Development and Validation of a Rome III Functional Gastrointestinal Disorders Questionnaire for Infants and Toddlers

Abstract

Objective: 

Functional gastrointestinal disorders (FGIDs) in infants and toddlers are common, but no questionnaire is available for use in clinic and research. The purpose of the present study was to develop and validate a questionnaire assessing symptoms associated with FGIDs in infants and toddlers.

Methods: 

Questions were developed based on the Rome III diagnostic criteria for FGIDs. A group of parents of children with FGIDs and experts in FGID reviewed the questionnaire for content, understandability, and completeness (face validity). Initial content validity was established by comparing physician and questionnaire diagnoses in a group of 332 consecutive new patients at a tertiary care clinic.

Results: 

Ten parents and 8 experts identified no major problems, indicating good face validity. Of 332 consecutive new patients, age 1 month to 4 year of age, 172 subjects (52% of the sample) qualified for a FGID by parent responses to the questionnaire (mean age = 1.23 year, 53% girls). All of these subjects also received an FGID diagnosis by their physician. Agreement between parent and doctor was fair to substantial (κ = 0.18–0.76), except for infant rumination and functional diarrhea in infants, which showed poor overlap.

Conclusions: 

The newly developed Rome III questionnaire for infants and toddlers had good initial face and content validity. This questionnaire will be an important addition to clinical care and research of infant/toddler FGIDs. Replication of these findings in primary care is needed.

What Is Known/What Is New

What Is Known

• Functional gastrointestinal disorders are common in infants and toddlers.
• The Rome III criteria classify functional gastrointestinal disorders and questionnaires are available to identify functional gastrointestinal disorders in children, adolescents, and adults.
• No questionnaire is available to assess symptoms associated with functional gastrointestinal disorders in infants and toddlers.
• What Is New
• The present study developed a questionnaire for infant and toddler functional gastrointestinal disorders based on Rome III criteria.
• Initial face and content validity among a sample of infants and toddlers in a pediatric gastrointestinal clinic were promising.
• This questionnaire will be helpful for research and clinician practice, when classification of functional gastrointestinal disorders according to Rome is desired.
• Replication is needed in primary care.

Functional gastrointestinal disorders (FGIDs) are common affecting more than one-quarter of infants and toddlers in a community sample ⁽¹⁾. Each FGID is diagnosed based on a set of symptoms defined by Rome diagnostic criteria ⁽²⁾. Rome III criteria are used to diagnose 7 infant/toddler FGIDs: infant rumination, infant colic, infant regurgitation, infant dyschezia, cyclic vomiting syndrome, functional diarrhea, and functional constipation.

Publication of Rome criteria increased FGID research in adults and children ⁽³⁾, but there are fewer studies in infants and toddlers. One reason for the limited scientific attention to FGIDs in infants and toddlers may be the lack of a validated questionnaire. Validated questionnaires exist for adults and children/adolescents, and have been used to collect evidence-based data. A similar instrument is needed in infants and toddlers. The aim of this study was to develop a questionnaire assessing symptoms associated with infant/toddler FGIDs and to provide preliminary evidence for its validity in a tertiary care setting.

METHODS

Questionnaire Development

The Questionnaire on Pediatric Gastrointestinal Symptoms—Rome III infant/toddler version (QPGS-RIII infant/toddler) was developed based on the Rome III diagnostic criteria in a process similar to the development of the QPGS for children and adolescents ^(4,5). First, the Rome III criteria were reworded in questions about symptoms that can be understood and answered by parents or caregivers. The questions were developed by pediatric gastroenterologists (A.R., P.H.) and a psychologist (M.v.T.) with expertise in FGIDs, the Rome criteria, and questionnaire development.

Subjects

We administered the questionnaire to 332 parents of consecutive new patients, 1 month to 4 year of age, presenting to a pediatric gastroenterology clinic at Children's Hospital of New Orleans from July 2012 to January 2014. The study was approved by the Louisiana State University and the Children's Hospital of New Orleans institutional review boards in June 2012 and renewed in July 2013. We identified new patients to pediatric gastroenterology clinic at Children's Hospital in the waiting room after registration. We defined new patients as those who had never been to our clinic or those who had not been to clinic in >3 years. We enrolled subjects from every gastrointestinal division physician's clinic population. Six faculty members and 4 fellows contributed subjects.

Measures

All of the parents completed demographic information and the QPGS-RIII infant/toddler. This questionnaire is licensed through the Rome Foundation (www.romecriteria.org) and a copy can be requested through them. It consists of 27 items, divided into 7 subsections (1 for each diagnosis). Each sections starts with a general question about the symptom (eg, “For the past 3 months when not vomiting, did your child bring up food in his/her mouth after it has already been chewed and swallowed?”). Parents who agree with the question are instructed to complete the other questions in that subsection; parents who do not agree are instructed to move to the next section. This modular questionnaire facilitates parents answering questions relevant to their child, reducing time to complete the questionnaire. One of the authors, A.R., scored all questionnaires.

For children with an FGID based on the questionnaire, physician's diagnoses were obtained. We did not use diagnoses in the medical chart as these were not always Rome-based diagnoses. Rather we requested the physician who evaluated the patient to identify the correct Rome diagnosis/diagnoses based on the Rome criteria (allowing for a write-in nonfunctional/organic disease option as well). Physicians did not have access to the questionnaire data.

Data Analyses

Content validity was established by examining concordance between physician and questionnaire diagnoses. Cohen's κ were calculated and agreement identified as: poor (κ <0.00), slight (κ = 0.0–0.20), fair (κ = 0.21–0.40), moderate (κ = 0.41–0.60), substantial (κ = 0.61–0.80), or almost perfect (κ = 0.81–1.00) ⁽⁶⁾. If a parent or physician skipped a question, the concerned was requested to provide the missing information; hence, there were no missing data.

RESULTS

Face validity was evaluated by asking 10 parents of infants/toddlers with FGIDs and 8 experts in FGID (pediatric gastroenterologists and psychologists) to review the questionnaire for content, understandability, and completeness. All of the parents understood the items. Four of 10 parents gave no further comments. Six parents suggested small textual/grammatical changes to 7 items. The majority of changes were made as suggested. Changes were not made if it would revise the content of the item (1 time) or the change would make the item or answer category deviate from previously validated QPGS in children and adolescents ^(4,5) (1 time). The experts suggested more detailed changes in wording and answer categories. These changes were minor and focused mainly on consistent wording (eg, using “your child” rather than “infant” or “toddler”) and use of answer categories (eg, create equal intervals). No one identified major problems. Changes were made based on the feedback.

Of 332 questionnaires, 168 patients (50.6% of the sample) qualified for an FGID. Mean age was 1.2 years with 53% of the sample being younger than 12 months of age, 20.8% between 1 and 2 years, 16.7% between 2 and 3 years, and 14.3% between 3 and 4 years old. Two subjects were 4 years old and excluded from further analyses as the infant/toddler Rome criteria are only applicable to 0 to 3 years of age. The sample consisted of 53% girls. Physicians provided a Rome III diagnosis for 100% of the sample. No subjects were identified as having an organic disorder.

The most common diagnoses based on both questionnaire and physician report was functional constipation in >70% of 1 to 3 years old toddlers. The most common diagnosis in infants was infant regurgitation, which was present in about half of all of the infants (Table 1). Functional diarrhea was the least reported in either age group (1%–2.5%). More than 1 diagnosis was reported for 23 (27.9%) of infants by questionnaire and 13 (16.5%) by physician. Of the toddlers 10 (11.5%) qualified for >1 disorder by questionnaire and 3 (3.4%) by physician report. Agreement between questionnaire and physician was fair to substantial for most disorders with the exception of functional diarrhea in infants and infant rumination, which showed low overlap (κ = 0.02; see Table 1).

TABLE 1: Physician and questionnaire agreement

DISCUSSION

In this article, we describe the development of the QPGS-RIII for infants and toddlers and provide preliminary evidence for its validity. The infant toddler version was developed through a rigorous process, similar in scope to the QPGS child/adolescent version ⁽⁵⁾. The questionnaire had good face validity based on parent and professional expert feedback. In addition, reasonable content validity was found: overlap between questionnaire and physician diagnoses was moderately good for all but 2 disorders in infants and better than results reported for the child/adolescent version of the Rome III questionnaire ⁽⁷⁾. For example, functional constipation showed one of the highest agreement between the QPGS and physician diagnosis in both age groups, but the concordance was substantial among infants/toddlers (κ = 0.61 for infants and 0.71 for toddlers), whereas only fair among children/adolescents (κ = 0.34 for concordance of both child-report and parent-report with physician diagnosis) ⁽⁷⁾.

The reasons for the low agreement between questionnaire and physician diagnosis for infant functional diarrhea and rumination need to be examined. First, the low number of children qualifying for these disorders likely plays a role. Better agreement was found for the disorders that were more common. It may also be the case that it is difficult for parents to know whether their infant is ruminating or has regurgitation. Similarly it may be difficult to determine normal versus loose stool consistency in children who are not potty trained. Therefore, using picture based stool charts, such as the Amsterdam Infant Stool Scale ⁽⁸⁾ may improve agreement. These issues need to be addressed in future studies.

Although preliminary validation of the QPGS-RIII infant/toddler version appears adequate, this study has important limitations. First, validation was performed in a tertiary care clinic. These patients likely present with more severe symptoms and may have a different sociodemographic makeup than patients in primary care or in the community. Unfortunately, we did not have information on sociodemographics of these families. A recent study found FGIDs to be common in the general US population, with more than one-quarter of infants and toddlers affected by FGIDs according to the QPGS without differences in sociodemographic status ⁽¹⁾. There was, however, no medical evaluation for these children, so we do not know how many children may have been misdiagnosed. More studies are needed to validate the questionnaire outside of tertiary care. Second, although none of the patients defined by the QPGS-RIII as experiencing an FGID had a physician diagnosis of an organic disease, we do not know how many children were missed by the QPGS (a physician diagnosis of an FGID but did not receive a diagnosis with the QPGS-RIII). Third, we have no follow-up on the subjects to learn whether the FGID diagnosis changed to one of organic disease. This needs to be evaluated in future studies. Furthermore, we did not include test-retest reliability or convergent validation. For FGIDs there is no criterion standard. Physicians may not agree on the correct diagnosis for any 1 FGID patient ⁽⁹⁾. Therefore, validation is needed in other ways. Compared with healthy children, infants and toddlers with an FGID, as determined with the new questionnaire, show decreased quality of life, increased medical visit, and hospitalizations ⁽¹⁾, which provides further validation for the questionnaire.

The QPGS-RIII for infants and toddlers is a parent-reported questionnaire of their child's symptoms, useful for diagnosing Rome III diagnoses for FGIDs. We have provided initial evidence for its validity. This questionnaire will be an important addition to clinical care and research of infant/toddler FGID.