- Email: [email protected]
Comparison of Gastroesophageal Reflux Disease Questionnaire and Multichannel Intraluminal Impedance pH Monitoring in Identifying Patients With Chronic Cough Responsive to Antireflux Therapy
CHEST
Original Research SIGNS AND SYMPTOMS OF CHEST DISEASE
Comparison of Gastroesophageal Reflux Disease Questionnaire and Multichannel Intraluminal Impedance pH Monitoring in Identifying Patients With Chronic Cough Responsive to Antireflux Therapy Xianghuai Xu, MD; Qiang Chen, MD; Siwei Liang, MD; Hanjing Lv, MD; and Zhongmin Qiu, MD, FCCP
Background: Empirical therapy has been recommended as an initial clinical approach for treating gastroesophageal reflux-induced chronic cough (GERC). This study compared the predictive accuracy of the Gastroesophageal Reflux Disease Questionnaire (GerdQ) with the accuracy of multichannel intraluminal impedance pH monitoring (MII-pH) for GERC. Methods: A total of 126 consecutive patients with potential GERC were recruited to undergo MII-pH and complete the GerdQ. A final diagnosis of GERC was made after favorable response to consequent medicinal antireflux therapy, regardless of laboratory findings. The predictive accuracy of the GerdQ for GERC was assessed and compared with that of MII-pH. Results: GERC was confirmed in 102 of 126 patients (81.0%); cough was due to acid reflux in 55 (53.9%) and nonacid reflux in 47 (46.1%). The optimal cutoff point of the GerdQ for predicting GERC was defined as 8.0 according to the highest Youden index of 0.584, with a sensitivity of 66.7%, specificity of 91.7%, positive predictive value of 97.1%, and negative predictive value of 42.9%. A subanalysis for only acid GERC showed further improvement in the predictive accuracy of the GerdQ, corresponding to a sensitivity of 90.9%, specificity of 78.6%, positive predictive value of 71.4%, and negative predictive value of 96.4%. However, a meaningful GerdQ cutoff point for prediction of nonacid GERC could not be determined. In general, MII-pH was superior to the GerdQ for predicting GERC and acid GERC. Conclusions: The GerdQ can be used for predicting acid GERC but not nonacid GERC and is inferior to MII-pH. Trial registry: Chinese Clinical Trial Registry; No.: ChiCTR-ODT-12001899; URL: www.chictr.org CHEST 2014; 145(6):1264–1270 Abbreviations: GERC 5 gastroesophageal reflux-induced chronic cough; GERD 5 gastroesophageal reflux disease; GerdQ 5 Gastroesophageal Reflux Disease Questionnaire; MII-pH 5 multichannel intraluminal impedance pH monitoring; SAP 5 symptom association probability
reflux is a common cause of chronic Gastroesophageal cough. Based on pH values of refluxate in the 1,2
esophagus, gastroesophageal reflux-induced chronic cough (GERC) can be classified as acid or nonacid.3 Multichannel intraluminal impedance pH monitoring (MII-pH) can detect both acid and nonacid reflux4 and is the most sensitive and specific test for diagnosing GERC. Unfortunately, MII-pH has not been extensively used clinically because of its invasiveness and limited availability. Alternatively, empirical therapy has been advocated as an initial approach for treating
1264
GERC.1,5,6 Because empirical trials frequently fail to relieve patients’ cough,7 screening patients with potential GERC prior to antireflux treatment may improve therapeutic gains. The Gastroesophageal Reflux Disease Questionnaire (GerdQ) is a patient-centered and self-reported diagnostic instrument for gastroesophageal reflux disease (GERD)8 and can achieve an overall diagnostic accuracy similar to that made by a gastroenterologist, with a sensitivity of 64.6% and a specificity of 71.4%.8,9 However, to our knowledge the usefulness of the GerdQ in Original Research
GERC diagnosis has never been addressed. Therefore, we conducted a prospective study to compare the predictive accuracies of the GerdQ and MII-pH for GERC. Materials and Methods Patients Consecutive patients referred to our respiratory clinic for potential GERC between January 2011 and March 2013 were recruited for this study. Requirements for participation included any one of the following: (1) presence of chronic cough and typical refluxrelated symptoms, (2) GERC considered when previous laboratory workups failed to identify other possible causes of chronic cough,10 and (3) coexisting GERC suspected when the treatment of current etiologies failed to completely resolve a cough. Patients refusing to undergo or who were intolerant of MII-pH were excluded. The study protocol was approved by the ethics committee of Tongji Hospital [No. LL(H)-11-13] and registered with the Chinese Clinical Trials Register. Written informed consent was obtained from all subjects before enrollment. Study Design This was a single-center observational study. Initial patient assessments included a medical history, physical examination, plain chest radiography, lung function tests, and evaluation of cough severity as determined by validated cough symptom scores that rate a cough using six incremental scales (0 5 no cough, 5 5 worst cough)11 (e-Appendix 1). Additionally, patients were evaluated with the previously validated Chinese version of the GerdQ12,13 (e-Appendix 2). The GerdQ comprises six symptom-related items consisting of four reflux-related items positively related to GERD and two items negatively related to GERD. Patients were asked to recall how often they experienced the events described in the questionnaire during the preceding week and to score their answers on a four-point scale ranging from 0 to 3 for positive predictors and from 3 to 0 for negative predictors. The total scores represented a GerdQ score ranging from 0 to 18. A higher score signifies a greater possibility of GERD.8 MII-pH was performed as previously described3 after the patients had stopped taking acid-suppressing medication for at least 1 week. Briefly, a combined MII-pH catheter was inserted transnasally into the patient’s esophagus, with six impedance channel sensors (K6011-E10632; Unisensor AG) located 3, 5, 7, 9, 15, and 17 cm above the lower esophageal sphincter and an antimony pH electrode (819100; Medical Measurement System BV) positioned 5 cm Manuscript received July 16, 2013; revision accepted December 19, 2013; originally published Online First January 23, 2014. Affiliations: From the Department of Respiratory Medicine, Tongji Hospital, Tongji University School of Medicine, Shanghai, China. Part of this article was presented orally at the Seventh International Cough Symposium 2012: Taming of Chronic Cough, July 5-7, 2012, London, England. Funding/Support: This study was supported by grants from the National Natural Science Foundation of China [81170079] and Shanghai Shenkang Hospital Development Center [SHDC12012211]. Correspondence to: Zhongmin Qiu, MD, FCCP, Department of Respiratory Medicine, Tongji Hospital, Tongji University School of Medicine, No. 389 Xincun Rd, Shanghai 200065, China; e-mail: [email protected] © 2014 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details. DOI: 10.1378/chest.13-1634 journal.publications.chestnet.org
above the proximal border of the lower esophageal sphincter. A connected portable data logger (Ohmega; Medical Measurement System BV) stored data from all seven channels over 24 h. Reflux episodes recorded on the tracings of MII-pH were manually characterized by their impedance value as liquid, gas, or mixed liquidgas reflux or characterized using a pH meter as acidic (pH , 4.0), weakly acidic (pH 4.0-7.0), or weakly alkaline (pH . 7.0) reflux, with the latter two collectively referred to as nonacid reflux. The DeMeester score was calculated as a global measure of esophageal acid exposure. Symptom association probability (SAP) was used to represent the temporal association between cough recorded by patients on diary cards and reflux that had occurred during the preceding 2-min period.3,14 Abnormal acid reflux was defined as a DeMeester score ⱖ 14.72, SAP for acid reflux ⱖ 95%, or both. These scores were used for diagnosing acid GERC, whereas abnormal nonacid reflux was defined as an SAP for nonacid reflux ⱖ 95%, which was used for diagnosing nonacid GERC.3 Regardless of MII-pH findings, all patients received an 8-week course of standard antireflux therapy consisting of omeprazole 20 mg bid and domperidone 10 mg tid. If the initial treatment failed and MII-pH revealed abnormal reflux, patients received an augmented antireflux trial by doubling the dose of omeprazole to 80 mg/d or replacing domperidone with baclofen 20 mg tid.15 All study patients and investigators were blinded to GerdQ scores throughout treatment. GERC was finally diagnosed only when a cough disappeared (resolved completely) or improved (combined daytime and nighttime cough symptom score decreased by ⱖ 2 points)3 as recommended by American College of Chest Physicians guidelines.1 The predictive accuracies of the GerdQ and MII-pH for GERC were analyzed and compared (Fig 1). Statistical Analysis Data with normal distributions are expressed as mean ⫾ SD, whereas data with skewed distributions are expressed as median
Figure 1. CONSORT (Consolidated Standards of Reporting Trials) flow diagram of the study and algorithm for GERC. GERC 5 gastroesophageal reflux-induced chronic cough; GerdQ 5 Gastroesophageal Reflux Disease Questionnaire; MII-pH 5 multichannel intraluminal impedance pH monitoring. CHEST / 145 / 6 / JUNE 2014
1265
values (interquartile range). Unpaired Student t and Mann-Whitney U tests were used for data comparison where applicable. Receiver operating characteristic analysis was used to evaluate the predictive value of the GerdQ. The optimal cutoff point was determined on the basis of the highest Youden index. The relation between the GerdQ score and DeMeester score or SAP for acid reflux was analyzed by linear regression. Predictive values of the GerdQ and MII-pH were compared with the x2 test. Diagnostic consistency was evaluated by k test. Statistical calculations were performed with SPSS, version 17.0 (IBM) software.
Results Demographic Characteristics A total of 167 of 583 patients with chronic cough (28.6%) met the inclusion criteria for this study. Among this group of eligible patients, 126 (76%) completed evaluations by MII-pH and the GerdQ. Demographic characteristics of the final study group are shown in Table 1. Abnormal reflux was revealed by MII-pH in 108 of 126 patients (85.7%), and GERC was confirmed by empirical antireflux treatment in 80 patients (74.1%). Twenty-eight patients (25.9%) received augmented acid suppression treatment with omeprazole 80 mg/d, and GERC was confirmed in 11 patients (39.3%). The remaining 17 patients (15.7%) accepted therapy with baclofen, and GERC was further confirmed in eight (47.1%). Among nine nonresponders, cough in seven (77.8%) could be explained by upper-airway cough syndrome in two (22.2%) and cough variant asthma in five (55.6%), but the causes were unknown in two (22.2%). In 18 patients (14.3%) with a negative DeMeester score and SAP, nonacid GERC was determined in three patients (16.7%) on the basis of increased weakly acidic reflux events and favorable response to antireflux treatment. In total, GERC was diagnosed in 102 patients (81.0%), including 55 with acid GERC (53.9%) and 47 with nonacid GERC (46.1%). Additionally, the cough was completely resolved in 38 patients (69.1%) with acid GERC and 25 (53.2%) with nonacid GERC. A plot of patient response vs treatment time is shown in Figure 2. Table 1—Patient Demographic Characteristics Characteristic Male (female) sex Age, y Cough duration, mo Cough symptom score Daytime Nighttime Absence of heartburn or regurgitation FEV1 % predicted FVC % predicted FEV1/FVC, %
Value 59 (67) 45.6 ⫾ 18.3 8.5 (5.3-36.0) 3 (2-4) 1 (1-3) 49 (38.9) 95.3 ⫾ 12.1 97.2 ⫾ 16.6 85.8 ⫾ 6.7
Data are presented as mean ⫾ SD, median (interquartile range), or No. (%) unless otherwise indicated. 1266
The GerdQ in Predicting GERC On the basis of the highest Youden index of 0.584, the optimal cutoff point of the GerdQ for predicting GERC was defined at 8.0. The sensitivity, specificity, positive and negative predictive values, and area under the curve for this score are shown in Table 2 and Figure 3A. GERC was confirmed in 68 of 70 patients with a positive GerdQ score (97.1%), including 50 with acid GERC (73.5%) and 18 with nonacid GERC (26.5%). Additionally, GERC was verified in 34 of 56 patients with a negative GerdQ score (60.7%), including five with acid GERC (14.7%) and 29 with nonacid GERC (85.3%). Patients with acid GERC had a higher GerdQ score than those with nonacid GERC (10.63 ⫾ 2.06 vs 7.38 ⫾ 1.39, respectively; t 5 6.700; P 5 .000). Moreover, patients with GERC and a positive GerdQ score presented with more frequent heartburn, regurgitation, and use of additional medication (Table 3). A subanalysis for predicting acid GERC showed the same optimal cutoff point of 8.0; its sensitivity, negative predictive value, and the area under the curve increased, whereas its specificity and positive predictive value decreased (Fig 3B, Table 4). Fifty of 55 patients with acid GERC (90.9%) had a positive GerdQ score. There was a significant positive linear correlation between the GerdQ and DeMeester scores (r 5 0.793, P 5 .000) (Fig 4) but not between the GerdQ score and SAP for acid reflux (r 5 0.159, P 5 .296). When predicting nonacid GERC alone, the area under the curve of the GerdQ was 0.310, and no objective meaningful cutoff point could be determined. Comparison of The GerdQ and MII-pH in Predicting GERC The GerdQ had a lower sensitivity, negative predictive value, and k value than MII-pH but showed higher specificity and a comparable positive predictive value (Table 2). When used only for predicting acid GERC, the GerdQ still had a lower sensitivity, specificity, positive predictive value, Youden index, and k value, but its negative predictive value was similar to that of MII-pH (Table 4). For prediction of GERC, there was a 60.8% overlap between the GerdQ and MII-pH. Additionally, 30.4% and 5.9% of GERC cases were exclusively predicted by MII-pH and GerdQ, respectively. MII-pH and GerdQ failed to predict only 2.9% of GERC cases (Fig 5). Discussion Cough is not a typical symptom of GERD and can be caused by many other diseases. Generally, patients with obvious regurgitation and heartburn have a higher likelihood that their cough is due to reflux.16 Thus, GERC may be predicted by the GerdQ because it Original Research
Figure 2. A-C, Changes in cough symptom score over time in the patients with definite GERC responsive to antireflux therapy. A, GERC. B, Acid GERC. C, Nonacide GERC. See Figure 1 legend for expansion of abbreviation.
primarily and quantitatively evaluates the frequency and intensity of these symptoms.8 The optimal cutoff GerdQ score (8.0) for predicting GERC was consistent with that previously reported in patients with GERD,8 corresponding to a sensitivity of 66.7% and a specificity of 91.7%. When used only for acid GERC, the GerdQ demonstrated higher sensitivity but at the cost of lower specificity. Therefore, GerdQ can be used for screening acid GERC and selecting the appropriate patients for the appropriate treatment and can reduce the need for MII-pH and conserve medical resources. Proton pump inhibitors alone or in combination with prokinetic agents are considered to be the cornerstone of antireflux pharmacotherapy.1,5,6 However, there is controversy regarding their therapeutic efficacy for GERC.17-19 A recent systematic review favored using proton pump inhibitors for treating GERC if patients have pathogenic esophageal acid exposure.20 This recommendation is supported by the present data, showing that standard or augmented acid suppression therapy containing omeprazole resolved cough in most patients with GERC, with a slight delay of cough resolution in nonacid GERC compared with acid GERC (9.91 ⫾ 1.59 weeks vs 8.36 ⫾ 0.75 weeks, respectively; t 5 6.455; P 5 .000). The patients may represent a more rigor-
ously selected population with a higher likelihood of GERC as demonstrated by MII-pH. The difference in patient recruitment may help to explain the high success rate of antireflux therapy. Microaspiration and esophageal-tracheobronchial reflex are the main mechanisms underlying GERC.1,5 Proton pump inhibitors can reduce the acidity and volume of refluxate and even decrease reflux episodes when combined with prokinetic agents.21 Even for cough caused by weakly acidic reflux, which accounts for the majority of nonacid reflux,22 a mild increase in the pH value of refluxate might result in a significant blockade of acid signal transduction in a hypersensitive esophagus followed by cough resolution.23 However, proton pump inhibitors cannot modify transient lower esophageal sphincter relaxations, which are more frequently involved in GERC.24 Baclofen, a potent agonist of g-aminobutyric acid B receptor, may help to control cough because it inhibits both acid and nonacid reflux through rectifying transient lower esophageal sphincter relaxations.15,25 In addition, baclofen has a nonspecific antitussive activity and is used to treat refractory cough caused by diverse etiologies.26 In the present study, cough in 8.0% of patients with GERC resistant to acid suppression was resolved with baclofen
Table 2—Prediction of GERC Between GerdQ Score ⱖ 8 and MII-pH Item
GerdQ MII-pH x2 Value
P Value
Sensitivity, % Specificity, % Positive predictive value, % Negative predictive value, % Youden index k Value
66.7 91.7 97.1 42.9 0.584 0.386
.000 .000 .121 .000 … All , .01
97.1 62.5 91.7 83.3 0.596 0.659
30.488 24.987 2.405 34.320 … …
GERC 5 gastroesophageal reflux-induced chronic cough; GerdQ 5 Gastroesophageal Reflux Disease Questionnaire; MII-pH 5 multichannel intraluminal impedance pH monitoring. journal.publications.chestnet.org
Figure 3. A, B, Receiver operating curve of the GerdQ for predicting GERC. AUC, sensitivity, and 1-specificity by cutoff point of GerdQ score for GERC (A) and acid GERC (B). AUC 5 area under the curve. See Figure 1 legend for expansion of other abbreviations. CHEST / 145 / 6 / JUNE 2014
1267
Table 3—Distribution of Items in Patients With GERC With Positive and Negative GerdQ Scores Item
GerdQ ⱖ 8
GerdQ , 8
t Value
P Value
1. How often did you have a burning feeling behind your breastbone (heartburn)? 2. How often did you have stomach contents (liquid or food) moving upward to your throat or mouth (regurgitation)? 3. How often did you have a pain in the center of the upper stomach? 4. How often did you have nausea? 5. How often did you have difficulty getting a good night’s sleep because of your heartburn and/or regurgitation? 6. How often did you take additional medication for your heartburn and/or regurgitation other than what the physician told you to take (eg, Tums [GlaxoSmithKline], Rolaids [Chattem, Inc], Maalox [Novartis])?
1.33 ⫾ 0.23 1.95 ⫾ 0.34
0.09 ⫾ 0.06 0.17 ⫾ 0.08
6.832 8.691
.000 .000
2.98 ⫾ 0.42 2.93 ⫾ 0.44 0.12 ⫾ 0.05
2.91 ⫾ 0.26 2.91 ⫾ 0.26 0.09 ⫾ 0.06
1.154 0.221 0.393
.253 .826 .695
0.52 ⫾ 0.11
0.17 ⫾ 0.08
2.195
.032
Data are presented as mean ⫾ SD unless otherwise indicated. See Table 2 legend for expansion of abbreviations.
as an add-on therapy. Although we cannot separate its antireflux effect from its nonspecific antitussive activity in the study, baclofen can become a treatment option for GERC refractory to proton pump inhibitors. Reflux-related symptoms may be associated with the type of reflux involved. Most regurgitation and heartburn are caused by acid reflux,27 and nonacid reflux is responsible for only a minority of these symptoms in patients.28,29 This fact is further reinforced by the present findings that patients with GERC with positive GerdQ scores presented with more frequent heartburn and regurgitation, and GerdQ scores were significantly related to DeMeester scores in patients with acid GERC. This finding may explain why the GerdQ was a more accurate predictor for acid GERC. The GerdQ demonstrated a suboptimal sensitivity and specificity in predicting GERC, possibly reflecting the inherent imbalance between cough and the other reflux symptoms in GERC. Cough is usually the sole or predominant complaint of patients with GERC, and the other reflux-related symptoms may be rare or absent.30 When patients lack or have no obvious regurgitation and heartburn, GerdQ cannot predict GERC. The GerdQ correctly identified only 66.7% of GERC cases, which is consistent with two studies of the GerdQ in patients with GERD.14,31 Therefore, the GerdQ is imperfect for predicting GERC because a negative GerdQ score cannot exclude the possibility of GERC. MII-pH provides information about the temporal association between cough and acid or nonacid reflux.32
Not surprisingly, MII-pH is superior to the GerdQ for predicting both acid and nonacid GERC and cannot be completely replaced by the GerdQ. Nevertheless, MII-pH is far from perfect because it depends on SAP in the diagnosis of GERC, and SAP is not adequately reliable because patients usually underestimate the frequency or misreport the timing of cough events.33,34 Considering its invasive and expensive nature, MII-pH is currently less cost-effective than the GerdQ in the diagnosis of GERC. There are several limitations of this study. GERC was defined by favorable responsiveness to medicinal therapy, which might have led to an underdiagnosis of some GERC cases resistant to medicinal treatments but responsive to other therapies, such as antireflux surgery. The strict inclusion criteria possibly resulted in selection bias; therefore, the characteristics of the enrollees may not resemble those of a typical population with cough visiting a physician. Nevertheless, these patients are usually potential candidates for empirical antireflux therapy and need to be screened with the GerdQ before therapy. The GerdQ is initially designed
Table 4—Prediction of Acid GERC Between GerdQ Score ⱖ 8 and MII-pH Item
GerdQ MII-pH x2 Value P Value
Sensitivity, % Specificity, % Positive predictive value, % Negative predictive value, % Youden index k value
90.9 100.0 78.6 93.0 71.4 91.7 96.4 100.0 0.695 0.944 0.623 0.920
See Table 2 legend for expansion of abbreviations. 1268
9.424 8.140 13.227 4.082 … …
.002 .004 .000 .121 … All , .01
Figure 4. Relationship between GerdQ score and DeMeester score in patients with acid GERC analyzed by linear regression (r 5 0.793, P 5 .000). See Figure 1 legend for expansion of abbreviation. Original Research
References
Figure 5. Predicted proportion of GERC by GerdQ and MII-pH. See Figure 1 legend for expansion of abbreviations.
for GERD and lacks the items related to cough. However, addition of cough evaluation items to the GerdQ would not likely improve its predictive accuracy because the characteristics and timing of a cough do not provide clues about the cause of cough.35 Furthermore, the GerdQ is useless for nonacid GERC. However, even though it is imperfect, use of the GerdQ should be considered first to avoid overuse of empirical antireflux therapy. If a patient with a positive GerdQ score does not respond to consequent empirical antireflux therapy or presents with a negative GerdQ score, MII-pH must be performed to identify reflux as a cause of chronic cough. In conclusion, the GerdQ can be used for predicting acid GERC but not nonacid GERC. GerdQ is inferior to MII-pH.
Acknowledgments Author contributions: Dr Xu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Dr Xu: contributed to the case collection and processing, statistical analysis of data, interpretation of results, and writing of the manuscript. Dr Chen: contributed to the case collection and critical review of the manuscript. Dr Liang: contributed to the case collection and critical review of the manuscript. Dr Lv: contributed to the case collection and critical review of the manuscript. Dr Qiu: contributed to the study design, program coordination, and review and correction of the manuscript. Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Role of sponsors: The sponsors had no role in the study design, cases collection, data analysis and manuscript submission. The opinions expressed in this paper reflect those of the authors and do not represent those of The National Natural Science Foundation of China and Shanghai Shenkang Hospital Development Center. Other contributions: The authors thank AstraZeneca AB (Sweden) for permission to use a Chinese version of the GerdQ in this study. Additional information: The e-Appendixes can be found in the “Supplemental Materials” area of the online article. journal.publications.chestnet.org
1. Irwin RS. Chronic cough due to gastroesophageal reflux disease: ACCP evidence-based clinical practice guidelines. Chest. 2006;129(1_suppl):80S-94S. 2. Wei W, Yu L, Lü H, et al. Comparison of cause distribution between elderly and non-elderly patients with chronic cough. Respiration. 2009;77(3):259-264. 3. Qiu Z, Yu L, Xu S, et al. Cough reflex sensitivity and airway inflammation in patients with chronic cough due to non-acid gastro-oesophageal reflux. Respirology. 2011;16(4):645-652. 4. Kahrilas PJ, Sifrim D. High-resolution manometry and impedance-pH/manometry: valuable tools in clinical and investigational esophagology. Gastroenterology. 2008;135(3):756-769. 5. Morice AH, Fontana GA, Sovijarvi AR, et al; ERS Task Force. The diagnosis and management of chronic cough. Eur Respir J. 2004;24(3):481-492. 6. Morice AH, McGarvey L, Pavord I; British Thoracic Society Cough Guideline Group. Recommendations for the management of cough in adults. Thorax. 2006;61(suppl 1):i1-i24. 7. Poe RH, Kallay MC. Chronic cough and gastroesophageal reflux disease: experience with specific therapy for diagnosis and treatment. Chest. 2003;123(3):679-684. 8. Jones R, Junghard O, Dent J, et al. Development of the GerdQ, a tool for the diagnosis and management of gastrooesophageal reflux disease in primary care. Aliment Pharmacol Ther. 2009;30(10):1030-1038. 9. Lacy BE, Chehade R, Crowell MD. A prospective study to compare a symptom-based reflux disease questionnaire to 48-h wireless pH monitoring for the identification of gastroesophageal reflux (revised 2-26-11). Am J Gastroenterol. 2011; 106(9):1604-1611. 10. Asthma Workgroup, Chinese Society of Respiratory Diseases (CSRD), Chinese Medical Association. The Chinese national guidelines on diagnosis and management of cough (December 2010). Chin Med J (Engl). 2011;124(20):3207-3219. 11. Hsu JY, Stone RA, Logan-Sinclair RB, Worsdell M, Busst CM, Chung KF. Coughing frequency in patients with persistent cough: assessment using a 24 hour ambulatory recorder. Eur Respir J. 1994;7(7):1246-1253. 12. Chinese GerdQ Study Group. Evaluation of gastroesophageal reflux disease questionnaire in diagnosis of gastroesophageal reflux disease [in Chinese]. Chin J Dig. 2009;29(12):793-798. 13. Bai Y, Du Y, Zou D, et al; Chinese GerdQ Research Group. Gastroesophageal Reflux Disease Questionnaire (GerdQ) in real-world practice: a national multicenter survey on 8065 patients. J Gastroenterol Hepatol. 2013;28(4):626-631. 14. Blondeau K, Dupont LJ, Mertens V, Tack J, Sifrim D. Improved diagnosis of gastro-oesophageal reflux in patients with unexplained chronic cough. Aliment Pharmacol Ther. 2007;25(6): 723-732. 15. Xu X, Chen Q, Liang S, Lü H, Qiu Z. Successful resolution of refractory chronic cough induced by gastroesophageal reflux with treatment of baclofen. Cough. 2012;8(1):8. 16. Pratter MR, Brightling CE, Boulet LP, Irwin RS. An empiric integrative approach to the management of cough: ACCP evidence-based clinical practice guidelines. Chest. 2006; 129(1_suppl):222S-231S. 17. Shaheen NJ, Crockett SD, Bright SD, et al. Randomised clinical trial: high-dose acid suppression for chronic cough - a double-blind, placebo-controlled study. Aliment Pharmacol Ther. 2011;33(2):225-234. 18. Faruqi S, Molyneux ID, Fathi H, Wright C, Thompson R, Morice AH. Chronic cough and esomeprazole: a doubleblind placebo-controlled parallel study. Respirology. 2011; 16(7):1150-1156. 19. Chang AB, Lasserson TJ, Gaffney J, Connor FL, Garske LA. Gastro-oesophageal reflux treatment for prolonged non-specific CHEST / 145 / 6 / JUNE 2014
1269
20.
21. 22. 23. 24. 25.
26. 27.
cough in children and adults. Cochrane Database Syst Rev. 2011;(1):CD004823. Kahrilas PJ, Howden CW, Hughes N, Molloy-Bland M. Response of chronic cough to acid-suppressive therapy in patients with gastroesophageal reflux disease. Chest. 2013; 143(3):605-612. Dellon ES, Shaheen NJ. Persistent reflux symptoms in the proton pump inhibitor era: the changing face of gastroesophageal reflux disease. Gastroenterology. 2010;139(1):7-13. Boeckxstaens GE, Smout A. Systematic review: role of acid, weakly acidic and weakly alkaline reflux in gastro-oesophageal reflux disease. Aliment Pharmacol Ther. 2010;32(3):334-343. Sifrim D, Zerbib F. Diagnosis and management of patients with reflux symptoms refractory to proton pump inhibitors. Gut. 2012;61(9):1340-1354. Hershcovici T, Mashimo H, Fass R. The lower esophageal sphincter. Neurogastroenterol Motil. 2011;23(9):819-830. Zhang Q, Lehmann A, Rigda R, Dent J, Holloway RH. Control of transient lower oesophageal sphincter relaxations and reflux by the GABA(B) agonist baclofen in patients with gastro-oesophageal reflux disease. Gut. 2002;50(1):19-24. Dicpinigaitis PV, Dobkin JB. Antitussive effect of the GABAagonist baclofen. Chest. 1997;111(4):996-999. Vela MF, Camacho-Lobato L, Srinivasan R, Tutuian R, Katz PO, Castell DO. Simultaneous intraesophageal impedance and pH measurement of acid and nonacid gastroesophageal reflux: effect of omeprazole. Gastroenterology. 2001;120(7):1599-1606.
1270
28. Bredenoord AJ, Weusten BL, Curvers WL, Timmer R, Smout AJ. Determinants of perception of heartburn and regurgitation. Gut. 2006;55(3):313-318. 29. Zerbib F, Duriez A, Roman S, Capdepont M, Mion F. Determinants of gastro-oesophageal reflux perception in patients with persistent symptoms despite proton pump inhibitors. Gut. 2008;57(2):156-160. 30. Irwin RS, French CL, Curley FJ, Zawacki JK, Bennett FM. Chronic cough due to gastroesophageal reflux. Clinical, diagnostic, and pathogenetic aspects. Chest. 1993;104(5):1511-1517. 31. Jonasson C, Wernersson B, Hoff DA, Hatlebakk JG. Validation of the GerdQ questionnaire for the diagnosis of gastrooesophageal reflux disease. Aliment Pharmacol Ther. 2013;37(5): 564-572. 32. Bredenoord AJ. Impedance-pH monitoring: new standard for measuring gastro-oesophageal reflux. Neurogastroenterol Motil. 2008;20(5):434-439. 33. Sifrim D, Dupont L, Blondeau K, Zhang X, Tack J, Janssens J. Weakly acidic reflux in patients with chronic unexplained cough during 24 hour pressure, pH, and impedance monitoring. Gut. 2005;54(4):449-454. 34. Kavitt RT, Higginbotham T, Slaughter JC, et al. Symptom reports are not reliable during ambulatory reflux monitoring. Am J Gastroenterol. 2012;107(12):1826-1832. 35. Mello CJ, Irwin RS, Curley FJ. Predictive values of the character, timing, and complications of chronic cough in diagnosing its cause. Arch Intern Med. 1996;156(9):997-1003.
Original Research
Original Research SIGNS AND SYMPTOMS OF CHEST DISEASE
Comparison of Gastroesophageal Reflux Disease Questionnaire and Multichannel Intraluminal Impedance pH Monitoring in Identifying Patients With Chronic Cough Responsive to Antireflux Therapy Xianghuai Xu, MD; Qiang Chen, MD; Siwei Liang, MD; Hanjing Lv, MD; and Zhongmin Qiu, MD, FCCP
Background: Empirical therapy has been recommended as an initial clinical approach for treating gastroesophageal reflux-induced chronic cough (GERC). This study compared the predictive accuracy of the Gastroesophageal Reflux Disease Questionnaire (GerdQ) with the accuracy of multichannel intraluminal impedance pH monitoring (MII-pH) for GERC. Methods: A total of 126 consecutive patients with potential GERC were recruited to undergo MII-pH and complete the GerdQ. A final diagnosis of GERC was made after favorable response to consequent medicinal antireflux therapy, regardless of laboratory findings. The predictive accuracy of the GerdQ for GERC was assessed and compared with that of MII-pH. Results: GERC was confirmed in 102 of 126 patients (81.0%); cough was due to acid reflux in 55 (53.9%) and nonacid reflux in 47 (46.1%). The optimal cutoff point of the GerdQ for predicting GERC was defined as 8.0 according to the highest Youden index of 0.584, with a sensitivity of 66.7%, specificity of 91.7%, positive predictive value of 97.1%, and negative predictive value of 42.9%. A subanalysis for only acid GERC showed further improvement in the predictive accuracy of the GerdQ, corresponding to a sensitivity of 90.9%, specificity of 78.6%, positive predictive value of 71.4%, and negative predictive value of 96.4%. However, a meaningful GerdQ cutoff point for prediction of nonacid GERC could not be determined. In general, MII-pH was superior to the GerdQ for predicting GERC and acid GERC. Conclusions: The GerdQ can be used for predicting acid GERC but not nonacid GERC and is inferior to MII-pH. Trial registry: Chinese Clinical Trial Registry; No.: ChiCTR-ODT-12001899; URL: www.chictr.org CHEST 2014; 145(6):1264–1270 Abbreviations: GERC 5 gastroesophageal reflux-induced chronic cough; GERD 5 gastroesophageal reflux disease; GerdQ 5 Gastroesophageal Reflux Disease Questionnaire; MII-pH 5 multichannel intraluminal impedance pH monitoring; SAP 5 symptom association probability
reflux is a common cause of chronic Gastroesophageal cough. Based on pH values of refluxate in the 1,2
esophagus, gastroesophageal reflux-induced chronic cough (GERC) can be classified as acid or nonacid.3 Multichannel intraluminal impedance pH monitoring (MII-pH) can detect both acid and nonacid reflux4 and is the most sensitive and specific test for diagnosing GERC. Unfortunately, MII-pH has not been extensively used clinically because of its invasiveness and limited availability. Alternatively, empirical therapy has been advocated as an initial approach for treating
1264
GERC.1,5,6 Because empirical trials frequently fail to relieve patients’ cough,7 screening patients with potential GERC prior to antireflux treatment may improve therapeutic gains. The Gastroesophageal Reflux Disease Questionnaire (GerdQ) is a patient-centered and self-reported diagnostic instrument for gastroesophageal reflux disease (GERD)8 and can achieve an overall diagnostic accuracy similar to that made by a gastroenterologist, with a sensitivity of 64.6% and a specificity of 71.4%.8,9 However, to our knowledge the usefulness of the GerdQ in Original Research
GERC diagnosis has never been addressed. Therefore, we conducted a prospective study to compare the predictive accuracies of the GerdQ and MII-pH for GERC. Materials and Methods Patients Consecutive patients referred to our respiratory clinic for potential GERC between January 2011 and March 2013 were recruited for this study. Requirements for participation included any one of the following: (1) presence of chronic cough and typical refluxrelated symptoms, (2) GERC considered when previous laboratory workups failed to identify other possible causes of chronic cough,10 and (3) coexisting GERC suspected when the treatment of current etiologies failed to completely resolve a cough. Patients refusing to undergo or who were intolerant of MII-pH were excluded. The study protocol was approved by the ethics committee of Tongji Hospital [No. LL(H)-11-13] and registered with the Chinese Clinical Trials Register. Written informed consent was obtained from all subjects before enrollment. Study Design This was a single-center observational study. Initial patient assessments included a medical history, physical examination, plain chest radiography, lung function tests, and evaluation of cough severity as determined by validated cough symptom scores that rate a cough using six incremental scales (0 5 no cough, 5 5 worst cough)11 (e-Appendix 1). Additionally, patients were evaluated with the previously validated Chinese version of the GerdQ12,13 (e-Appendix 2). The GerdQ comprises six symptom-related items consisting of four reflux-related items positively related to GERD and two items negatively related to GERD. Patients were asked to recall how often they experienced the events described in the questionnaire during the preceding week and to score their answers on a four-point scale ranging from 0 to 3 for positive predictors and from 3 to 0 for negative predictors. The total scores represented a GerdQ score ranging from 0 to 18. A higher score signifies a greater possibility of GERD.8 MII-pH was performed as previously described3 after the patients had stopped taking acid-suppressing medication for at least 1 week. Briefly, a combined MII-pH catheter was inserted transnasally into the patient’s esophagus, with six impedance channel sensors (K6011-E10632; Unisensor AG) located 3, 5, 7, 9, 15, and 17 cm above the lower esophageal sphincter and an antimony pH electrode (819100; Medical Measurement System BV) positioned 5 cm Manuscript received July 16, 2013; revision accepted December 19, 2013; originally published Online First January 23, 2014. Affiliations: From the Department of Respiratory Medicine, Tongji Hospital, Tongji University School of Medicine, Shanghai, China. Part of this article was presented orally at the Seventh International Cough Symposium 2012: Taming of Chronic Cough, July 5-7, 2012, London, England. Funding/Support: This study was supported by grants from the National Natural Science Foundation of China [81170079] and Shanghai Shenkang Hospital Development Center [SHDC12012211]. Correspondence to: Zhongmin Qiu, MD, FCCP, Department of Respiratory Medicine, Tongji Hospital, Tongji University School of Medicine, No. 389 Xincun Rd, Shanghai 200065, China; e-mail: [email protected] © 2014 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details. DOI: 10.1378/chest.13-1634 journal.publications.chestnet.org
above the proximal border of the lower esophageal sphincter. A connected portable data logger (Ohmega; Medical Measurement System BV) stored data from all seven channels over 24 h. Reflux episodes recorded on the tracings of MII-pH were manually characterized by their impedance value as liquid, gas, or mixed liquidgas reflux or characterized using a pH meter as acidic (pH , 4.0), weakly acidic (pH 4.0-7.0), or weakly alkaline (pH . 7.0) reflux, with the latter two collectively referred to as nonacid reflux. The DeMeester score was calculated as a global measure of esophageal acid exposure. Symptom association probability (SAP) was used to represent the temporal association between cough recorded by patients on diary cards and reflux that had occurred during the preceding 2-min period.3,14 Abnormal acid reflux was defined as a DeMeester score ⱖ 14.72, SAP for acid reflux ⱖ 95%, or both. These scores were used for diagnosing acid GERC, whereas abnormal nonacid reflux was defined as an SAP for nonacid reflux ⱖ 95%, which was used for diagnosing nonacid GERC.3 Regardless of MII-pH findings, all patients received an 8-week course of standard antireflux therapy consisting of omeprazole 20 mg bid and domperidone 10 mg tid. If the initial treatment failed and MII-pH revealed abnormal reflux, patients received an augmented antireflux trial by doubling the dose of omeprazole to 80 mg/d or replacing domperidone with baclofen 20 mg tid.15 All study patients and investigators were blinded to GerdQ scores throughout treatment. GERC was finally diagnosed only when a cough disappeared (resolved completely) or improved (combined daytime and nighttime cough symptom score decreased by ⱖ 2 points)3 as recommended by American College of Chest Physicians guidelines.1 The predictive accuracies of the GerdQ and MII-pH for GERC were analyzed and compared (Fig 1). Statistical Analysis Data with normal distributions are expressed as mean ⫾ SD, whereas data with skewed distributions are expressed as median
Figure 1. CONSORT (Consolidated Standards of Reporting Trials) flow diagram of the study and algorithm for GERC. GERC 5 gastroesophageal reflux-induced chronic cough; GerdQ 5 Gastroesophageal Reflux Disease Questionnaire; MII-pH 5 multichannel intraluminal impedance pH monitoring. CHEST / 145 / 6 / JUNE 2014
1265
values (interquartile range). Unpaired Student t and Mann-Whitney U tests were used for data comparison where applicable. Receiver operating characteristic analysis was used to evaluate the predictive value of the GerdQ. The optimal cutoff point was determined on the basis of the highest Youden index. The relation between the GerdQ score and DeMeester score or SAP for acid reflux was analyzed by linear regression. Predictive values of the GerdQ and MII-pH were compared with the x2 test. Diagnostic consistency was evaluated by k test. Statistical calculations were performed with SPSS, version 17.0 (IBM) software.
Results Demographic Characteristics A total of 167 of 583 patients with chronic cough (28.6%) met the inclusion criteria for this study. Among this group of eligible patients, 126 (76%) completed evaluations by MII-pH and the GerdQ. Demographic characteristics of the final study group are shown in Table 1. Abnormal reflux was revealed by MII-pH in 108 of 126 patients (85.7%), and GERC was confirmed by empirical antireflux treatment in 80 patients (74.1%). Twenty-eight patients (25.9%) received augmented acid suppression treatment with omeprazole 80 mg/d, and GERC was confirmed in 11 patients (39.3%). The remaining 17 patients (15.7%) accepted therapy with baclofen, and GERC was further confirmed in eight (47.1%). Among nine nonresponders, cough in seven (77.8%) could be explained by upper-airway cough syndrome in two (22.2%) and cough variant asthma in five (55.6%), but the causes were unknown in two (22.2%). In 18 patients (14.3%) with a negative DeMeester score and SAP, nonacid GERC was determined in three patients (16.7%) on the basis of increased weakly acidic reflux events and favorable response to antireflux treatment. In total, GERC was diagnosed in 102 patients (81.0%), including 55 with acid GERC (53.9%) and 47 with nonacid GERC (46.1%). Additionally, the cough was completely resolved in 38 patients (69.1%) with acid GERC and 25 (53.2%) with nonacid GERC. A plot of patient response vs treatment time is shown in Figure 2. Table 1—Patient Demographic Characteristics Characteristic Male (female) sex Age, y Cough duration, mo Cough symptom score Daytime Nighttime Absence of heartburn or regurgitation FEV1 % predicted FVC % predicted FEV1/FVC, %
Value 59 (67) 45.6 ⫾ 18.3 8.5 (5.3-36.0) 3 (2-4) 1 (1-3) 49 (38.9) 95.3 ⫾ 12.1 97.2 ⫾ 16.6 85.8 ⫾ 6.7
Data are presented as mean ⫾ SD, median (interquartile range), or No. (%) unless otherwise indicated. 1266
The GerdQ in Predicting GERC On the basis of the highest Youden index of 0.584, the optimal cutoff point of the GerdQ for predicting GERC was defined at 8.0. The sensitivity, specificity, positive and negative predictive values, and area under the curve for this score are shown in Table 2 and Figure 3A. GERC was confirmed in 68 of 70 patients with a positive GerdQ score (97.1%), including 50 with acid GERC (73.5%) and 18 with nonacid GERC (26.5%). Additionally, GERC was verified in 34 of 56 patients with a negative GerdQ score (60.7%), including five with acid GERC (14.7%) and 29 with nonacid GERC (85.3%). Patients with acid GERC had a higher GerdQ score than those with nonacid GERC (10.63 ⫾ 2.06 vs 7.38 ⫾ 1.39, respectively; t 5 6.700; P 5 .000). Moreover, patients with GERC and a positive GerdQ score presented with more frequent heartburn, regurgitation, and use of additional medication (Table 3). A subanalysis for predicting acid GERC showed the same optimal cutoff point of 8.0; its sensitivity, negative predictive value, and the area under the curve increased, whereas its specificity and positive predictive value decreased (Fig 3B, Table 4). Fifty of 55 patients with acid GERC (90.9%) had a positive GerdQ score. There was a significant positive linear correlation between the GerdQ and DeMeester scores (r 5 0.793, P 5 .000) (Fig 4) but not between the GerdQ score and SAP for acid reflux (r 5 0.159, P 5 .296). When predicting nonacid GERC alone, the area under the curve of the GerdQ was 0.310, and no objective meaningful cutoff point could be determined. Comparison of The GerdQ and MII-pH in Predicting GERC The GerdQ had a lower sensitivity, negative predictive value, and k value than MII-pH but showed higher specificity and a comparable positive predictive value (Table 2). When used only for predicting acid GERC, the GerdQ still had a lower sensitivity, specificity, positive predictive value, Youden index, and k value, but its negative predictive value was similar to that of MII-pH (Table 4). For prediction of GERC, there was a 60.8% overlap between the GerdQ and MII-pH. Additionally, 30.4% and 5.9% of GERC cases were exclusively predicted by MII-pH and GerdQ, respectively. MII-pH and GerdQ failed to predict only 2.9% of GERC cases (Fig 5). Discussion Cough is not a typical symptom of GERD and can be caused by many other diseases. Generally, patients with obvious regurgitation and heartburn have a higher likelihood that their cough is due to reflux.16 Thus, GERC may be predicted by the GerdQ because it Original Research
Figure 2. A-C, Changes in cough symptom score over time in the patients with definite GERC responsive to antireflux therapy. A, GERC. B, Acid GERC. C, Nonacide GERC. See Figure 1 legend for expansion of abbreviation.
primarily and quantitatively evaluates the frequency and intensity of these symptoms.8 The optimal cutoff GerdQ score (8.0) for predicting GERC was consistent with that previously reported in patients with GERD,8 corresponding to a sensitivity of 66.7% and a specificity of 91.7%. When used only for acid GERC, the GerdQ demonstrated higher sensitivity but at the cost of lower specificity. Therefore, GerdQ can be used for screening acid GERC and selecting the appropriate patients for the appropriate treatment and can reduce the need for MII-pH and conserve medical resources. Proton pump inhibitors alone or in combination with prokinetic agents are considered to be the cornerstone of antireflux pharmacotherapy.1,5,6 However, there is controversy regarding their therapeutic efficacy for GERC.17-19 A recent systematic review favored using proton pump inhibitors for treating GERC if patients have pathogenic esophageal acid exposure.20 This recommendation is supported by the present data, showing that standard or augmented acid suppression therapy containing omeprazole resolved cough in most patients with GERC, with a slight delay of cough resolution in nonacid GERC compared with acid GERC (9.91 ⫾ 1.59 weeks vs 8.36 ⫾ 0.75 weeks, respectively; t 5 6.455; P 5 .000). The patients may represent a more rigor-
ously selected population with a higher likelihood of GERC as demonstrated by MII-pH. The difference in patient recruitment may help to explain the high success rate of antireflux therapy. Microaspiration and esophageal-tracheobronchial reflex are the main mechanisms underlying GERC.1,5 Proton pump inhibitors can reduce the acidity and volume of refluxate and even decrease reflux episodes when combined with prokinetic agents.21 Even for cough caused by weakly acidic reflux, which accounts for the majority of nonacid reflux,22 a mild increase in the pH value of refluxate might result in a significant blockade of acid signal transduction in a hypersensitive esophagus followed by cough resolution.23 However, proton pump inhibitors cannot modify transient lower esophageal sphincter relaxations, which are more frequently involved in GERC.24 Baclofen, a potent agonist of g-aminobutyric acid B receptor, may help to control cough because it inhibits both acid and nonacid reflux through rectifying transient lower esophageal sphincter relaxations.15,25 In addition, baclofen has a nonspecific antitussive activity and is used to treat refractory cough caused by diverse etiologies.26 In the present study, cough in 8.0% of patients with GERC resistant to acid suppression was resolved with baclofen
Table 2—Prediction of GERC Between GerdQ Score ⱖ 8 and MII-pH Item
GerdQ MII-pH x2 Value
P Value
Sensitivity, % Specificity, % Positive predictive value, % Negative predictive value, % Youden index k Value
66.7 91.7 97.1 42.9 0.584 0.386
.000 .000 .121 .000 … All , .01
97.1 62.5 91.7 83.3 0.596 0.659
30.488 24.987 2.405 34.320 … …
GERC 5 gastroesophageal reflux-induced chronic cough; GerdQ 5 Gastroesophageal Reflux Disease Questionnaire; MII-pH 5 multichannel intraluminal impedance pH monitoring. journal.publications.chestnet.org
Figure 3. A, B, Receiver operating curve of the GerdQ for predicting GERC. AUC, sensitivity, and 1-specificity by cutoff point of GerdQ score for GERC (A) and acid GERC (B). AUC 5 area under the curve. See Figure 1 legend for expansion of other abbreviations. CHEST / 145 / 6 / JUNE 2014
1267
Table 3—Distribution of Items in Patients With GERC With Positive and Negative GerdQ Scores Item
GerdQ ⱖ 8
GerdQ , 8
t Value
P Value
1. How often did you have a burning feeling behind your breastbone (heartburn)? 2. How often did you have stomach contents (liquid or food) moving upward to your throat or mouth (regurgitation)? 3. How often did you have a pain in the center of the upper stomach? 4. How often did you have nausea? 5. How often did you have difficulty getting a good night’s sleep because of your heartburn and/or regurgitation? 6. How often did you take additional medication for your heartburn and/or regurgitation other than what the physician told you to take (eg, Tums [GlaxoSmithKline], Rolaids [Chattem, Inc], Maalox [Novartis])?
1.33 ⫾ 0.23 1.95 ⫾ 0.34
0.09 ⫾ 0.06 0.17 ⫾ 0.08
6.832 8.691
.000 .000
2.98 ⫾ 0.42 2.93 ⫾ 0.44 0.12 ⫾ 0.05
2.91 ⫾ 0.26 2.91 ⫾ 0.26 0.09 ⫾ 0.06
1.154 0.221 0.393
.253 .826 .695
0.52 ⫾ 0.11
0.17 ⫾ 0.08
2.195
.032
Data are presented as mean ⫾ SD unless otherwise indicated. See Table 2 legend for expansion of abbreviations.
as an add-on therapy. Although we cannot separate its antireflux effect from its nonspecific antitussive activity in the study, baclofen can become a treatment option for GERC refractory to proton pump inhibitors. Reflux-related symptoms may be associated with the type of reflux involved. Most regurgitation and heartburn are caused by acid reflux,27 and nonacid reflux is responsible for only a minority of these symptoms in patients.28,29 This fact is further reinforced by the present findings that patients with GERC with positive GerdQ scores presented with more frequent heartburn and regurgitation, and GerdQ scores were significantly related to DeMeester scores in patients with acid GERC. This finding may explain why the GerdQ was a more accurate predictor for acid GERC. The GerdQ demonstrated a suboptimal sensitivity and specificity in predicting GERC, possibly reflecting the inherent imbalance between cough and the other reflux symptoms in GERC. Cough is usually the sole or predominant complaint of patients with GERC, and the other reflux-related symptoms may be rare or absent.30 When patients lack or have no obvious regurgitation and heartburn, GerdQ cannot predict GERC. The GerdQ correctly identified only 66.7% of GERC cases, which is consistent with two studies of the GerdQ in patients with GERD.14,31 Therefore, the GerdQ is imperfect for predicting GERC because a negative GerdQ score cannot exclude the possibility of GERC. MII-pH provides information about the temporal association between cough and acid or nonacid reflux.32
Not surprisingly, MII-pH is superior to the GerdQ for predicting both acid and nonacid GERC and cannot be completely replaced by the GerdQ. Nevertheless, MII-pH is far from perfect because it depends on SAP in the diagnosis of GERC, and SAP is not adequately reliable because patients usually underestimate the frequency or misreport the timing of cough events.33,34 Considering its invasive and expensive nature, MII-pH is currently less cost-effective than the GerdQ in the diagnosis of GERC. There are several limitations of this study. GERC was defined by favorable responsiveness to medicinal therapy, which might have led to an underdiagnosis of some GERC cases resistant to medicinal treatments but responsive to other therapies, such as antireflux surgery. The strict inclusion criteria possibly resulted in selection bias; therefore, the characteristics of the enrollees may not resemble those of a typical population with cough visiting a physician. Nevertheless, these patients are usually potential candidates for empirical antireflux therapy and need to be screened with the GerdQ before therapy. The GerdQ is initially designed
Table 4—Prediction of Acid GERC Between GerdQ Score ⱖ 8 and MII-pH Item
GerdQ MII-pH x2 Value P Value
Sensitivity, % Specificity, % Positive predictive value, % Negative predictive value, % Youden index k value
90.9 100.0 78.6 93.0 71.4 91.7 96.4 100.0 0.695 0.944 0.623 0.920
See Table 2 legend for expansion of abbreviations. 1268
9.424 8.140 13.227 4.082 … …
.002 .004 .000 .121 … All , .01
Figure 4. Relationship between GerdQ score and DeMeester score in patients with acid GERC analyzed by linear regression (r 5 0.793, P 5 .000). See Figure 1 legend for expansion of abbreviation. Original Research
References
Figure 5. Predicted proportion of GERC by GerdQ and MII-pH. See Figure 1 legend for expansion of abbreviations.
for GERD and lacks the items related to cough. However, addition of cough evaluation items to the GerdQ would not likely improve its predictive accuracy because the characteristics and timing of a cough do not provide clues about the cause of cough.35 Furthermore, the GerdQ is useless for nonacid GERC. However, even though it is imperfect, use of the GerdQ should be considered first to avoid overuse of empirical antireflux therapy. If a patient with a positive GerdQ score does not respond to consequent empirical antireflux therapy or presents with a negative GerdQ score, MII-pH must be performed to identify reflux as a cause of chronic cough. In conclusion, the GerdQ can be used for predicting acid GERC but not nonacid GERC. GerdQ is inferior to MII-pH.
Acknowledgments Author contributions: Dr Xu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Dr Xu: contributed to the case collection and processing, statistical analysis of data, interpretation of results, and writing of the manuscript. Dr Chen: contributed to the case collection and critical review of the manuscript. Dr Liang: contributed to the case collection and critical review of the manuscript. Dr Lv: contributed to the case collection and critical review of the manuscript. Dr Qiu: contributed to the study design, program coordination, and review and correction of the manuscript. Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Role of sponsors: The sponsors had no role in the study design, cases collection, data analysis and manuscript submission. The opinions expressed in this paper reflect those of the authors and do not represent those of The National Natural Science Foundation of China and Shanghai Shenkang Hospital Development Center. Other contributions: The authors thank AstraZeneca AB (Sweden) for permission to use a Chinese version of the GerdQ in this study. Additional information: The e-Appendixes can be found in the “Supplemental Materials” area of the online article. journal.publications.chestnet.org
1. Irwin RS. Chronic cough due to gastroesophageal reflux disease: ACCP evidence-based clinical practice guidelines. Chest. 2006;129(1_suppl):80S-94S. 2. Wei W, Yu L, Lü H, et al. Comparison of cause distribution between elderly and non-elderly patients with chronic cough. Respiration. 2009;77(3):259-264. 3. Qiu Z, Yu L, Xu S, et al. Cough reflex sensitivity and airway inflammation in patients with chronic cough due to non-acid gastro-oesophageal reflux. Respirology. 2011;16(4):645-652. 4. Kahrilas PJ, Sifrim D. High-resolution manometry and impedance-pH/manometry: valuable tools in clinical and investigational esophagology. Gastroenterology. 2008;135(3):756-769. 5. Morice AH, Fontana GA, Sovijarvi AR, et al; ERS Task Force. The diagnosis and management of chronic cough. Eur Respir J. 2004;24(3):481-492. 6. Morice AH, McGarvey L, Pavord I; British Thoracic Society Cough Guideline Group. Recommendations for the management of cough in adults. Thorax. 2006;61(suppl 1):i1-i24. 7. Poe RH, Kallay MC. Chronic cough and gastroesophageal reflux disease: experience with specific therapy for diagnosis and treatment. Chest. 2003;123(3):679-684. 8. Jones R, Junghard O, Dent J, et al. Development of the GerdQ, a tool for the diagnosis and management of gastrooesophageal reflux disease in primary care. Aliment Pharmacol Ther. 2009;30(10):1030-1038. 9. Lacy BE, Chehade R, Crowell MD. A prospective study to compare a symptom-based reflux disease questionnaire to 48-h wireless pH monitoring for the identification of gastroesophageal reflux (revised 2-26-11). Am J Gastroenterol. 2011; 106(9):1604-1611. 10. Asthma Workgroup, Chinese Society of Respiratory Diseases (CSRD), Chinese Medical Association. The Chinese national guidelines on diagnosis and management of cough (December 2010). Chin Med J (Engl). 2011;124(20):3207-3219. 11. Hsu JY, Stone RA, Logan-Sinclair RB, Worsdell M, Busst CM, Chung KF. Coughing frequency in patients with persistent cough: assessment using a 24 hour ambulatory recorder. Eur Respir J. 1994;7(7):1246-1253. 12. Chinese GerdQ Study Group. Evaluation of gastroesophageal reflux disease questionnaire in diagnosis of gastroesophageal reflux disease [in Chinese]. Chin J Dig. 2009;29(12):793-798. 13. Bai Y, Du Y, Zou D, et al; Chinese GerdQ Research Group. Gastroesophageal Reflux Disease Questionnaire (GerdQ) in real-world practice: a national multicenter survey on 8065 patients. J Gastroenterol Hepatol. 2013;28(4):626-631. 14. Blondeau K, Dupont LJ, Mertens V, Tack J, Sifrim D. Improved diagnosis of gastro-oesophageal reflux in patients with unexplained chronic cough. Aliment Pharmacol Ther. 2007;25(6): 723-732. 15. Xu X, Chen Q, Liang S, Lü H, Qiu Z. Successful resolution of refractory chronic cough induced by gastroesophageal reflux with treatment of baclofen. Cough. 2012;8(1):8. 16. Pratter MR, Brightling CE, Boulet LP, Irwin RS. An empiric integrative approach to the management of cough: ACCP evidence-based clinical practice guidelines. Chest. 2006; 129(1_suppl):222S-231S. 17. Shaheen NJ, Crockett SD, Bright SD, et al. Randomised clinical trial: high-dose acid suppression for chronic cough - a double-blind, placebo-controlled study. Aliment Pharmacol Ther. 2011;33(2):225-234. 18. Faruqi S, Molyneux ID, Fathi H, Wright C, Thompson R, Morice AH. Chronic cough and esomeprazole: a doubleblind placebo-controlled parallel study. Respirology. 2011; 16(7):1150-1156. 19. Chang AB, Lasserson TJ, Gaffney J, Connor FL, Garske LA. Gastro-oesophageal reflux treatment for prolonged non-specific CHEST / 145 / 6 / JUNE 2014
1269
20.
21. 22. 23. 24. 25.
26. 27.
cough in children and adults. Cochrane Database Syst Rev. 2011;(1):CD004823. Kahrilas PJ, Howden CW, Hughes N, Molloy-Bland M. Response of chronic cough to acid-suppressive therapy in patients with gastroesophageal reflux disease. Chest. 2013; 143(3):605-612. Dellon ES, Shaheen NJ. Persistent reflux symptoms in the proton pump inhibitor era: the changing face of gastroesophageal reflux disease. Gastroenterology. 2010;139(1):7-13. Boeckxstaens GE, Smout A. Systematic review: role of acid, weakly acidic and weakly alkaline reflux in gastro-oesophageal reflux disease. Aliment Pharmacol Ther. 2010;32(3):334-343. Sifrim D, Zerbib F. Diagnosis and management of patients with reflux symptoms refractory to proton pump inhibitors. Gut. 2012;61(9):1340-1354. Hershcovici T, Mashimo H, Fass R. The lower esophageal sphincter. Neurogastroenterol Motil. 2011;23(9):819-830. Zhang Q, Lehmann A, Rigda R, Dent J, Holloway RH. Control of transient lower oesophageal sphincter relaxations and reflux by the GABA(B) agonist baclofen in patients with gastro-oesophageal reflux disease. Gut. 2002;50(1):19-24. Dicpinigaitis PV, Dobkin JB. Antitussive effect of the GABAagonist baclofen. Chest. 1997;111(4):996-999. Vela MF, Camacho-Lobato L, Srinivasan R, Tutuian R, Katz PO, Castell DO. Simultaneous intraesophageal impedance and pH measurement of acid and nonacid gastroesophageal reflux: effect of omeprazole. Gastroenterology. 2001;120(7):1599-1606.
1270
28. Bredenoord AJ, Weusten BL, Curvers WL, Timmer R, Smout AJ. Determinants of perception of heartburn and regurgitation. Gut. 2006;55(3):313-318. 29. Zerbib F, Duriez A, Roman S, Capdepont M, Mion F. Determinants of gastro-oesophageal reflux perception in patients with persistent symptoms despite proton pump inhibitors. Gut. 2008;57(2):156-160. 30. Irwin RS, French CL, Curley FJ, Zawacki JK, Bennett FM. Chronic cough due to gastroesophageal reflux. Clinical, diagnostic, and pathogenetic aspects. Chest. 1993;104(5):1511-1517. 31. Jonasson C, Wernersson B, Hoff DA, Hatlebakk JG. Validation of the GerdQ questionnaire for the diagnosis of gastrooesophageal reflux disease. Aliment Pharmacol Ther. 2013;37(5): 564-572. 32. Bredenoord AJ. Impedance-pH monitoring: new standard for measuring gastro-oesophageal reflux. Neurogastroenterol Motil. 2008;20(5):434-439. 33. Sifrim D, Dupont L, Blondeau K, Zhang X, Tack J, Janssens J. Weakly acidic reflux in patients with chronic unexplained cough during 24 hour pressure, pH, and impedance monitoring. Gut. 2005;54(4):449-454. 34. Kavitt RT, Higginbotham T, Slaughter JC, et al. Symptom reports are not reliable during ambulatory reflux monitoring. Am J Gastroenterol. 2012;107(12):1826-1832. 35. Mello CJ, Irwin RS, Curley FJ. Predictive values of the character, timing, and complications of chronic cough in diagnosing its cause. Arch Intern Med. 1996;156(9):997-1003.
Original Research