A New Measure of Health Status for Clinical Trials in Inflammatory Bowel Disease

Transcription

1 GASTROENTEROLOGY 1989;96: A New Measure of Health Status for Clinical Trials in Inflammatory Bowel Disease GORDON GUYATT, ALBA MITCHELL, E. JAN IRVINE, JOEL SINGER, NANCY WILLIAMS, ROBERT GOODACRE, and CATHY TOMPKINS School of Nursing, Department of Clinical Epidemiology and Biostatistics, and Department of Medicine, McMaster University, Hamilton, Ontario, Canada We have developed a measure of subjective health status (quality' of life) for patients with inflammatory bowel disease (IBD). Ninety-seven patients with IBD described problems they had experienced as a result of the disease; the 32 most frequent and important items were included in the Inflammatory Bowel Disease Questionnaire (IBDQ). Sixty-one IBD patients were evaluated twice. One month separated the evaluations, at which disease activity indices, the IBDQ, and a number of other questionnaires were administered. Reproducibility studies in 19 stable patients showed improvement in scores, but also a small within-person standard deviation. Responsiveness studies revealed large changes in scores in patients who had improved or deteriorated and suggested that the IBDQ was more responsive than a general health status measure. Responsiveness appeared greater in patients with ulcerative colitis than in those with Crohn's disease. Predicted and observed correlations between changes in IBDQ score and changes in other measures were similar. We conclude that although further testing is required, particularly in examining the relation between changes in the IBDQ and changes in the activity of Crohn's disease, the IBDQ shows promise as a measure of health status for clinical trials in IBD. M easures of disease activity that are based on laboratory or endoscopic variables have, appropriately, served as primary measures of outcome in clinical trials in inflammatory bowel disease (IBD), defined here as Crohn's disease and ulcerative colitis. There are a number of instruments available that measure disease activity (1-4), all of which have limitations (including limited reproducibility) (5). A major limitation of these indices is that they do not provide a detailed picture of the patient's subjective, including emotional and social problems associated with IBD. The importance of measuring subjective aspects of health status (often referred to as quality of life) has become increasingly recognized (6). Qualitative and semiquantitative descriptions of the problems affecting IBD patients have documented limitations in work and social activities, home and married life, and emotional (7-22). Although measures of patient information and psychologic adjustment in IBD are available (13), there is no instrument that examines the broad range of problems experienced by patients with IBD in a form suitable for use as an outcome measure in clinical trials. A variety of measures of health status for general populations are available (23,24), but these measures may not focus adequately on the specific problems of IBD patients. In other area.s, including cancer, arthritis, and chronic lung disease, this potential problem of general instruments has led investigators to construct disease-specific questionnaires (25-27). The purpose of the work described here was to develop a health status measure for clinical trials that examines disease-related dys in IBD patients. The questionnaire, the Inflammatory Bowel Disease Questionnaire (IBDQ), examines four aspects of patients' lives: symptoms directly related to the primary bowel disturbance, systemic symptoms, and emotional and social. In this paper the development of the questionnaire is described, and data regarding its reproducibility, validity (the extent to which it is measuring what it is supposed to measure), and responsiveness are presented. Abbreviations used in this paper: IBD, inflammatory bowel disease; IBDQ, Inflammatory Bowel Disease Questionnaire by the American Gastroenterological Association /89/$3.50

2 March 1989 MEASURING HEALTH STATUS IN IBD 805 Methods Principles of Questionnaire Development The questionnaire was designed to meet the following criteria (6,28): 1. Items must reflect areas of that are important to patients with IBD. 2. Summary scores should be amenable to statistical analysis. 3. Repeated administration in stable patients must yield similar results. 4. When even a small clinically important change in score has occurred, the questionnaire score should reflect it. 5. The questionnaire should be valid. 6. The questionnaire should be relatively short and simple. A summary of the process of instrument development and testing is presented in Figure 1. Recruitment Strategy We intended that the final questionnaire would be applicable to a broad spectrum of patients with IBD. However, we specifically excluded patients with proctitis, and those with an ileostomy. The rationale for these exclusions was that the problems of these groups are substantially different from those of other IBD patients. Recruitment strategies, described below, reflected the target population. Item Selection and Item Reduction A list of problems important to patients with IBD was constructed. These items were generated by administering an open-ended questionnaire to clinicians in daily contact with IBD patients and to 77 patients with IBD, from a review of the literature describing the problems of IBD patients (7-22), from general quality of life questionnaires (23,24), and from other disease-specific instruments (28). Once this problem list had been generated, 97 subjects with IBD (54 with Crohn's disease and 43 with ulcerative colitis; 46 men and 51 women; mean age, 39 yr) were asked to identify all physical, emotional, or social problems they had experienced as a result of their IBD and to grade the severity of these problems in their daily lives. The results of this process were used to construct the IBDQ. Item Selection: Survey of clinicians, 77 IBD patients, literature review Construction of Item Reduction Questionnaire (150 items) from items identified ~ Item Reduction: Questionnaire administered to 54 patients with Crohn's Disease 43 patients with Ulcerative Colitis l 32 most frequent and important items included in IBDQ ~ Testing of IBDQ 61 patients with IBD interviewed twice 1 mo apart Reproducibility Testing 19 Global Rating "No Change" 15 Ulcerative Colitis 4 Crohn's Disease Responsiveness and Validity Testing 33 Global Rating "Improved" 9 Global Rating "Deterioration" Of total of 42: 23 Ulcerative Colitis (disease activity available in 19) 19 Crohn's Disease (disease activity available in 11) Figure 1. Development and testing of IBDQ.

3 806 GUY ATT ET AL. GASTROENTEROLOGY Vol. 96, No.3 Inflammatory Bowel Disease Questionnaire Testing When development and pretesting of the IBDQ had been complete, 61 patients with ulcerative colitis or Crohn's disease were interviewed twice by a single interviewer, 1 mo elapsing between interviews. Seven patients were recruited from the membership of the Canadian Foundation for Ileitis and Colitis, 11 were initially recruited while in hospital. and 43 were recruited from gastroenterology outpatient clinics. All patients had histologically confirmed IBD. At each interview the IBDQ and a general measure of physical and emotional developed by the Rand Corporation were administered (24). For patients whose disease was considered active by their gastroenterologist at the time of the initial assessment, the Van Hees index of disease activity (Crohn's disease) (4) or the st. Mark's index (ulcerative colitis) (29) was calculated. However, disease activity measures were available only for those subjects seen by a physician at both assessments. Of the 61 patients tested, this was true for 19 patients with ulcerative colitis and 11 with Crohn's disease. Because there was an insufficient number of patients with Crohn's disease among this group for meaningful analysis, the data relating to assessment of disease activity include only the patients with ulcerative colitis. At the follow-up interview patients were asked whether their disease activity had deteriorated, remained the same, or improved. If they had improved or deteriorated, the degree of change was quantified using a seven-point Likert scale from "almost the same, hardly any better (or worse)" through "moderately better (or worse)" to "a very great deal better (or worse)." Similar global ratings were elicited for fatigue and emotional. A relative or spouse who lived in the same household was also asked to make global assessments of the patient's change in disease activity, fatigue, and emotional. At the second interview, patients were classified into two groups by disease activity as changed or unchanged. The former group was considered clinically "unstable," the latter clinically "stable." Measures of reproducibility, responsiveness, and validity were extracted from these data. Results Item Selection and Item Reduction The Item Reduction Questionnaire included 150 items and was administered to 97 IBD patients. Of these 97,5 had participated in the item generation process. The results of the Item Reduction Questionnaire are presented in detail in another publication (30). In summary, the items chosen most frequently and rated most important by the subjects fell into four dimensions: gastrointestinal symptoms, symptoms not directly related to the bowel disturbance (which we shall call systemic symptoms), symptoms of emotional dys, and social dys. The items chosen most frequently, and rated as most important, were included in the IBDQ. The IBDQ that emerged from this process contained 30 items that were serially pretested to identify poorly worded questions and to improve item presentation. Clinicians who had practices heavily weighted with IBD patients were presented with the questionnaire and asked for feedback; on the basis of their responses, two additional items were added. The final IBDQ includes 10 questions relating to bowel symptoms, five questions relating to systemic symptoms, 12 questions relating to emotional, and five questions relating to social. Initial administration of the IBDQ takes a maximum of 30 min, and usually between 15 and 25 min. Follow-up administration takes a maximum of 25 min, and usually between 10 and 20 min. The structure and content of the questionnaire are described in the Appendix. The response options for each question are framed as a seven-point scale in which 7 represents best and 1 represents worst. Thus, the maximum (best) score is 70 for the bowel symptoms dimension; 35 for the systemic symptoms dimension; 84 for the emotional dimension; and 35 for the social dimension. Reproducibility and Responsiveness Of the 61 patients who participated in the study of reproducibility and responsiveness, 48 had completed the Item Reduction Questionnaire. Of these 61 patients, 19 reported no change in their disease activity after 1 mo of observation, 33 improved, and 9 deteriorated. ReprodUcibility data came from the 19 patients whose self-reported global rating of disease activity was unchanged. Of these 19 patients, 9 were male, 10 female; 15 had ulcerative colitis and 4 had Crohn's disease; their mean age was 37 yr. Reproducibility encompasses two issues: whether there were systematic changes in scores over time, and the variability of the changes. The first issue is addressed in Table 1, which illustrates that scores on all four dimensions improved with time, the smallest change being seen in social. For two dimensions, bowel symptoms and systemic symptoms, a paired t-test demonstrated that the differ- Table 1. Repeated IBDQ Administration in 19 Subjects Whose Global Rating Showed No Change Over an Interval of 1 Month Baseline Follow-up Difference p Bowel symptoms Systemic symptoms Emotional Social Higher scores represent better.

4 March 1989 MEASURING HEALTH STATUS IN IBD 807 Table 2. Reproducibility and Responsiveness: Variability in IBDQ Score in 19 Subjects Whose Global Rating Showed No Change in Relation to Changes in IBDQ Score in 42 Subjects Who Improved or Deteriorated Standard deviation Bowel symptoms 3.7 Systemic symptoms 3.3 Emotional 6.9 Social 1.8 Coefficient of variation b Difference in patients who changed/ standard deviation 11.1/3.7 = /3.3 = /6.9 = /1.8 = 1.7 The standard deviation referred to is the standard deviation of the differences in 19 subjects whose global rating suggested they were stable. b The coefficient of variation is the standard deviation divided by the mean of the baseline and follow-up scores. ences between baseline and follow-up were unlikely to have occurred by chance. A second component of reproducibility is the variability in scores seen in stable subjects over time. We have used the standard deviation of the differences between baseline and follow-up in the 19 stable subjects to quantify reproducibility (Table 2). The standard deviation of the differences can be compared to the mean score across both administrations to yield a within-person coefficient of variation. The coefficient of variation for the four dimensions varied between 0.06 and We examined the responsiveness of the IBDQ using data from the 42 patients whose self-reported global ratings indicated change in their clinical state over the course of the 4-wk follow-up. Of the 42 patients, 22 were male and 20 female; 23 had ulcerative colitis and 19 had Crohn's disease; their mean age was 37 yr. Because the number of patients who deteriorated was too small to allow separate analysis, the sign of the difference score in every such patient was reversed and all patients were included in a single analysis. The results, shown in Table 3, demonstrate substantial consistent change in all four dimensions. Indeed, these changes are times as great as the increase in score seen in stable subjects. In each case, the probability value from a paired t-test revealed that the differences between baseline and follow-up were very unlikely to have occurred by chance. To determine whether questionnaire responsiveness may differ in patients with ulcerative colitis and Crohn's disease, we analyzed results from the 23 patients with ulcerative colitis and the 19 patients with Crohn's disease separately. The results, shown iii Table 4, suggest that although the questionnaire appears responsive in both groups, the magnitude of changes is greater in patients with ulcerative colitis. One important rationale for developing a disease- Table 3. Responsiveness: Changes in IBDQ Score Over a 1-Month Interval in 42 Subjects Whose Global Rating Showed Improvement or Deterioration Baseline Follow-up Difference Bowel symptoms <0.001 Systemic symptoms <0.001 Emotional <0.001 Social Rand physical Rand emotional The lower score in the questionnaire represents improvement in. specific measure is that it is likely to be more responsive than a general measure (31). We found that changes in Rand physical in the patients who changed did not reach conventional levels of statistical significance, and that the Rand emotional questionnaire proved less powerful than all four IBD dimensions (Table 3). A second measure of responsiveness relates the magnitude of the differences in subjects who are clinically changed to the standard deviation of the changes in clinically stable subjects (31). The larger the ratio, the smaller the sample size required in controlled trials of therapy. The ratio of the magnitude of change in IBDQ score in subjects whose global rating suggested change in disease activity to the standard deviation of change in IBDQ score in stable subjects is presented, for all four dimensions, in the final column of Table 2. For all four dimensions the ratio is greater than one and suggests that investigators attempting to detect changes with treatment of the magnitude we observed in our 42 subjects reporting improvement need to enroll less than 20 subjects per group in a parallel groups design Table 4. Responsiveness: Changes in IBDQ Score Over a 1-Month Interval in 23 Subjects With Ulcerative Colitis and 19 Subjects With Crohn's Disease Whose Global Rating Showed Improvement or Deterioration Baseline Follow-up Difference p Ulcerative colitis Bowel symptoms <0.001 Systemic <0.001 symptoms Emotional <0.001 Social Crohn's disease Bowel symptoms Systemic symptoms Emotional Social p

5 808 GUY ATT ET AL. GASTROENTEROLOGY Vol. 96, No.3 (31). Larger sample sizes will be required to detect smaller differences; nevertheless, these initial data suggest that the IBDQ is a powerful measure for detecting change in health status of IBD patients. Validity Data from the 42 subjects who reported change in their global rating of disease activity were used for validity assessment. Before examining our data, we made a number of predictions concerning the IBDQ findings we would expect if it is really measuring health status. These predictions are listed below and for each, the observed result follows. 1. The patient's global rating of change in disease activity should relate closely (correlation, >0.5) with change in the bowel symptoms dimension of the IBDQ. Correlation observed, 0.42; p = The patient's global rating of change in tiredness should relate closely (correlation, >0.5) to change in the systemic symptoms dimension of the IBDQ. Correlation observed, 0.36; p = The patient's global rating of change in emotional should relate closely (correlation, >0.5) to change in the emotional dimension of the IBDQ. Correlation observed: 0.52, p < The physician's global rating of change in IBD activity should relate moderately well (correlation, >0.4) to change in the bowel symptoms dimension of the IBD. Correlation observed, 0.30; p = The relative's global rating of change in IBD should relate moderately well (correlation, >0.4) to change in the bowel symptoms dimension of the IBDQ. Correlation observed, 0.38; p = The relative's global rating of change in tiredness should relate moderately well (correlation, >0.4) to change in the systemic symptoms dimension of the IBDQ. Correlation observed, 0.17; p = The relative's global rating of change in emotional should relate moderately well (correlation, >0.4) to change in the emotional dimension of the IBDQ. Correlation observed, 0.35; p = There should be some relation (correlation, >0.3) between change in the disease activity index and change in the bowel symptoms dimension of the IBDQ. Correlation observed, 0.33; p = There should be some relation (correlation, ;:::0.3) between change in the disease activity index and change in the systemic symptoms dimension of the IBD. Correlation observed, 0.036; p = Change in the emotional dimension of the IBDQ should relate closely (correlation, >0.5) with change in the emotional dimension of the Rand questionnaire. Correlation observed, 0.76; p < In general, the correlations were slightly lower than those we predicted. In 2 cases (relation between IBDQ systemic symptoms and both relative's global rating of fatigue and change in ulcerative colitis disease activity) the result was substantially different from that predicted. Discussion Crohn's disease and ulcerative colitis are chronic debilitating diseases with a significant impact on health status. We have developed a questionnaire based on reports from IBD patients about how their illness affects their lives. We have demonstrated that the variability of changes in IBDQ score when administered to patients whose global rating of clinical status suggested they were stable is small. The coefficient of variation in these stable subjects (Table 2) is smaller than the coefficient of variation seen in many questionnaires, and comparable to that seen in a number of physiologic measurements (32). However, mean IBDQ scores improved in these apparently stable patients. There are at least two ways these data can be interpreted. The first is to assume that the patients were indeed stable and that the changes observed were a of repeated questionnaire administration. A second possibility is that the global rating of change was unresponsive. This would be true if at least some of the 19 patients who reported no alteration in clinical status over the previous month had in fact improved. Were this the case, one would conclude that the IBDQ was better able to detect small degrees of improvement than was the global rating. It is not possible to definitively distinguish these possibilities on the basis of the present data, and further study of the properties of the IBDQ will be required. In the meantime we recommend at least two, and preferably three, practice administrations of the IBDQ before baseline data are obtained. We have demonstrated that the IBDQ is able to detect change in health status when change occurs, and that the instrument shows substantial correlation with other related measures. As predicted, proxy assessments by physicians and relatives showed weaker correlations with IBDQ scores than did patients' own global ratings. It is important to note the possibility of a training or learning effect that might show apparent improvement in scores in

6 March 1989 MEASURING HEALTH STATUS IN IBD 809 stable subjects, and the fact that the correlations with other measures, while substantial, were lower than predicted. The former problem should be ameliorated by several administrations of the questionnaire before obtaining baseline data. One may question the need for a detailed questionnaire such as the IBDQ for measuring health status in clinical trials in IBD. Would not a simple transition index, such as the global ratings of change we used in the present study, suffice? But simple global ratings of change have a number of disadvantages. First, single-item measures are inevitably less reproducible and valid than are questionnaires that rely on multiple items to tap an area such as bowel symptoms or emotional. An analogy can be drawn here to single vs. multiple measurements of clinical variables such as blood pressure. Second, a transition index cannot be used as a measure of outcome if the patient's status is to be sampled repeatedly. Such repeated sampling is crucial in trials that are longer than a few weeks in duration. In summary, our results, although limited by the small sample size of the study, suggest that the IBDQ may be of use in randomized trials in IBD. Further studies of reproducibility, validity, and responsiveness are required, particularly in patients with Crohn's disease. Appendix: Inflammatory Bowel Disease Questionnaire (IBDQ) Summary The IBDQ includes 32 questions. The wording is deliberately repetitious, as experience has taught us that the repetition ensures subjects' understanding. The questions are grouped into four categories: bowel symptoms, systemic symptoms, emotional, and social (SF). Response options are consistently presented as seven-point scales. An example of the way the questions are structured follows: 1. How frequent have your bowel movements been during the last two weeks? Please indicate how frequent your bowel movements have been during the last two weeks by picking one of the options from the WHITE card in front of you. 1 BOWEL MOVEMENTS AS OR MORE FREQUENT THAN THEY HAVE EVER BEEN 2 EXTREMELY FREQUENT 3 VERY FREQUENT 4 MODERATE INCREASE IN FREQUENCY OF BOWEL MOVEMENTS 5 SOME INCREASE IN FREQUENCY OF BOWEL MOVEMENTS 6 SLIGHT INCREASE IN FREQUENCY OF BOWEL MOVEMENTS 7 NORMAL, NO INCREASE IN FREQUENCY OF BOWEL MOVEMENTS The working structure of the other questions is identical, and appropriate seven-point scales are offered for each question. The content of the remaining 31 questions is as follows: 2. How often has the feeling of fatigue or of being tired and worn out been a problem for you during the last two weeks? 3. How often during the last two weeks have you felt frustrated, impatient, or restless? (SF) 4. How often during the last two weeks have you been unable to attend school or work because of your bowel problem? 5. How much of the time during the last two weeks have your bowel movements been loose? 6. How much energy have you had during the last two weeks? 7. How often during the last two weeks did you feel worried about the possibility of needing to have surgery because of your bowel problem? (SF) 8. How often during the last two weeks have you had to delay or cancel a social engagement because of your bowel problem? 9. How often during the last two weeks have you been troubled by cramps in your abdomen? 10. How often during the last two weeks have you felt generally unwell? 11. How often during the last two weeks have you been troubled because of fear of not finding a washroom? (SF) 12. How much difficulty have you had, as a result of your bowel problems, doing leisure or sports activities you would have liked to have done during the last two weeks? 13. How often during the last two weeks have you been troubled by pain in the abdomen? 14. How often during the last two weeks have you had problems getting a good night's sleep, or been troubled by waking up during the night? 15. How often during the last two weeks have you felt depressed or discouraged? (SF) 16. How often during the last two weeks have you had to avoid attending events where there was no washroom close at hand? 17. Overall, in the last two weeks, how much of a problem have you had with passing large amounts of gas? 18. Overall, in the last two weeks, how much of a problem have you had maintaining, or getting to, the weight you would like to be at? 19. Many patients with bowel problems often have worries and anxieties related to their illness. These include worries about getting cancer, worries about never feeling any better, and worries about have a relapse. In general, how often during the last two weeks have you had felt worried or anxious? 20. How much of the time during the last two weeks have you been troubled by a feeling of abdominal bloating? 21. How often during the last two weeks have you felt relaxed and free of tension? 22. How much of the time during the last two weeks

7 810 GUYATT ET A1. GASTROENTEROLOGY Vol. 96, No.3 have you had a problem with rectal bleeding with your bowel movements? 23. How much of the time during the last two weeks have you felt embarrassed as a result of your bowel problem? 24. How much of the time during the last two weeks have you been troubled by a feeling of having to go to the bathroom even though your bowels are empty? 25. How much of the time during the last two weeks have you felt tearful or upset? 26. How much of the time during the last two weeks have you been troubled by accidental soiling of your underpants? 27. How much of the time during the last two weeks have you felt angry as a result of your bowel problem? (SF) 28. To what extent has your bowel problem limited sexual activity during the last two weeks? 29. How much of the time during the last two weeks have you been troubled by feeling sick to your stomach? 30. How much of the time during the last two weeks have you felt irritable? 31. How often during the last two weeks have you felt lack of understanding from others? 32. How satisfied, happy, or pleased have you been with your personal life during the past two weeks? References 1. Best WR, Becktel JM, Singleton JW, Kern F. Development of a Crohn's Disease Activity Index; National Cooperative Crohn's Disease Study. Gastroenterology 1976;70: Harvey RF, Bradshaw JM. A simple index of Crohn's disease activity. Lancet 1980;i: Myren J, Bouchier lad, Watkinson G, et al. The OMGE Multinational Inflammatory Bowel Disease Survey A further report on 2657 cases. Scand J Gastroenterol 1984;29:95:S Van Hees PAM, Van Elteren PH, Van Lier HJJ, Van Tongeren JHM. An index of inflammatory activity in patients with Crohn's disease. Gut 1980;21: De Dombal FT, Softley A. IOIED Report No.1: Observer variation in calculating indices of severity and activity in Crohn's disease. Gut 1987;28: Kirshner B, Guyatt G. A methodological framework for assessing health indices. J Chronic Dis 1985;38: Kroner K. Are you prepared for your ulcerative colitis patient? Nursing 1980;20: Sparacino L1. Psychosocial considerations for the adolescent and young adult with inflammatory bowel disease. Nurs Clin North Am 1984;19: Sutherland D. Ulcerative colitis. Nursing Times 1982: Murray JB. Psychological factors in ulcerative colitis. J Genet Psychol 1984;110: Schoenberg P. An experience of ulcerative colitis. Br J Psychiatry 1983;143: Rigatelli M. A global psychosomatic study of 16 consecutive patients with ulcerative colitis. Psychother Psychos om 1981;35: Olbrisch ME, Ziegler SW. Psychological adjustment to inflammatory bowel disease: informational control and private self-consciousness. J Chronic Dis 1982;35: Farmer RG, Whelan G, Fazio VW. Long-term follow-up of patients with Crohn's disease. Gastroenterology 1985;88: Lind E, Fausa 0, Gjone E, Mogensen SB. Crohn's disease; treatment and outcome. Scand J Gastroenterol 1985;20: Krause U, Ejerblad S. Bergman L. Crohn's disease; a long-term study of the clinical course in 186 patients. Scand J GastroenteroI1985;20: Joachim G, Milne B. The effects of inflammatory bowel disease on lifestyle. Canadian Nurse 1985;81: Gazzard BD, Price HL, Libby GW, Dawson AM. The social toll of Crohn's disease. Br Med J 1978;2: Hendriksen C, Binder V. Social prognosis in patients with ulcerative colitis. Br Med J 1980;281: Mallett SJ, Lennard-Jones JE. Bingley I. Gilon E. Colitis. Lancet 1978: Sales DJ, Kirsner JB. The prognosis of inflammatory bowel disease. Arch Intern Med 1983;143: Sorensen VZ, Olsen BG. Binder V. Life prospects and quality of life in patients with Crohn's disease. Gut 1987;28: Bergner M, Bobbitt RA, Carter WB, Gilson BS. The Sickness Impact Profile: development and final revision of a health status measure. Med Care 1981;19: Ware JE, Brook RH, Davies-Avery A, et al. Conceptualization and measurement of health for adults in the health insurance study. Volume 1. Model of health and methodology. Santa Monica. Calif.: Rand Corporation. May Spitzer WO, Dobson AI. Hall J, et al. Measuring the quality of life of cancer patients. J Chronic Dis 1981;34: Liang MH, Larson MG, Cullen KE. Schwartz JA. Comparative measurement efficiency and sensitivity of five health status instruments for arthritis research. Arthritis Rheum 1985;28: Guyatt GH, Berman LB, Townsend M, Pugsley SO, Chambers LW. A measure of quality of life for clinical trials in chronic lung disease. Thorax 1987;42: Guyatt GH, Bombardier C. Tugwell PX. Measuring dis,easespecific quality of life in clinical trials. Can Med Assoc J 1986;134: Powell-Tuck J, Bown RL, Lennard-Jones JE. A comparison of oral prednisone given as single or multiple daily d o ~ for e s active proctocolitis. Scand J Gastroenterol 1978;13: Mitchell A. Guyatt GH, Singer J, et al. Quality ohife in patients with inflammatory bowel disease. J Clin Gastroenterol 1988;10: Guyatt GH, Walter S, Norman G. Measuring change over time: assessing the usefulness of evaluative instruments. J Chronic Dis 1987;40: Guyatt GH, Thompson PI. Berman LB, et al. How should we measure in patients with chronic heart and lung disease? J Chronic Dis 1985;38: Received November 23, Accepted October Address requests for reprints to: Dr. Gordon Guyatt, Department of Clinical Epidemiology and Biostatistics, McMaster University Health Sciences Centre, Room 3H7, 1200 Main Street West. Hamilton, Ontario, Canada, L8N 3Z5. This study was supported in part by the Canadian Foundation for Ileitis and Colitis. Dr. Guyatt is a career scientist of the Ontario Ministry of Health. The authors thank Marysia Donnelly, Helen Pikula, and Drs. R.A. Chambers, C.S. Kumaranayake. T.1. Seaton, M.E. Castelli. S.M. Collins, R.H. Hunt, B.J. Lumb. R.L. Rossman, F.G. Saibil, and E.J. Prokipchuk for their help in recruiting patients for this study.