Over the past several decades a dramatic increase in the employment of QOL outcome measures has been evident in the adult and pediatric clinical trials literature. In part this is a result of the trend to expand the traditionally selected, “objective” outcome measures of morbidity (i.e. days hospitalized, number of infections) and/or mortality, to include assessment of the emotional and functional status of participants. The single-minded focus on mortality and morbidity as outcomes in health is being steadily superseded by broader considerations of QOL.

One of the first stages in evaluating a new measure is to determine the “phenomenon of interest” to define the conceptual framework underlying the measure [14]. In IBD the primary outcome measure traditionally selected for use in clinical trials has been a multi-item disease activity index [15], such as the pediatric Crohn disease activity index [16, 17]. For the disease activity measures, the concept is to use the degree of intestinal inflammatory activity as a surrogate measure of the patient’s health status. This framework is based heavily on physician perceptions, with little input about the patient’s perception of the disease on their health status [15]. QOL measures try to address this deficiency. Because existing measures of disease activity are not sensitive enough to assess the full impact of the disorder, HRQOL measures have been developed to do this [18].

The ideal assessment of HRQOL would involve lengthy, detailed interviews between the patient and an independent interviewer, an impractical procedure in day-to-day clinical care. A self-administered questionnaire, that is easy to understand and complete, and that covers all the important aspects of the patient’s HRQOL, is a more attractive means of assessing HRQOL. The questionnaire should include all relevant elements, or “domains,” of HRQOL. These domains may cover physical, functional, emotional, and cognitive well-being and, in case of disease, disease-related aspects. Each domain consists of a number of “dimensions” or questions. A balance needs to be struck between including a sufficient number of dimensions so that a complete assessment of HRQOL can be made, but not to create too lengthy a questionnaire so that it becomes burdensome for the patient to complete. The advantage of combining questions into domains is that interventions can be directed at these domains, attempting to ameliorate that aspect of HRQOL.

There are two basic types of HRQOL measures, generic and disease-specific. A generic measure is designed to measure all aspects of health and its related QOL, and can include items and domains that are broadly applicable to various diseases and populations. Although disease-specific questionnaires include some of these same issues, they also address issues specific to the particular disease. Disease-specific questionnaires are more sensitive to disease-related changes in patients’ health status than generic questionnaires.

Generic measures can take several forms, from instruments with: global assessments using single indicators (e.g., “What is the quality of your life on a scale of 1 to 10?”); utilities (e.g., Standard Gamble, Time Trade-Off); or multi-item measures which give a health profile, such as the Medical Outcomes Study Short-Form 36 (SF-36) [19]. One of the advantages of a generic measure is its generalizability. Generic measures permit comparisons between “healthy populations” and different disease groups, interventions, and demographic and cultural groups [20, 21]. Generic questionnaires, such as the SF-36 for adults [20, 22] or the Child Health Questionnaire for children [23], have been applied to groups with no defined illness, allowing normative values to be generated for these healthy populations. When such normative data is available it offers the potential to make comparisons as to burden of illness between populations affected with and without chronic illness [24]. The chief disadvantage to generic measures is their insensitivity to important clinical change. This stems from their inherent lack of specificity; a result of the inclusion of many items which may not be relevant to the individual patient with an isolated disease. This can be addressed by the use of a disease-specific measure that focuses on concerns relevant to a particular patient group. “Specificity” is achieved by the inclusion of dimensions and domains which are targeted to the disease in question. For example, in the Pediatric Asthma QOL Questionnaire, a measure developed by Juniper et al. [25], the symptoms domain includes questions such as “How much did tightness in your chest bother you during the past week?” and “How often did your asthma wake you up during the night during the past week?” This specificity makes the questionnaire more sensitive to important clinical change in asthma, which is an important criterion when choosing an outcome measure in a clinical trial. In the adult literature, disease-specific questionnaires have been developed for a number of diseases, including IBD [26], rheumatoid arthritis [27], breast cancer [28, 29], and asthma [30, 31]. In comparison very few disease-specific questionnaires have been developed for use in the pediatric population [30, 32–35].

Any measurement tool should be tested prior to use to ensure it fulfills the fundamental psychometric characteristics of a good measure. A HRQOL questionnaire would be one example of a measurement tool. The psychometric characteristics to be assessed include sensibility, reliability, validity, and responsiveness to change (Table 45.1). Sensibility is a measurement characteristic with many aspects, and for a questionnaire should include assessment of feasibility for both the person administering and completing the questionnaire (i.e. time to complete and mark, readability), as well as a critical review of the appropriateness of items included or omitted. Reliability looks at whether a measure has reproducibility, that is, if the same result is obtained when the same (unchanged) entity is measured again (17). For example, assuming that HRQOL is influenced by disease activity or medication use, one would expect a reliable HRQOL questionnaire to show very similar scores when given to a patient at time one and again at time two if no interval change in disease activity or medication has occurred. Validity is concerned with whether a questionnaire actually measures what it is intended to measure. Ideally one would like to measure the validity of a HRQOL measure comparing it with a gold standard. Unfortunately, HRQOL is a concept for which no gold standard exists. Thus, a process of construct validity testing must be carried out. This involves generating hypotheses, called constructs, and studying whether the measure acts as one would expect. The final characteristic, responsiveness to change, relates to the ability of the questionnaire to detect change over time, characteristics important for use in clinical settings. The smaller the change in disease status that can be detected with a questionnaire, the more responsive the questionnaire. This latter characteristic is especially important in determining the sample size for studies in which HRQOL is a main outcome.

Making QOL assessments in a pediatric population requires the awareness of several key methodological issues: whether to ask children directly [36, 37]; and how to allow for varying developmental level and age [10, 38]. It is not always possible to obtain the child’s assessment of their QOL, whether due to age, developmental, or disability limits to comprehension. In these instances a proxy is sought. The proxy reporter of the child’s QOL is most often their parent/caregiver, but in some cases may be another individual such as a teacher or physician [37]. Research has shown, however, that proxies tend to have a low to modest agreement with the child’s reported QOL self-assessment [39]. The degree of agreement seems to relate to the objectiveness of the dimension of QOL being assessed. Pantell et al. showed that parents and teachers agreed fairly well in reporting on child functioning but markedly less well for recent functional status, certain types of subjective feelings in regard to illness, information needs, emotional states, and family functioning [40]. Agreement among raters may differ as a result of factors such as child sex, age, and condition [41].

Other issues relate to the wide developmental spectrum seen across the pediatric age group. The quality of a child’s self-report is highly dependent on their expressive and receptive language abilities [10]. As well, differences in time ­perception and memory related to their developmental stage will affect a child’s ability to respond to questions based upon experiences during a specific time period [37]. Within a given culture developmental tasks can vary by age such that some QOL items may be appropriate for a specific age range, but not for another. For example, perceptions on relationships with the opposite sex will vary with age. Other issues likely related to developmental age include position bias, the tendency to choose the first answer; acquiescence response bias, the tendency to agree with the interviewer; and limited understanding of negatively worded items [42]. Realizing there are low agreements for proxy ratings in some areas of response, it is unclear to what extent differences in response pattern are due to limitations in abstract reasoning or true differences in perspective or opinion.

Measurement of HRQOL in IBD has received a lot of attention over the several decades, with the result that there are now validated outcome measures that have been used in clinical trials or cross-sectional studies (Table 45.2). Early attempts at assessing HRQOL in IBD, however, were hampered by a number of methodological issues: healthy or medical comparison groups were not used; studies were done by retrospective analyses [44, 45]; nonstandardized instruments [45–48] and unskilled interviewers were used to obtain the data; and insensitive outcome factors (i.e. ability to work) [44, 48, 49] were used as measures of HRQOL.

The main disease-specific HRQOL instrument used currently is the IBD questionnaire (IBDQ) which was developed for IBD patients and to be used in clinical trials [26, 50, 51]. It is a 32-item questionnaire, consisting of 30 items chosen most frequently and rated most important by adult IBD patients, and two items added based on feedback obtained by clinicians who had practices heavily weighted with IBD patients. The four domains covered in IBDQ include bowel symptoms (10 items), systemic symptoms (5 items), emotional function (12 items), and social function (5 items). Responses are based on a 7-point Likert scale in which 1 represents worst function, and with 7 indicating best function, thus the higher the score the better the QOL. The questionnaire can be self-administered [52, 53] and takes approximately 15 min to complete. The IBDQ has undergone extensive testing of its measurement characteristics, including several randomized controlled trials [50, 54, 55] and cross-sectional studies [56]. From a large multicentered Canadian trial of maintenance therapy, an intraclass correlation coefficient of 0.70 was calculated for test–retest reliability in 280 patients with stable disease over an 8-week period [50]. Responsiveness testing using a modified responsiveness index developed by Guyatt et al. indicated that all IBDQ indices reflected deterioration for those patients whose condition worsened during the study [57]. Construct validity testing in the original publication of the measure, and with subsequent use of the IBDQ in trials, has shown it to be a valid measure of HRQOL in adult patients with IBD. It has been shown to have strong correlation with patient, relative, and physician global ratings and discriminate between the groups of patients who did or did not require surgery [50]. Some researchers have expressed concern about the use of a single measure to describe the HRQOL for IBD, because of the frequently disparate natures of its component diseases, Crohn disease and ulcerative colitis [43]. For example, because Crohn disease can affect variable locations in the bowel the range of symptoms can also vary greatly, with differences exacerbated by relapsing and remitting disease activity. This is compared with ulcerative colitis in which the bowel disease is limited to the colon. Given these differences some researchers have suggested that a different or separate approach to HRQOL evaluation for these two diseases may be required. This issue was apparently not addressed in the development of the IBDQ [26, 50].

If HRQOL is to be considered as an outcome measure in a multinational trial, translation of the questionnaire to the mother tongue of the studied patient population is required to reliably compare data across populations. Translation alone, without consideration of cultural differences, may not be sufficient. Cultural differences with respect to disease perception and illness experience are becoming more apparent with the increasing immigrant population residing in western countries [58]. The exclusion from a study of a group or population, based on culture or language, could lead to a systematic bias in studies of health care utilization or QOL [59]. Given the increasing number of cross-culturally adapted versions of the IBDQ [60–66] and the development [67] and subsequent validation [68, 69] of the 10-item Short IBDQ (SIBDQ), it seems unlikely that other HRQOL measures for adult IBD patients will be developed, unless it is to target specific subgroups missed in the IBDQ item generation such as patients with ileostomy.

As is the case with pediatric QOL assessment in general, consideration of QOL issues in pediatric IBD has lagged behind that of the adult IBD cohort. The earliest semblances of QOL inquiry were from a number of centers which reported the results of long-term follow-up or cross-sectional assessments of their pediatric IBD populations [2, 3, 70–76]. In many instances, rather than actually describing the QOL, they were describing the functional status of the patients [70–72, 75, 76].

Goel et al. [70] and Lindquist et al. [72] did not use a formal measure to describe QOL, but rather, in their description of the current status [70] or clinical course [72] of the patients, included limitations on social activities, school attendance, or occupation as descriptors. Farmer and Michener [73, 74] developed a simple measure which provided three categories of QOL: “Good—meaning ability to function in a nearly normal manner with minimal interference from the illness and its sequelae; Poor—indicated severe effect on life style, requiring medication and often frequent hospitalization; Fair—suboptimal but adequate functioning, i.e., chronic illness and partial disability.” Patients were categorized based on interviews by trained personnel. The researchers acknowledged that their view of QOL was a composite of several elements of the patient’s life and that patients might experience varying degrees of QOL over a long period of time. Patients were asked to consider the cumulative effect of the illness and treatment and to describe their current state of health. Farmer and Michener’s long-term follow-up study of 522 patients (followed from 1955 to 1974) with onset of CD under age 21 found that approximately two-thirds of patients considered their functioning to be in the fair level, with only 6% rating their functioning as poor [73, 74]. Given the marked changes in management over the past five decades, it is unclear what relevance QOL outcomes in such a cohort would have compared to a similar present day cohort. More recently researchers have sought to assess QOL in pediatric IBD using measures with domains which encompass a broader concept of QOL [77–80]. MacPhee et al. [80] completed an assessment of thirty pediatric IBD patients using a number of generic psychological and QOL questionnaires. Their study emphasized social supports and coping strategies. They used the QOL for Adolescents and Parents questionnaire [81], a generic measure which gives a total satisfaction score with health status and similar scores for subscales.

Thomas et al. [77, 78] describe the early stages of development of a disease-specific QOL questionnaire which they used to assess QOL in their pediatric Crohn disease cohort. Focus group meetings were held with pediatric patients of ages from 8 to 17 years (two groups, 8–12 years and 12–17 years of age) to learn how their disease and its treatment affected their lives at school, at home, and with friends. An 88-item questionnaire was constructed based on the areas identified in the focus groups. The questionnaire contained six domains of HRQOL including symptoms and treatment, social life, emotional state, family life, education, and future aspects. No data on validity, reliability, or sensibility was provided for this questionnaire [77, 78]. The questionnaire was used in one pilot study involving 16 children from one academic IBD program in England. Acknowledging the limitations of a small sample size, they found that CD appeared to most adversely affect the HRQOL of children as manifested through school absenteeism, fatigue limiting sports activities, and difficulties in taking holidays.

Moody et al. [79] studied QOL in pediatric Crohn disease using a questionnaire they developed in conjunction with a British national lay committee of Crohn in Childhood Research Association members. Limited information is provided on the questionnaire’s development, and it’s length and exact format is unclear from the published report [79]. Results from 64 valid questionnaires were received in a pilot study. The mean age of the children in this study was 14.1  ±  2.8 years (range 6–17 years). In this cohort 60% of the children reported prolonged absences from school, with a mean 3  ±  2.8 months’ absence in the previous 12 months. Eighty percent of those who had taken examinations felt that their marks had suffered due to ill health. Seventy percent of patients with CD were unable to participate in sports on a regular basis, 60% did not feel comfortable leaving their homes, and 50% did not feel they could play outside with their friends, because of the illness. Forty percent of children also reported concerns about taking holidays and being able to have sleepovers at friends’ homes. This study would suggest that CD has a major impact on the QOL of pediatric patients. However, caution should be exercised in making these conclusions as there are several limitations of the published study. It is not clear if the questionnaire underwent any validity testing to ensure it was measuring what it intended to measure. Given the study design, in which a general mailing was sent to members of a society, there may be a strong response bias in favor of those whose QOL is poor. As well, the authors do not tell us the number of questionnaires distributed, nor do they clarify the response rate.

Preliminary development of the Pediatric IBDQ (PEDIBDQ) for children and teens [82, 83], and a computer-based animated program to assess HRQOL for young children 5–11 years of age [84] have been reported in abstract or manuscript form. Further work has not been reported using these questionnaires, however. In the mid-1990s, researchers at the Hospital for Sick Children in Toronto, Canada began work on a disease-specific measure, the IMPACT questionnaire [85], which today is the most commonly employed disease-specific measure for assessing HRQOL in the pediatric IBD population.