Gastroesophageal reflux disease (GERD) in the pediatric population (children ages 2–17 years) is defined as the presence of troublesome gastrointestinal symptoms and/or complications (1). Predominant symptoms of GERD differ among age subgroups in this population (1,2). Common symptoms among children and adolescents include regurgitation, retrosternal chest pain, dysphagia, wheezing, chronic cough, and recurrent pneumonia (3). Diagnostic tests such as endoscopy and pH monitoring are often indicated for the diagnosis of GERD (4). Although in clinical practice a provisional diagnosis of GERD is usually made on the basis of symptoms (5), patients presenting with severe symptoms or symptoms suggestive of GERD complications should undergo specialized testing before treatment and not be diagnosed on history and physical examination alone (4). Those without such features may not require diagnostic testing, at least not at first presentation. A diagnosis based on symptoms alone allows a time limited but not long-term treatment (4). Further caution must be exercised with symptom-based diagnosis and treatment because the symptoms of GERD are often indistinguishable from those of other disorders, especially eosinophilic esophagitis (4,6).
For those patients not requiring investigation immediately, the use of patient-reported outcome (PRO) measures mirrors clinical pediatric practice, while permitting more systematic assessment. Data gathered via PROs can provide information on symptoms and the effects of treatment known only to the patient (eg, pain, discomfort). Use of PROs in GERD adds value by providing insight into both the frequency and severity of symptoms and the impact of symptoms and treatment on the patient's functioning and health-related quality of life (HRQL) (7,8). Systematic development of psychometrically sound PROs allows their use for collecting reliable and valid data on the symptoms and impact of pediatric GERD, and confirmation of the responsiveness of PRO instruments to a change in clinical status supports their potential usefulness for assessing treatment effects. There are many instruments to measure GERD in adults, but few available instruments for children. In children beyond infancy (older than 1 year), available questionnaires for measuring pediatric GERD symptoms include the MM-GERD (9), GERD Symptom Questionnaire-Young Child for children ages 1 to 4 (10), and the GERD Assessment of Symptoms in Pediatrics Questionnaire for children ages 5 to 11 years (11). Evidence that parents or children had input into the conceptual framework and item generation for these instruments is not available. Such patient input is strongly recommended by the Food and Drug Administration for PROs that are to be used in pharmaceutical labeling or promotional claims (12). Furthermore, published data on the measurement qualities of these questionnaires are limited, and no information is available on responsiveness needed before confident application of these questionnaires in clinical trials. Finally, there is no instrument publicly available to assess the effect of GERD symptoms on the daily lives of children, including interference with school, social function, and social activities.
The objectives of the present study were to develop and evaluate the psychometric properties of a pediatric GERD instrument suitable for measuring both the symptoms and the effect of those symptoms on daily life and functioning in children ages 2 to 17 years. Because both the symptoms and the patient's ability to self-report the symptoms vary by age, the instrument was developed in 2 age-based versions that together encompass the entire age range of interest. This study describes the development of a new pediatric GERD instrument and the evaluation of the psychometric characteristics of its 2 age-stratified versions.
MATERIALS AND METHODS
The Pediatric Gastroesophageal Reflux Disease Symptom and Quality of Life Questionnaire (PGSQ) was developed as 2 age-appropriate versions because of age-related differences in symptoms, child development considerations, and the necessity for surrogate reporting for the younger children. The version aimed at the child ages 2 to 8 years is completed by the patient's parent (PGSQ-Cp; view the questionnaire at http://links.lww.com/MPG/A37), and the version for older children and adolescents ages 9 to 17 years (PGSQ-A; view the questionnaire at http://links.lww.com/MPG/A36) is completed by the patient. All of the research activities obtained appropriate institutional review board approval and signed consents and assents (for participants ages 9–17). All of the research was conducted as per good clinical practice and the Declaration of Helsinki.
The PGSQ development followed standard methods (13,14) and was an iterative process, with an emphasis on eliciting information from children with GERD and parents of children with GERD. Instrument development began with a literature review to identify the symptoms and effects of pediatric GERD and the outcome of treatment. The results of this review were used to guide the collection of qualitative data from children and parents of children with GERD through focus groups. Following analysis of the data, expert clinicians reviewed items generated from the focus group data and an initial questionnaire was drafted. The initial questionnaire was then cognitively debriefed in one-to-one interviews with children and the parents of children with GERD. Questionnaire development occurred in the United States, and only English-speaking individuals participated.
Focus Groups and Expert Review of Items
Focus group participants were recruited from 4 pediatric gastroenterologists' offices located in different geographic areas of the United States (Texas, Ohio, Illinois, and Pennsylvania). Participating children were required to have either current symptomatic GERD confirmed within the last 3 months by pH probe, upper gastrointestinal endoscopy, or radionuclide milk scan or current symptomatic GERD responsive to acid-suppressive medication. On the basis of age, either the children themselves or their parents were part of the focus groups. Focus groups aimed at an open group discussion with the moderator eliciting responses from the group using nonleading methods.
Ten focus groups, with a total of 92 participants, were convened: 5 with parents of pediatric patients with GERD (n = 25) and 5 with pediatric patients with GERD themselves (n = 67). The majority of the focus group patients were white (91%), slightly more than half were girls, and all ages were represented.
Focus group findings suggest that patients with pediatric GERD experience a wide range of symptoms. Frequency and severity of symptoms were variable and described as affected by the types of foods and drinks consumed, empty stomach, timing and quantity of food consumption, stress, activity level, and other illnesses.
Participants reported that GERD had a major impact on many aspects of patients' lives, particularly school attendance/performance and participation in extracurricular and social activities. GERD also contributed to general feelings of frustration regarding symptoms, their effect on daily life, and the need to take medication.
Focus group findings were used to draft a comprehensive set of items that measured symptoms of pediatric GERD and their effects. All of the items were then reviewed by an expert panel consisting of 7 pediatric gastroenterologists, 1 pediatrician, and 2 specialists in outcomes research. The panel found that the items generated from child and parent qualitative data were comprehensive and clinically relevant for patients with pediatric GERD.
Two age-related versions of the instrument were developed for further testing, 1 for children ages 2 to 7 to be completed by a parent and the other for older children ages 8 to 17 years. Twenty-one English-speaking participants were enrolled for cognitive debriefing interviews, including 11 children ages 8 to 17 years and 10 parents of children ages 2 to 7 years. Upon completion of the appropriate version of the PGSQ, participants were asked a series of questions by an interviewer using a scripted guide. Participants were asked to provide feedback regarding the applicability, relevance, comprehensiveness, clarity, and acceptability of the questionnaire instructions, item content, and response scales. Results indicated that both age-related versions of the PGSQ were clear, easy to complete, and appropriate for assessing pediatric GERD symptoms and their effect on everyday life. Participants considered the items highly relevant and comprehensive. Slight modifications were made based on the interviews, and the draft instruments were then included in a larger psychometric evaluation study.
The psychometric evaluation study was designed as a 3-week, longitudinal, observational nondrug study of children with (“GERD”) and without (“controls”) GERD, using parental responders for the younger children. The study was conducted at 11 clinical sites between March and November 2007. These sites included some staffed solely by pediatric gastroenterologists and some with access to general pediatricians (for access to controls). The recruitment goal was to enroll a sample size of 110 for each age-group questionnaire: 70 parents of children ages 2 to 8 with GERD and 40 parents of control children ages 2 to 8; 70 adolescents ages 9 to 17 with GERD and 40 control adolescents ages 9 to 17. This sample size was adequate for conducting factor analysis and other planned analyses (15,16).
All of the participants were required to be able to read, write, and speak English. The GERD and control subjects for the 2 age groups were operationally defined as follows:
1. Primary caregivers living with a child with GERD who at visit 1 was 2 to 8 years old, either newly diagnosed as having pediatric GERD and receiving a newly prescribed medication or previously diagnosed as having pediatric GERD but being switched to a new treatment or dosage because of lack of response to previous treatment.
2. Primary caregivers living with a child without GERD who at visit 1 was 2 to 8 years old and the child did not demonstrate any gastrointestinal symptoms compatible with GERD as the reason for the visit and did not have a current diagnosis of GERD.
3. Children or adolescents with GERD who at visit 1 were 9 to 17 years old and either newly diagnosed as having pediatric GERD and receiving a newly prescribed medication or previously diagnosed as having pediatric GERD but being switched to a new treatment or dosage because of lack of response to previous treatment.
4. Children or adolescents without GERD ages 9 to 17 years who at visit 1 were 9 to 17 years old and did not demonstrate any gastrointestinal symptoms compatible with GERD as the reason for the visit and did not have a current diagnosis of pediatric GERD.
Participants were excluded from the study if they had any condition such as visual or cognitive problems that may prohibit questionnaire completion. Control patients were seen for a non-GERD–related complaint and were excluded if gastrointestinal symptoms such as reflux or abdominal pain were reasons for the clinic visit. These control patients were included to evaluate known-group validity and ensure a full range of PGSQ scores for sensitivity tests. All of the patients underwent a brief clinical examination to identify signs of GERD and their severity.
Although the study-inclusion criteria required that patients have a GERD diagnosis, the study protocol did not mandate endoscopy or other procedures to verify diagnosis. Therefore, it was necessary to rely on a clinical diagnosis. If a patient newly diagnosed at visit 1 was later determined not to have GERD during the study (eg, at visit 2 after a trial of proton pump inhibitors), then the patient was excluded from the analyses.
Instruments and Assessment Schedule
PGSQ Initial Draft Versions
Initial versions of both PGSQ-Cp and PGSQ-A consisted of 21 items measuring symptoms and 20 items measuring effect of symptoms on everyday life. Response options were based on the number of days/week, including “none/0 days,” “1 or 2 days,” “3 or 4 days,” “5 or 6 days,” and “every day/7 days.” For the impact items, response options were “never,” “almost never,” “sometimes,” “almost always,” and “always” (item scores ranged from 0 to 4). Subscale scores consisted of item scores summed and divided by the number of items in the subscale. The recall period was the last 7 days. These draft PGSQ versions were administered to patients and parents at the baseline visit (visit 1) and at the follow-up visit (visit 2) 3 to 6 weeks after the baseline visit.
Global Symptom Visual Analog Scales
A global symptom questionnaire in the form of 4 visual analog scale (VAS) items asked about 4 key symptoms of pediatric GERD: burning in the stomach and chest, tasting vomit in the mouth, vomiting, and loss of appetite. The VAS was a 10-cm vertical line anchored by 0 and 10, on which 0 represented the symptom “not present at all” and 10 represented the symptom “as bad as it can be.” These 4 global symptom VAS items were administered at visits 1 and 2.
Overall Treatment Effect Scale
Parent- or patient-rated and clinician-rated overall changes in pediatric GERD symptoms were assessed at visit 2 (17–20). The first question asked participants to indicate whether the patient's pediatric GERD symptoms had improved, remained the same, or worsened since the last evaluation. If improvement was indicated, then they were asked to rate the degree of improvement on a 7-point scale from “almost the same, hardly better at all” (1) to “a very great deal better” (7). If worsening was indicated, they were asked to rate the degree of worsening on a 7-point scale from “almost the same, hardly worse at all” (−1) to “a very great deal worse” (−7).
Pediatric Quality of Life Inventory
The Pediatric Quality of Life Inventory (PedsQL) measures HRQL in children and adolescents (21,22). The PedsQL generic core scales include 4 domains: physical functioning (8 items), emotional functioning (5 items), social functioning (5 items), and school functioning (5 items). Parents of children ages 2 to 8 years completed parent versions of the PedsQL. Adolescents ages 9 to 17 years self-completed the PedsQL. The PedsQL was administered at visits 1 and 2.
At visit 1, all of the participants completed a brief sociodemographic form with items including age, race, sex, education level, and other household members.
Clinicians completed a clinical form for each child at both visits 1 and 2. Data collected included the child's height and weight, comorbidities, date of diagnosis of pediatric GERD (if applicable), severity of pediatric GERD, other health conditions, current medications, change in medications, and any lifestyle change recommendations.
SAS version 9.1.3 (SAS Institute, Cary, NC) was used for all of the analyses. This psychometric analysis focused on the reliability and validity of each version of the PGSQ.
Item Selection and Factor Analyses
Characteristics of individual PGSQ items were examined by calculating the mean, minimum and maximum response (assessing for floor and ceiling effects), percentage of missing, and item-total and interitem scale correlations using MAP-R (23). Exploratory factor analysis (EFA) with orthogonal and oblique rotations was then used to examine the underlying structure among the PGSQ items. Items were selected for deletion based on predetermined rules for nonresponse, correlations, factor loadings, and clinical relevance. Once factors were derived from the EFA and items were deleted, item response theory (IRT) analysis was used to assess item parameters and characteristics within each scale and for the total score (24,25).
Internal consistency reliability addresses the homogeneity of items in a scale and was assessed for the PGSQ using Cronbach alpha. Values >0.70 are generally considered acceptable for aggregate data (16,26).
The reproducibility of PGSQ scores was examined in the subgroup of test-retest caregivers and patients who indicated little or no change in their GERD symptoms based on the overall treatment effect scale (OTE) at visit 2. Intraclass correlation coefficients (ICCs) were calculated between visit 1 and visit 2 to evaluate test-retest reliability. ICCs above 0.70 are considered good and those above 0.60 are acceptable.
Construct validity was assessed by examining the magnitude of correlations between the PGSQ and the PedsQL and GERD symptoms. Concurrent validity was supported when the PGSQ was substantially correlated (>0.40) with subscales measuring similar concepts. Conversely, scales measuring different concepts should be less correlated (<0.40).
To assess known-group validity at visit 1, PGSQ symptom and impact scores were compared between GERD and control groups, with mean score comparisons made using t tests and confidence intervals. Participants were stratified by the clinician severity rating (0 = none, 1 = mild, 2 = moderate, and 3 = severe), and analysis of covariance models, controlling for age and baseline visit score, were used to compare severity groups, with post hoc category comparisons via the Scheffe test.
The extent to which PGSQ scores were responsive in patients experiencing a change in GERD symptoms was explored. Patients with GERD were identified who were classified as responders based on the parent- or patient-rated OTE and the clinician-rated OTE. Responders were defined as patients having an improvement of 2 (“a little better”) or greater. Effect sizes were then calculated using the following formula: difference in mean score for responders from visit 1 to visit 2 divided by the standard deviation of visit 1 scores for all of the subjects (27). Effect sizes of 0.5 to 0.8 are consistent with a moderate improvement, and those >0.8 are consistent with a large improvement (28).
A total of 231 participants were enrolled at 11 sites, comprising parents of 2- to 8-year-old children with pediatric GERD (n = 75) and without pediatric GERD (n = 41) and 9- to 17-year-old adolescents with pediatric GERD (n = 75) and without pediatric GERD (n = 40). Demographic characteristics were comparable between the participants with and without GERD for age and race/ethnicity (Table 1). For the patients with GERD, the mean duration since diagnosis was 1.2 (SD 1.8) years, 13.3% also had asthma, and 23.3% had another medical condition. GERD severity as rated by clinicians was 48.7% mild, 49.3% moderate, and 2% severe.
Item Selection and Analyses
Results from item descriptive statistics, item differentiation between patients and controls, IRT analyses, and expert clinician review were used to reduce the number of items in the draft questionnaires to produce the final PGSQ instruments (data not shown). During this process, the 41-item draft PGSQ-Cp was reduced to 37 items, and the draft PGSQ-A was reduced to 35 items. Results of the EFA identified 4 symptom subscales (regurgitation, sleep, heartburn, and extraesophageal) for the PGSQ-Cp (data not shown). For the PGSQ-A, 3 distinct symptom subscales were found (regurgitation, heartburn/sleep/general, and extraesophageal) (data not shown).
Each questionnaire demonstrated subscales for the Impact and School items, which were scored separately because the school items can only be completed for children currently attending school. Sample Impact items include “missed out on doing things with friends,” “unable to do physical activities,” “had to lay down,” and “unable to eat” what the child wanted. Sample School items include “hard time paying attention at school,” “absent from school,” “late to school,” and “leave school early.”
With the exception of a few items that had missing data (<2%), analyses were conducted on the complete dataset. Acceptable internal consistency reliabilities were observed for both PGSQ-Cp and PGSQ-A subscale scores (Table 2).
Reproducibility was acceptable for the PGSQ-Cp total symptom (ICC 0.81) and subscale (ICCs 0.65–0.85) scores, except for total Impact and School scores. For the PGSQ-A, only the extraesophageal and School score demonstrated acceptable reproducibility (Table 2).
Correlations between the PGSQ regurgitation subscale and the global questions “taste throw-up” and “sick to stomach/need to throw-up” were moderate to high with correlations of 0.79 and 0.82 for the PGSQ-Cp and 0.42 and 0.92 for the PGSQ-A. Correlations between the PGSQ Heartburn subscale and the global question “hurting/burning in chest” were also moderate to high at 0.74 for the PGSQ-Cp and 0.64 for the PGSQ-A. Correlations between the PGSQ total and subscale scores and the PedsQL scores were low to moderate, ranging from −0.18 to −0.69 for the PGSQ-Cp and −0.06 to −0.77 for the PGSQ-A (Table 3).
Both versions of the PGSQ significantly differentiated between patients with GERD and controls for all of the subscales (P < 0.001; Figs. 1 and 2). No significant differences were found by clinical level of severity for either questionnaire (data not shown).
For all of the PGSQ-Cp subscales except for regurgitation, mean visit 1 to visit 2 changes were significantly different between those patients who improved and those who were stable, based on the parent-rated OTE (Fig. 3).
For the PGSQ-Cp scores, effect sizes were consistently greater for responders compared to nonresponders for all of the scores (data not shown). For the PGSQ-Cp total symptom score, the effect size was 0.95 for responders and 0.13 for nonresponders. For Impact, the effect sizes were 1.2 and 0.05 for responders and nonresponders, respectively, and for School, the effect sizes were 0.81 and 0.06 for responders and nonresponders, respectively.
For all of the PGSQ-A subscales, mean visit 1 to visit 2 changes were significantly different between those subjects who improved and those who were stable, based on the patient-rated OTE (Fig. 4).
Responders exhibited greater improvements in symptom and impact scores than nonresponders (data not shown). The effect size observed for responders on total symptom scores was 1.14 compared to 0.02 for nonresponders. For Impact, the effect sizes were 1.2 and 0.19 for responders and nonresponders, respectively, and for School, the effect sizes were 0.72 and 0.27 for responders and nonresponders, respectively.
These 2 age-stratified PGSQ instruments are unique among the existing pediatric GERD questionnaires in the following respects: the breadth of age coverage (2–17 years), the use of detailed parent and patient information in the development of measures, the inclusion of subscales on the impact of symptoms on daily life and school, and the evidence of responsiveness to changes in clinical status. In contrast, existing symptom questionnaires for pediatric GERD cover a more limited age range (eg, 1–4 years (10) or 5–11 years (11)), are unsupported by detailed published evidence of parent or patient input into questionnaire development, address only symptoms, and have no evidence supporting responsiveness (10,11).
The PGSQ improves upon existing measurement tools by using qualitative research to determine appropriate item content and by considering the effect that these symptoms have on the everyday lives of children, including interference with school, social functioning, and family activities. A recent study by Acierno et al (29) confirms these areas of symptoms and effects based on individual patient and parent interviews. Furthermore, the wider span of ages (2–17) covered by the PGSQ than by other questionnaires allows researchers greater flexibility in designing pediatric GERD clinical studies. The PGSQ was developed as a disease-specific measure and can be used alone or in combination with a pediatric-specific HRQL instrument such as the PedsQL.
The final versions of the PGSQ-Cp and the PGSQ-A established on the basis of this study contain 37 and 35 items, respectively. The final items were selected based on a review of item psychometric characteristics, ability to discriminate between patients with and without GERD, and clinical input from experienced pediatric gastroenterologists. This study demonstrated that both age-stratified versions of the PGSQ possess good psychometric characteristics, excellent internal consistency reliability, and acceptable reproducibility. The PGSQ-Cp and PGSQ-A also have evidence supporting good construct and known-group validity.
Responsiveness was demonstrated for the symptom and impact subscales of the PGSQ-Cp and PGSQ-A. Effect sizes were in the large range for almost every subscale except Regurgitation for the PGSQ-A version. Preliminary minimally important difference estimates have been evaluated (data not shown) and need to be confirmed in clinical trials to provide further information on interpretation and clinically important differences for the PGSQ-Cp and PGSQ-A.
There are several limitations of the questionnaires and the psychometric evaluation study. First, the caregiver surrogate reporting for the younger age group may affect reliability, an issue for any questionnaire aimed at young children. Children younger than 8 or 9 years are often unable to describe or localize symptoms and are more suggestible than older children or adults (1,30,31). For children who spend a good part of their daytime hours in school or daycare situations, the surrogate may have minimal knowledge of social or school disruption due to GERD symptoms.
Second, the length of the 3- to 6-week period between visit 1 and visit 2, although important to the assessment of responsiveness, likely resulted in the lower-than-expected reproducibility results for the PGSQ-A. Usually, test-retest reliability is established using a period of 1 week, and further study of reproducibility in PGSQ scores is needed.
Third, although both age-stratified versions of the PGSQ successfully differentiated between patients and controls, the scores did not differentiate by level of clinical severity. This may have resulted from the relatively small sample sizes (N < 5) found in some of the severity groupings.
Fourth, although patients were required to be either newly diagnosed as having GERD and receiving treatment or previously diagnosed and receiving new treatment according to the study inclusion criteria, no evidence of clinical testing underlying the diagnosis of GERD was collected at the baseline visit. However, recent North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines note that a history and physical examination may be sufficient for the diagnosis of GERD if the symptoms are typical (4).
In summary, PGSQ-Cp and PGSQ-A were developed to assess pediatric GERD symptoms and their effect on children and adolescents with pediatric GERD. The qualitative work behind the development of the PGSQ, including multiple focus groups and interviews with patients with GERD and parents/caregivers, supports the content validity, relevance, and respondent comprehension of the PGSQ items. This focus on qualitative research addresses concerns raised in the Food and Drug Administration guidance for development of PROs for labeling claims (12). Patient and parent information was complemented by expert input to ensure the clinical importance of the items, the utility of the tool for clinical practice and research, the psychometric appropriateness of the instruments, and the comprehensiveness of their concept coverage.
Both age-stratified versions of the PGSQ are able to identify and quantify important GERD symptoms and their effects on children's lives, can discriminate between patients with and without GERD, and, critical to application in clinical trials, demonstrated responsiveness to changes in clinical status. Given the reliability and validity evidence provided in this report, these 2 newly developed instruments are uniquely qualified as useful tools for the measurement of symptoms and impact associated with pediatric GERD in future research.
1. Sherman PM, Hassall E, Fagundes-Neto U, et al. A global, evidence-based consensus on the definition of gastroesophageal reflux disease in the pediatric population. Am J Gastroenterol 2009; 104:1278–1295, quiz 96.
2. Gupta SK, Hassall E, Chiu YL, et al. Presenting symptoms of nonerosive and erosive esophagitis in pediatric patients. Dig Dis Sci 2006; 51:858–863.
3. Michail S. Gastroesophageal reflux. Pediatr Rev 2007; 28:101–110.
4. Vandenplas Y, Rudolph CD, Di Lorenzo C, et al. Pediatric gastroesophageal reflux clinical practice guidelines: joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN). J Pediatr Gastroenterol Nutr 2009; 49:498–547.
5. Diaz DM, Winter HS, Colletti RB, et al. Knowledge, attitudes and practice styles of North American pediatricians regarding gastroesophageal reflux disease. J Pediatr Gastroenterol Nutr 2007; 45:56–64.
6. Furuta GT, Liacouras CA, Collins MH, et al. Eosinophilic esophagitis in children and adults: a systematic review and consensus recommendations for diagnosis and treatment. Gastroenterology 2007; 133:1342–1363.
7. Kleinman L, Revicki DA, Flood E. Validation issues in questionnaires for diagnosis and monitoring of gastroesophageal reflux disease in children. Curr Gastroenterol Rep 2006; 8:230–236.
8. Revicki DA, Wood M, Maton PN, et al. The impact of gastroesophageal reflux disease on health-related quality of life. Am J Med 1998; 104:252–258.
9. Malaty HM, O'Malley KJ, Abudayyeh S, et al. Multidimensional measure for gastroesophageal reflux disease (MM-GERD) symptoms in children: a population-based study. Acta Paediatr 2008; 97:1292–1297.
10. Deal L, Gold BD, Gremse DA, et al. Age-specific questionnaires distinguish GERD symptom frequency and severity in infants and young children: development and initial validation. J Pediatr Gastroenterol Nutr 2005; 41:178–185.
11. Tolia V, Bishop PR, Tsou VM, et al. Multicenter, randomized, double-blind study comparing 10, 20 and 40 mg pantoprazole in children (5-11 years) with symptomatic gastroesophageal reflux disease. J Pediatr Gastroenterol Nutr 2006; 42:384–391.
12. Food and Drug Administration. Guidance for Industry on Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims. Fed Regist 2009;74:65132–3.
13. Patrick DL, Burke LB, Powers JH, et al. Patient-reported outcomes to support medical product labeling claims: FDA perspective. Value Health 2007; 10(Suppl 2):S125–S137.
14. Revicki DA, Osoba D, Fairclough D, et al. Recommendations on health-related quality of life research to support labeling and promotional claims in the United States. Qual Life Res 2000; 9:887–900.
15. Gorsuch RL. Factor Analysis. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum; 1983.
16. Nunnally JC, Bernstein IH. Psychometric Theory. 3rd ed. New York: McGraw-Hill; 1994.
17. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials 1989; 10:407–415.
18. Guyatt GH, Juniper EF, Walter SD, et al. Interpreting treatment effects in randomised trials. BMJ 1998; 316:690–693.
19. Juniper EF, Guyatt GH, Willan A, et al. Determining a minimal important change in a disease-specific Quality of Life Questionnaire. J Clin Epidemiol 1994; 47:81–87.
20. Revicki DA, Sorensen S, Maton PN, et al. Health-related quality of life outcomes of omeprazole versus ranitidine in poorly responsive symptomatic gastroesophageal reflux disease. Dig Dis 1998; 16:284–291.
21. Varni JW, Seid M, Kurtin PS. PedsQL 4.0: reliability and validity of the Pediatric Quality of Life Inventory version 4.0 generic core scales in healthy and patient populations. Med Care 2001; 39:800–812.
22. Youssef NN, Murphy TG, Langseder AL, et al. Quality of life for children with functional abdominal pain: a comparison study of patients' and parents' perceptions. Pediatrics 2006; 117:54–59.
23. Hays RD, Hayashi T. Beyond internal consistency reliability: rationale and user's guide for multi-trait analysis program on the microcomputer. Behav Res Meth Instr Comp 1990; 22:167–175.
24. Reeve BR, Fayers P. Applying item response theory modeling for evaluating questionnaire item and scale properties. In: Fayers P, Hays R, editors. Assessing Quality of Life in Clinical Trials. 2nd ed. New York: Oxford University Press; 2005. pp. 55–73.
25. Van der Linden WJ, Hambleton RK, eds. Handbook of Modern Item Response Theory. New York: Springer-Velag; 1997.
26. Hays RD, Revicki DA. Reliability and validity, including responsiveness. In: Fayers P, Hays RD, editors. Assessing Quality of Life in Clinical Trials. 2nd ed. New York: Oxford University Press; 2005.
27. Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpreting changes in health status. Med Care 1989; 27(3 Suppl):S178–S189.
28. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Erlbaum; 1988.
29. Acierno S, Chilcote H, Edwards T, et al. Development of a quality of life instrument for pediatric gastroesophageal reflux disease: qualitative interviews. J Pediatr Gastroenterol Nutr 2010; 50:486–492.
30. Stanford EA, Chambers CT, Craig KD. The role of developmental factors in predicting young children's use of a self-report scale for pain. Pain 2006; 120:16–23.
31. von Baeyer CL, Spagrud LJ. Systematic review of observational (behavioral) measures of pain for children and adolescents aged 3 to 18 years. Pain 2007; 127:140–150.