = 0.85, p = 0.03) while concordance for liquid reflux tended to be significant (W = 0.75,p = 0.08). This was comparable to the reproducibility of the number of acid reflux episodes andpercentage of time with pH < 4 (W = 0.78, p = 0.05 and W = 0.88, p = 0.02, respectively).
Postprandial gastroesophageal reflux data assessed with impedance monitoring are asreproducible as assessed with pH monitoring.
INTRODUCTION
is good (9–15). The aim of this study was to investigate thereproducibility of esophageal electrical impedance monitor- Patients with symptoms of heartburn and regurgitation often ing and to compare this with the reproducibility of acid reflux suffer from reflux of acidic gastric contents into the esoph- agus. This can be shown with a pH study. However, thereis evidence that reflux symptoms may also occur in associ- MATERIALS AND ME THODS
ation with nonacid reflux (1). Furthermore, a subset of pa-tients with reflux disease experiences no relief after effective Patients
acid suppression. It has been suggested that reflux of nonacid Twenty healthy volunteers (14 males and 6 females: mean substances such as pepsin and bile acids may play a role in age 28 yr, range 19–46 yr) underwent two separate recordings the genesis of reflux symptoms (2). Furthermore, it has been of esophageal impedance and pH within an interval of 1–2 shown that nonacid components of the refluxate also play a wk. Patients were free of any gastrointestinal symptoms and role in the genesis of Barrett’s epithelium (3).
were not taking any medication. Informed written consent Multichannel intraluminal electrical impedance monitor- was obtained before the start of the study and the protocol was ing makes it possible to detect nonacid liquid and gas flow approved by the medical ethics committee of the University through the esophagus (4). Recent studies have shown that combined pH and impedance monitoring provide additionalinformation compared to pH measurements alone (5–7). The Study Protocol
fact that impedance monitoring can also detect gas transport After an overnight fast a routine esophageal manometry was through the esophagus makes it a useful tool to study air performed to determine the distance from the nostrils to the swallowing and excessive belching as well (8).
lower esophageal sphincter (LES). Thereafter, the impedance Currently, pH monitoring is a generally accepted test for and the pH catheters were introduced transnasally and po- the quantification of gastroesophageal reflux, and various sitioned based on the manometric findings (see below). A studies have shown that reproducibility of pH monitoring standardized meal was offered consisting of one hamburger Bredenoord et al.
(McDonald’s Quarter Pounder), 20 g of fresh onions, 44 g of (5, 17, 18). Furthermore, using the pH tracings, liquid re- potato chips, and 475 ml of orange juice (in total 967 kcal).
flux, mixed gas-liquid reflux, and gas reflux were classi- This meal was used to provoke reflux in a previous study of fied as acidic or nonacidic, using a threshold of pH < 4.
our group (16). The meal had to be finished in 30 min. After Percentage of time with pH < 4 was also assessed. Analysis the meal, impedance and pH signals were recorded for 90 was performed manually by two physicians with experience min. Patients remained seated for the duration of the study.
in this area; disagreement was solved by consensus. Theseobservers were blinded to the outcome of the other measure- Impedance and pH Monitoring
For impedance monitoring a 7-channel impedance catheter “Air swallows” were defined as swallows in which the liq- was used (Aachen University of Technology, FEMU, Aachen, uid bolus, identifiable by a decrease in impedance, was pre- Germany). This catheter (outer diameter 2.3 mm) enabled ceded by an increase in impedance of at least 1,000 recording from seven segments, each recording segment be- ing 2-cm long. The recording segments were located at 0–2, was found to be well above the amplitude of baseline noise.
2–4, 4–6, 8–10, 10–12, 14–16, and 17–19 above the upper The number of regular swallows as well as the number of “air border of the manometrically localized LES. Impedance sig- nals were stored in a digital system (Aachen University of Evaluation of Reproducibility
Technology, FEMU, Aachen, Germany) using a sample fre- Standard deviations (SD) and percentage coefficient of varia- quency of 1000 Hz. Intraluminal pH monitoring was per- tion (100 × SD/mean: %COV) were calculated for impedance formed with a glass pH electrode (Ingold A.G., Urdorf, and pH data. The mean %COV of the 20 values of the first Switzerland) and data were stored in a digital datalogger measurement and the mean %COV of the 20 values of the sec- (Orion, MMS, Enschede, the Netherlands) using a sampling ond measurement were calculated. An overall mean %COV frequency of 2 Hz. The pH glass catheter was positioned 5 was derived as a measure of interindividual variation. Fur- cm above the upper border of the LES. Using a cable that thermore, the mean %COV of the first and the second mea- connected the pH datalogger with the impedance datalog- surement in the 20 volunteers was calculated. This value, ger the pH signals were stored on both dataloggers enabling calculated from the values of the 20 individuals, was used as a measure of intraindividual reproducibility (19).
Data Analysis
As a second measure for reproducibility, Kendall’s Previously established criteria were used to identify swal- coefficients of concordance (W value) were calculated us- lows, gas reflux, liquid reflux, and mixed gas-liquid reflux ing the mean values for the named variables from individual Table 1. Intraindividual and Interindividual Reproducibility of Measurements of Mixed Gas-Liquid Reflux (Rates/h)
SD = standard deviation; %COV = percentage of covariance. The mean SD and mean %COV for the two columns (day 1 and day 2) constitute an index of interindividual variation,whereas the equivalent values for the 20 individual rows are an index of intraindividual variation.
Reproducibility of Esophageal Impedance Monitoring
swallowing. For virtually all parameters, the interindivid-ual %COV was considerably higher than the intraindividual %COV. Concordance between the measurements was sub- stantial for all parameters with Kendall’s W values varying between 0.75 and 0.90 (Table 2). Statistically significant con- cordance was reached for hourly rate of gas reflux episodes, mixed reflux episodes, acid reflux episodes, swallowing, air swallowing, and percentage of time with pH below 4. For three parameters the reproducibility was borderline signifi- cant: hourly rate of liquid reflux episodes, mixed acid reflux episodes, and liquid acid reflux episodes. The hourly rate of these three types of events was rather low, as can be seen in Figures 2–5 show Bland-Altman plots for hourly rates of liquid reflux episodes, gas reflux episodes, swallows, and air Figure 1. Reproducibility of mixed reflux rate (Bland-Altman plot).
swallows. In these figures data points are closely scatteredaround the x-axis, indicating a relatively small difference be- recordings and tested for significance. An error probability of tween the two measurements as compared to the mean of the p ≤ 0.05 was considered statistically significant. Throughout two measurements. Furthermore, the scattering around the the manuscript data are presented as mean ± SEM.
x-axis is symmetrical, indicating that the values of the sec- Assessment of reproducibility was facilitated by present- ond measurement are randomly distributed around the mean ing data in Bland-Altman plots. In these plots the difference of the two measurements and that no upward or downward between the first and second measurement is plotted against the mean value of the two measurements, which makes itpossible to graphically compare these two values. When thedifference between measurements on the first day and the DISCUSSION
second day is small, data points are scattered closely to the x-axis. Symmetrical scattering around the x-axis indicates that Previous studies showed that pH monitoring is a fairly there is no trend toward a difference of the measurements of reproducible method for detection of gastroesophageal reflux the second day compared to the measurements of the first day (9–15). In a study by Wiener et al., in which a cutoff value and the difference between the two measurements occurs in of 4% of the time with pH < 4 was used for the diagnosis of gastroesophageal reflux disease (GERD), reproducibilityfor this diagnosis was 80% (15). Most studies that investi- gated reproducibility of the percentage of time with pH < 4,reported a higher reproducibility.
As shown in Table 1 the mean rate at which mixed gas-liquid We studied reproducibility of impedance monitoring for reflux episodes occurred was 4.0/h. As illustrated the vari- the detection of gastroesophageal reflux. As shown in Table 2, ation between different patients was much larger than the %COV is much larger between patients than it is within pa- variation occurring within the same patient (Fig. 1). Scatter- tients. Furthermore, testing for concordance using Kendall’s ing of the data points occurs close to the x-axis in a rather W test showed that the concordance between the two mea- surements of gas and mixed reflux was statistically signifi- As shown in Table 2 similar findings were made for the cant, while concordance between the two measurements of other reflux parameters, for swallowing rate, and for air liquid reflux tended to be statistically significant.
Table 2. Reproducibility of Swallow and Reflux Parameters
SD = standard deviation; %COV = percentage of covariance.
Bredenoord et al.
Figure 4. Reproducibility of swallow rate (Bland-Altman plot).
Figure 2. Reproducibility of liquid reflux rate (Bland-Altman plot).
Results from our pH studies confirmed those obtained in testing with intraluminal impedance monitoring can be con- previous reports, and both percentage of time with pH < 4 and sidered as an important validation of this technique. Apart hourly rate of reflux episodes were found to be reproducible.
from reflux parameters, swallowing rate and air swallowing Reproducibility of the rate of liquid reflux, mixed gas-liquid rate as determined with intraluminal impedance were also acid reflux, and liquid gas-liquid acid reflux episodes were found to be reproducible in healthy volunteers. The frequen- borderline significant. This is most likely due to a type 2 er- cies found in this study were well within the reported range ror, resulting from the rather low rate of (acid) reflux episodes in these healthy volunteers as well as to the relatively short The reproducibility of reflux monitoring using esophageal recording time (90 min). Others have shown that reproducibil- impedance measurement was found to be comparable to the ity increases with the length of the study, showing much reproducibility of reflux testing with pH monitoring. This higher intraindividual concordance in 24-h measurements indicates that esophageal impedance monitoring can be an than in 3-h measurements (10). Since in our 90 min study important new tool for studying and diagnosing GERD (10).
standardized meals and position were used, the reproducibil- The additional value of impedance measurements in conju- ity found may theoretically differ from the reproducibility of gation with pH monitoring explains the increasing use of this 24-h ambulatory impedance monitoring.
technique. A recent publication of normal values and the con- Reproducibility of a biomedical test is determined not only sensus report concerning terminology have opened doors to by patient-related factors, but also by technical factors. Dif- more widespread clinical application (21, 22). Furthermore, ferences in catheter position, data acquisition, and data anal- the reproducibility of impedance measurements indicates that ysis can influence the results of impedance monitoring. The this technique might be suitable for investigation of the ef- degree of reproducibility is only acceptable when variations fects of new drugs and other treatment modalities on the caused by technical factors are rather small. The fact that occurrence of gastroesophageal reflux. However, one should we found a good reproducibility of gastroesophageal reflux take into account that, although intraindividual variations are Figure 3. Reproducibility of gas reflux rate (Bland-Altman plot).
Figure 5. Reproducibility of air swallow rate (Bland-Altman plot).
Reproducibility of Esophageal Impedance Monitoring
small compared to interindividual variations, a normal day- 5. Sifrim D, Holloway R, Silny J, et al. Acid, nonacid, and to-day variance within patients does occur and that this may gas reflux in patients with gastroesophageal reflux disease limit the statistical power of studies with small sample sizes.
during ambulatory 24-hour pH-impedance recordings. Gas-troenterology 2001;120(7):1588–98.
In ambulatory studies physiological intraindividual variation 6. Sifrim D, Silny J, Holloway RH, et al. Patterns of gas and could be larger than the variations found in this study, since liquid reflux during transient lower oesophageal sphincter in our setting patients were studied in the same position and relaxation: A study using intraluminal electrical impedance.
used the same standardized meal during each of the two mea- 7. Vela MF, Camacho-Lobato L, Srinivasan R, et al. Simulta- neous intraesophageal impedance and pH measurement of The Porto consensus report introduced a new nomenclature acid and nonacid gastroesophageal reflux: Effect of omepra- for reflux monitoring (22). In this nomenclature, definitions zole. Gastroenterology 2001;120(7):1599–606.
of reflux are based on the combined use of pH and impedance 8. Bredenoord AJ, Weusten BL, Sifrim D, et al. Aerophagia, monitoring. Reflux is thus divided in acid reflux (fall of pH gastric and supragastric belching. A study using intraluminal below 4), superimposed reflux (when pH is already below electrical impedance monitoring. Gut. 2004;53:1561–65.
9. Emde C, Armstrong D, Castiglione F, et al. Repro- 4), weakly acidic reflux (when pH is between 4 and 7), and ducibility of long-term ambulatory esophageal combined weakly alkaline reflux (pH above 7). Reproducibility of these pH/manometry. Gastroenterology 1991;100(6):1630–7.
four types of reflux depends on the combination of pH and 10. Johnsson F, Joelsson B. Reproducibility of ambulatory impedance monitoring techniques. The aim of this study was oesophageal pH monitoring. Gut 1988;29(7):886–9.
to investigate reproducibility of impedance monitoring alone 11. Wang H, Beck IT, Paterson WG. Reproducibility and physiological characteristics of 24-hour ambulatory and therefore we did not investigate reproducibility of these esophageal manometry/pH-metry. Am J Gastroenterol In conclusion, we showed that intraluminal impedance 12. Franz´en T, Tibbling Grahn L. Reliability of 24-hour measurement is a reproducible method for studying (air) oesophageal pH monitoring under standardized conditions.
swallowing, liquid reflux, gas reflux, and mixed gas-liquid Scand J Gastroenterol 2002;37(1):6–8.
13. Vandenplas Y, Helven R, Goyvaerts H, et al. Reproducibility reflux. The reproducibility of measurement of these reflux of continuous 24 hour oesophageal pH monitoring in infants events is comparable to the reproducibility of acid reflux mea- 14. Vaezi MF, Schroeder PL, Richter JE. Reproducibility of proximal probe pH parameters in 24-hour ambu-latory esophageal pH monitoring. Am J Gastroenterol ACKNOWLEDGMENTS
15. Wiener GJ, Morgan TM, Copper JB, et al. Ambulatory 24- Ms. A. Baron is gratefully acknowledged for her contribu- hour esophageal pH monitoring. Reproducibility and vari-ability of pH parameters. Dig Dis Sci 1988;33(9):1127–33.
tions to the conduct of this study. A.J. Bredenoord was sup- 16. Van Herwaarden MA, Samsom M, Rydholm H, et al.
ported by a clinical research trainee grant from Janssen-Cilag, The effect of baclofen on gastro-oesophageal reflux, lower oesophageal sphincter function and reflux symptoms inpatients with reflux disease. Aliment Pharmacol Ther2002;16(9):1655–62.
Reprint requests and correspondence: A.J. Bredenoord, M.D.,
17. Skopnik H, Silny J, Heiber O, et al. Gastroesophageal re- Department of Gastroenterology, St. Antonius Hospital, P.O. Box flux in infants: Evaluation of a new intraluminal impedance 2500, 3430 EM Nieuwegein, The Netherlands.
technique. J Pediatr Gastroenterol Nutr 1996;23(5):591–8.
Received August 17, 2004; August 19, 2004. 18. Fass J, Silny J, Braun J, et al. Measuring esophageal motility with a new intraluminal impedance device. Firstclinical results in reflux patients. Scand J Gastroenterol REFERENCES
19. Kramer MS, Feinstein AR. Clinical biostatistics. LIV.
1. Weusten BL, Smout AJ. Ambulatory monitoring of The biostatistics of concordance. Clin Pharmacol Ther esophageal pH and pressure. In: Castell DO, Richter JE, eds. The esophagus. Philadelphia(PA): Lippincott Williams 20. Lear CSC, Flanagan JB, Moorrees CFA. The frequency of deglutition in man. Arch Oral Biol 1965;10:83–99.
2. Katz PO. Review article: The role of non-acid reflux in 21. Shay S, Tutuian R, Sifrim D, et al. Twenty-four hour ambu- gastro-oesophageal reflux disease. Aliment Pharmacol Ther latory simultaneous impedance and pH monitoring: A mul- ticenter report of normal values from 60 healthy volunteers.
3. Falk GW. Barrett’s esophagus. Gastroenterology 2002; Am J Gastroenterol 2004;99(6):1037–43.
22. Sifrim D, Castell D, Dent J, et al. Gastro-oesophageal re- 4. Silny J. Intraluminal multiple electric impedance procedure flux monitoring: Review and consensus report on detec- for measurement of gastrointestinal motility. J Gastrointest tion and definitions of acid, non-acid, and gas reflux. Gut

Source: http://www.invive.in/wp-content/uploads/2011/07/Reproducibility-of-MIIpH-Monitoring2005.pdf