- Email: [email protected]
Inflammatory bowel disease: developing a short disease specific scale to measure health related quality of life
International Journal of Nursing Studies 39 (2002) 583–590
Inflammatory bowel disease: developing a short disease specific scale to measure health related quality of life G.D. Smitha,*, R. Watsonb, K.R. Palmera a
Gastrointestinal Unit, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, UK b Faculty of Health, School of Nursing, University of Hull, Hull, UK
Received 30 June 2000; received in revised form 16 February 2001; accepted 11 July 2001
Abstract Aim: To report on the preliminary development of a short scale, the Edinburgh inflammatory bowel disease questionnaire (EIBDQ) to measure disease specific aspects of inflammatory bowel disease (IBD) in terms of both physical impact and health related quality of life (HRQoL) consequences. Methods: A survey of individuals with Crohn’s disease (n ¼ 50), ulcerative colitis (n ¼ 50) and psoriatic arthritis (n ¼ 28) was carried out using the EIBDQ. The data were subject to factor analysis in order to investigate underlying dimension of the EIBDQ which were then analysed for internal consistency. Data for disease specific aspects of the EIBDQ were compared between IBD patients and psoriatic arthritis patients using contingency tables and the underlying dimension of the EIBDQ were correlated with measures of quality of life and psychological morbidity and a disease specific measure. Results: There are three underlying dimensions to the EIBDQ: a disease specific factor, a bowel specific factor and an information factor. The disease specific factor and the bowel specific factor are internally consistent and correlate with other measures of disease activity, quality of life and psychological morbidity. The EIBDQ is able to discriminate between IBD and another inflammatory disease: psoriatic arthritis. Conclusion: The EIBDQ is a reliable and valid instrument for measuring disease specific aspects of IBD but further development is required. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Inflammatory bowel disease; Crohn’s disease; Ulcerative colitis; Chronic illness; Health related quality of life; Factor analysis
1. Introduction Inflammatory bowel disease (IBD) is a general term for two idiopathic chronic diseases: Crohn’s disease (CD) and ulcerative colitis (UC) which are distinct, but related, inflammatory disorders of the gastrointestinal tract. IBD has been shown to have physical and psychosocial implications which can adversely effect the health related quality of life (HRQoL) of the sufferer (Drossman, 1994). The aetiology of both CD and UC is *Corresponding author. Tel.:+44-131-537-1756; fax: +44-131-537-1007. E-mail address: [email protected] (G.D. Smith).
largely unknown although genetic, environmental, lifestyle and dietary factors have all been implicated. Disease specific questionnaires are viewed as the most effective type of HRQoL instrument to define fluctuations in health status (Guyatt, 1989). One such tool, the inflammatory bowel disease questionnaire (IBDQ) has been developed and validated and shown to be of value in the assessment of disease outcomes in patients with IBD. The purpose of this paper is to report on the preliminary development of a disease specific scale for the measurement of HRQoL in IBD. The relevance of this study comes from the fact that nurses working in the speciality of gastroenterology are in an ideal position, if they possess the necessary knowledge and skills, to deal
0020-7489/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 7 4 8 9 ( 0 1 ) 0 0 0 4 2 - 6
584
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
with both the physical and psychosocial aspects of chronic illness, such as IBD (Smith, 1997). The development of the scale reported here has been nurseled and it is necessary for nurses to familiarise themselves with the issues surrounding measurement of human attributes and attitudes if they are to implement and gauge the effectiveness of treatment for both the physical and psychosocial aspects of chronic illness. Quality of life is a vague term which is difficult to define. However, it has been described by Shin and Johnson (1978, p. 475) in the following terms: the possession of the resources necessary to the satisfaction of individual needs, wants and desires, participation in activities enabling personal development and self-actualisation and satisfactory comparison between oneself and others. The above provides a very broad definition and one which encapsulates more factors than are probably useful in measuring the impact of a condition such as IBD on the quality of life of the individual. Indeed, it has been demonstrated that only part of quality of life is health related (McGee et al., 1991) and the importance of health aspects increase in those individuals with chronic disease, this has led to the emergence of the concept of HRQoL (Berg, 1975). Following acceptance and acknowledgement by Index Medicus in 1977 there has been an increase in interest of measurement of HRQoL within the medical profession (Bech, 1992). The relevance of HRQoL measurement is that it may provide information on patient population needs, health care delivery requirements, the nature of chronic disease, therapeutic efficacy in clinical trials and as an essential constituent of cost-utility analyses (Fletcher et al., 1992). All quality of life and HRQoL measurement scales require rigorous evaluation prior to their acceptance as valid diagnostic or evaluative tools in disease assessment. Psychometric assessment is the evaluation of the quality of a scale in terms of its reliability, validity, sensitivity and specificity (Bryman and Cramer, 1994; Polit and Hungler, 1995). 1.1. Generic and/or disease specific scales in IBD? Controversy exists over the use of either generic or disease specific questionnaires for the measurement of HRQoL in IBD (Garrett et al., 1990). Generic scales have been seen to have a place in the assessment of chronic disease but they are not always directly applicable and have limited insight into specific conditions (Drummond, 1987). Generic scales do however offer the possibility of making comparisons across disease categories and they are also more likely to fulfil psychometric criteria of reliability and validity.
Despite this, the role of generic scales has received criticism mainly on the grounds of specificity (McKenna, 1993). By their very nature generic instruments will contain irrelevant questions and while the scale may be modified by omitting or ignoring the results of irrelevant questions, generic scales will always lack the specificity of a disease specific scale. Disease specific scales, therefore, have the advantage of specificity because they will have been developed specifically for the condition under study. The value of disease specific questionnaires used in conjunction with generic measures has been proposed by Guyatt et al. (1989a, b).
1.2. Disease specific scales for measuring HRQoL in IBD Disease specific measures used in IBD, such as the Crohn’s disease activity index (CDAI) measure the physical impact of IBD rather than measures of HRQoL (Garrett and Drossman, 1990). The CDAI places emphasis upon the objective markers of disease which may not always be the best indicator of patient well-being. This is not to say that the CDAI is clinically unimportant, disease indices provide the basis for many clinicians’ assessment and management of IBD (Best et al., 1976). It has, however, been recognised in recent years that more attention should be paid to more subjective assessment of psychosocial factors in chronic illness. Indeed there is no clear relationship between laboratory data, such as haematocrit, ESR and overall functional ability. Drossman et al. (1990) highlighted the importance of combining physical disease assessment with psychosocial factors because they define different aspects of medical care. Interest in disease specific HRQoL scales developed as a result of the theory that sufferers’ psychological status influences their response to chronic conditions (Talal and Drossman, 1995) and that sufferers’ psychological and social disability in IBD may be at least as important as physical disability. In part, this is supported by the observation in IBD that objective markers of disease are not clearly related to functional ability (Drossman, 1994). Kunsebeck et al. (1990) supported this theory and concluded that, while objective measures of socio-economic status and disease activity are easy to measure, they lack insight into the personal impact of IBD. Early studies of HRQoL in IBD were methodologically flawed in that they were largely observational, uncontrolled and did not use reliable and validated measures (Mitchell et al., 1988). The outcome of these studies was an indication of the need to quantify psychosocial aspects of HRQoL. Since HRQoL depends upon the sufferer’s perspective, which is subjective and multidimensional, including such aspects as well-being, daily functioning, psychological state, job satisfaction and interpersonal relationships, the necessity for vali-
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
dated questionnaires is indicated. Several scales have therefore been developed to measure HRQoL in IBD. Initial attempts to measure HRQoL in IBD used crude assessment scales asking patients to rate their HRQoL as being ‘good’, ‘fair’ or ‘poor’ (Farmer et al., 1985). Early studies focused on dysfunction following surgery for IBD, these were cross-sectional studies, taking only a ‘snapshot’ of the condition and making no attempt to study the impact of the disease over time using follow-up studies (Irvine, 1995). Drossman et al. (1991a, b) evaluated the health status of a national sample of patients with IBD who were members of the Crohn’s and Colitis Foundation of America. These sufferers were a self-selected group with a range of disease severity who were sufficiently concerned about their condition to join this organisation. CD sufferers had poorer health status and more occupational disability than UC sufferers. This finding was supported by Farmer et al. (1992) who, using the Cleveland Clinic Questionnaire, used a direct-interview technique with IBD sufferers and found HRQoL to be poorer in CD than in those with UC. An attempt has been made to differentiate between the concerns of CD and UC sufferers and this has been carried out using the Rating Form of IBD Patient Concerns (RFIPC) which is a specialised questionnaire identifying the 25 most important worries and concerns of IBD patients (Drossman et al. (1991a, b). The RFIPC contains four dimensions including disease-related, body related, inter/intrapersonal and sex related which are measured on visual analogue scales. The RFIPC was applied by Drossman et al. (1989) and the results indicated that the major problems for IBD sufferers were related to surgery, energy/fatigue and then body image and psychosexual consequences of having a stoma. They concluded that CD and UC patients had similar concerns except that UC patients were more concerned with loss of bowel control and the development of cancer whilst CD patients were most concerned with pain and fertility issues. The results from the RFIPC correlated well with other measures of health status including the clinicians’ global assessment, functional impairment and number of hospitalisations (Drossman, 1992). However, the RFPIC demonstrated flaws related to the content validity of the questionnaire and in relation to lack of reliability. Love et al. (1992) developed the UC/CD Disease Health Status Scale which was designed to discriminate mild from moderate illness and to measure health care use, daily function and psychological distress. Results indicated that quality of life in UC and CD patients was worse than in a group of matched controls. The IBDQ was developed because it was felt that previous questionnaires had not adequately addressed the psychosocial aspects of the disease.
585
The IBDQ examines bowel disturbance, systemic effects, emotional and social function. The psychometric properties of the IBDQ appear to be good in terms of reliability and validity and it correlates well with objective measurement of disease activity. However with 32 items its length has placed practical limitations upon its use within out-patient departments where the vast majority of IBD patients are reviewed. Therefore development of a simpler, shorter instrument, which is less cumbersome to administer may provide benefits in the assessment and management of IBD patients. A shorter ten item IBDQ called the shortened inflammatory bowel disease questionnaire (SIBDQ) was formulated and found to correlate well with both IBDQ and CDAI (Irvine et al., 1996). Further to this Turnbull et al. (1996) have described an even shorter questionnaire containing five items, claiming the instrument which has been designed for use in the out-patient environment to be as reliable and as sensitive as the original 32 item IBDQ. It should also be noted that all the work reported above is North American in origin and a perspective on IBD from outside that context would be useful. 1.3. Development of the EIBDQ Despite the above developments it is reported that an ideal disease specific scale for measurement of HRQoL in IBD in the United Kingdom remains to developed. Such developments would permit more specific assessment of IBD related aspects of HRQoL and also comparison with other chronic diseases whilst retaining sensitivity to treatment effects (Irvine, 1995). Schipper et al. (1990) provide a concise yet adequate HRQoL as ‘‘the functional impact of an illness, and its consequent therapy, upon the patient, as perceived by the patient’’. The EIBDQ was developed, specifically, to encapsulate the above. The EIBDQ contains fifteen questions based upon agreement–disagreement scales (Atkinson, 1991), no copyright is reserved on this scale. The factor structure of the EIBDQ was examined and construct validity was examined using established and validated measures of disease activity and psychological morbidity in IBD. A comprehensive review of the IBD literature and a pilot study of a preliminary version of the EIBDQ resulted in the formulation of the finalised questionnaire which is shown in Table 1. 1.4. Present study The present study was designed to develop and undertake preliminary psychometric testing of a disease specific scale to measure HRQoL in IBD. Turnbull et al. (1996) have indicated the need to develop more user friendly scales to measure facets of HRQoL in IBD and a review of other relevant measures of disease activity
586
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
Table 1 Edinburgh (EIBDQ) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Inflammatory
Bowel
Disease
Questionnaire
Do you feel that sufficient information was given to you about your disease? Are you familiar with the National Association of Colitis and Crohn’s disease (NACC)? Over the past two weeks have you had diarrhoea? Over the past two weeks have you had tummy pain? Over the past two weeks have you been overly tired? Over the past two weeks have you had any joint pain? Do you feel that you have full control over your bowel function? Do you ever have to rush to the toilet in regarding your bowels? Do you ever have accidents regarding your bowels? Does your bowel function affect your relationship at home? Does your bowel function affect your social life? Does your bowel function affect any recreational activities you normally do? Do you feel your recreational activities are restricted by your disease? Do you feel your disease has affected your relationships at home? Do you feel your overall ‘quality of life’ is affected by your disease?
Responses to Items 1, 2, 7, 8, 9, 10, 11 and 12=‘‘Yes’’ and ‘‘No’’. Responses to Items 3, 6, 13, 14, 15 and 16=‘‘Not at all’’, ‘‘A little’’, ‘‘Moderately’’ or ‘‘Severely’’.
and psychological morbidity have been presented. The design of the present study was by means of a survey which utilised the EIBDQ with a sample of IBD sufferers and psoriatic arthritics who were used to represent another group of sufferers of a chronic inflammatory condition.
2. Methods One hundred and twenty-eight subjects, 50 CD (age range 23–82; 38 female), 50 UC (age range 19–78; 37 female) and 28 psoriatic arthritics (age range 17–62; 27 female) were recruited into the study. The CD and UC subjects were recruited from an out-patient gastrointestinal clinic and psoriatic arthritics were recruited from a rheumatology clinic within the same hospital. The following questionnaires were administered to all subjects: a modified CDAI, HAD, SF-36, SS (styles and strategies questionnaire), and the EIBDQ.
medication and also objective clinical measures such as the haematocrit. The modified version used only those aspects which the patients could self-report. The HAD has 14 questions and two scales, one for anxiety and one for depression, it is simple to use and relatively insensitive to physical symptoms. The SF-36 comprises 36 questions measuring eight dimensions of HRQoL including physical function, social function, mental health, role limitation due to physical problems, energy/vitality, role limitations due to emotional problems, pain and health perception. The scale was developed in America but a modified United Kingdom version exists (Brazier et al., 1972). The psychometric properties of the SF-36 were established by Garratt et al. (1993) showing good construct validity and factor analysis supports the underlying dimensions. The SS questionnaire, developed by Roger et al. (1993) is a 16 item questionnaire which measures adaptive and maladaptive styles of coping in terms of being detached and rational, on the one hand, and emotional and avoiding, on the other. The EIBDQ has been described above. Data from the above questionnaires were entered into a database for analysis using SPSS for Windows. The following statistical tests were applied: Pearson’s Product Moment Correlation Coefficient and Cronbach’s alpha to measure reliability using internal consistency of multiple item scales or factors. Principal components analysis with oblique rotation was applied as described by Watson and Deary (1994) to the responses in the EIBDQ, with further manoeuvres to eliminate redundant questions and reduce the number of factors while accounting for as much of the variance as possible as described by Agius et al. (1996). Oblique rotation was selected in preference to Varimax rotation as this obtained the simplest solution with highest loadings on putative factors. Using this application Item 6 of the EIBDQ, which relates solely to joint pain, was omitted and Item 15 was omitted because it may have been misinterpreted due to ambiguity surrounding ‘quality of life’. The omission of both these item from the factor analysis improved the structure of the derived matrix (Table 3). The disease specific nature of selected items from the EIBDQ which were specific to either IBD or psoriatic arthritis (PA) was tested using the chi square test of proportion. The chi-square test of proportions was used to analyse the association between symptoms of IBD in IBD and PA patients.
3. Results 3.1. Disease specificity
2.1. Instruments The CDAI is an eight item instrument including questions on pain, stool consistency and use of
Table 2 shows selected items from the EIBDQ related to symptomatic reporting of diarrhoea, abdominal pain, tiredness and joint pain. Diarrhoea, abdominal pain and
587
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590 Table 2 Response to EIBDQ items specific and non-specific to IBD in IBD and PA patients
Item 3 (diarrhoea) IBD PA Item 4 (abdominal pain) IBD PA Item 5 (tiredness) IBD PA Item 6 (joint pain) IBD PA
Absent
Present
Chisquared
d.f.
31 26
79 2
27
2
o0.05
29 26
71 2
36
2
o0.05
8 13
92 15
66
2
o0.05
39 1
61 27
12
2
o0.05
p
taken to indicate reliability. The intercorrelation between the three rotated factors is very low (Table 3) indicating that, despite the use of oblique rotations, the factors were representing different underlying dimensions of the EIBDQ. 3.4. Validity
d.f.=degrees of freedom. CD and UC sufferers were collapsed into one category of IBD sufferers and responses to the IBDQ were dichotomised into ‘‘absent’’ and ‘‘present’’, the latter category being created by collapsing the ‘‘mild’’, ‘‘moderate’’ or ‘‘severe’’ responses to items.
tiredness which are all recognised as symptoms of IBD, are more prevalent in both CD and UC patients than in those suffering from psoriatic arthritis. Conversely, as shown in Table 2, joint pain is much more prevalent in sufferers of psoriatic arthritis than either group of IBD sufferers. The association between disease state and item responses is statistically significant (po0:05) using the chi square test of proportions.
In order to demonstrate convergent validity, the putative factors were correlated with validated measures of physical and psychological morbidity in IBD. Factor scores were calculated by summing the scores for all items loading on the factors. The correlation matrixes shown in Tables 4 and 5 show the intercorrelation of the bowel specific factor (Factor 2) with the CDAI, the two dimensions of the HAD and one dimension from each of the SS questionnaire (emotional coping) and SF-36 (physical function) for CD and UC respectively. Factor scores for the bowel specific factor were generated by summing the scores for all items loading on Factor 2. Significant correlations (po0:05) above 0.3 are shown in bold as these represent at least a moderate correlation between variables. For both CD and UC Factor 2 correlated with the CDAI score. Factor 2 also correlated with the HAD score for anxiety in the case of CD. The SF-36 dimension related to physical role function correlated with Factor 2 for both CD and UC and the Table 3 Principal components analysis of the EIBDS data followed by oblique rotation Factor 1 (Disease specific)
Facto 2 (Bowel specific)
Factor 3 (Information)
14 13 12 4 5
0.811 0.775 0.702 0.640 0.581
0.030 0.035 0.035 0.113 0.001
0.006 0.006 0.078 0.098 0.032
10 8 9 7 11 3
0.119 0.050 0.023 0.072 0.153 0.362
0.860 0.846 0.833 0.800 0.778 0.408
0.130 0.182 0.032 0.173 0.167 0.114
0.029 0.051 0.79
0.831 0.787 0.55
3.3. Reliability
2 0.193 1 0.205 Cronbach’s a 0.86 Correlation 1 1 2 0.24 3 0.15
Values of Cronbach’s alpha for the three rotated factors are also shown in Table 3, and were acceptable for Factors 1 and 2, a Cronbach’s alpha of 0.8 being
Factor loading greater than 0.400 are shown in bold with the internal consistency of the rotated factors (Cronbach’s a) and the correlation between factors being presented at the foot of the table.
3.2. Factor analysis The results of principal components analysis followed by oblique rotation showed three putative factors, and these factors are, respectively, a non-bowel disease specific factor (Factor 1), a bowel specific factor (Factor 2) and an information related factor (Factor 3). The loadings of variables which are greater than 0.4 are shown in bold in Table 3. The percentage of variance extracted by each factor, before rotation was 39%, 14% and 10%, respectively.
Item
1 0.04
1
588
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
Table 4 Correlation matrix of EIBDS Factor 2 with CDAI, HAD, SS and selected items for SF-36 for Crohn’s disease patients (n=50)
EIBDS Factor 2 Crohn’s activity index HAD (anxiety) HAD (depression) SS (emotional) SF-36 (physical function)
EIBDS Factor 2
Crohn’s activity index
HAD (anxiety)
HAD (depression)
SS (emotional)
SF-36 (physical function)
1 0.52 0.16 0.55 0.23 0.48
1 0.24 0.37 0.04 0.35
1 0.45 0.36 0.06
1 0.07 0.23
1 0.04
1
Correlations greater than 0.3 and po0.05 are shown in bold. See text for explanation of abbreviations used in the table.
Table 5 Correlation matrix of EIBDS Factor 2 with CDAI, HAD, and selected items for SF-36 for ulcerative colitis patients (n=50)
EIBDS Factor 2 Crohn’s activity index HAD (anxiety) HAD (depression) SS (emotional) SF-36 (physical function)
EIBDS Factor 2
Crohn’s activity index
HAD (anxiety)
HAD (depression)
SS (emotional)
SF-36 (physical function)
1 0.30 0.40 0.43 0.30 0.32
1 0.42 0.13 0.45 0.16
1 0.45 0.65 0.06
1 0.25 0.35
1 0.01
1
Correlations greater than 0.3 and po0.05 are shown in bold. See text for explanation of abbreviations used in the table.
SS questionnaire item related to emotional coping with stress correlated with Factor 2 in UC sufferers only.
4. Discussion The purpose of the work described in this paper was the development of a short, user-friendly disease specific scale to measure HRQoL in IBD. Although the IBDQ is widely accepted as the ‘gold standard’ in the measurement of HRQoL in IBD (Irvine, 1993), it is commonly accepted that there is the need for a shorter more practical questionnaire which can be easily administered in the out-patient environment. Whilst acknowledging that generic scales have a place in the measurement of HRQoL in IBD in terms of enabling comparisons to be made across diseases, there is an established requirement for scales which will specifically measure and reflect the impact of HRQoL in patients with IBD. The study was carried out by reviewing the literature for dimensions of HRQoL which appear to be affected by IBD and incorporating these with other aspects of IBD into a new instrument called the EIBDQ. A range of multivariate statistical techniques was employed to establish the specificity, reliability and convergent validity of the EIBDQ.
This study has identified a disease specific dimension of the EIBDQ, as demonstrated by factor analysis, from other aspects of IBD such as bowel activity and the provision of information to IBD patients. It was possible to discriminate between IBD sufferers and sufferers of another inflammatory condition, psoriatic arthritis, by means of individual disease specific items in the EIBDQ and the HRQoL related items of the EIBDQ were more prevalent in IBD patients than in patients with psoriatic arthritis. The reliability of the bowel specific dimension of the EIBDQ (Factor 2) is reliable as demonstrated by Cronbach’s alpha. Factor 2 correlates well with a reliable and valid measure of depression and for UC patients and it also correlates well with a reliable and valid measure of anxiety and depression. In both cases the HAD was used to measure anxiety and depression (Table 4). Likewise, an important dimension of the SF36 correlates with the bowel activity specific as does the CDAI which is an established objective measure of physical disease activity in IBD (Table 5). The correlations suggest that the bowel activity specific scale (Factor 1) has construct validity in terms of other measures of psychosocial, HRQoL and disease activity measures. Correlations greater than 0.3 are taken to represent at least moderate correlations between variables (Bryman and Cramer, 1996).
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
The work reported here is preliminary in the sense that any scale can be improved and other aspects of the questionnaire from which the bowel activity specific scale is derived may also be improved to provide a comprehensive measure of the full impact of IBD upon HRQoL. Further development could proceed by including new items, particularly in the dimension of information provision, but also to the domain of the EIBDS from which the bowel specific factor is derived. The importance of developing such disease specific scales extends beyond that of mere research and surveying the impact of disease upon HRQoL. If such scales are sufficiently sensitive, and further research and development could be aimed at investigating this aspect, then it could be used in further research into the efficacy of treatment in IBD. This could include treatment aimed at alleviating the underlying condition and psychosocial interventions which are aimed at the manifestations of IBD, especially upon HRQoL, such as counselling.
References Agius, R., Blenkin, H., Deary, I., Zealley, H., Wood, R., 1996. Survey of perceived stress and work demands of consultant doctors. Occupational and Environmental Medicine 53, 217–224. Atkinson, F.I., 1991. Survey Design and Sampling. In: Cormack, D.F.S. (Ed.), The Research Process in Nursing, 2nd Edition.. Blackwell Science Publication, Oxford, pp. 196–206 (Chapter 21). Bech, P., 1992. Measuring quality of life: the medical perspective. Nordic Journal of Psychiatry 46, 85–89. Berg, O., 1975. Health and qulaity of life. Acta Sociologica 18, 3–22. Best, W.R., Becktel, J.M., Singleton, J.W., Kem, F., 1976. Development of a crohn’s disease activity index. Gastroenterology 70, 439–444. Brazier, J.E., Harper, R., Jones, N.M.B., O’Cathain, A., Thomas, K.J., Usherwood, T., 1972. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. British Medical Journal 305, 160–164. Bryman, A., Cramer, D., 1994. Quantitative Data Analysis for Social Scientists, Routledge, London, pp. 152–190. Drossman, D.A., 1994. Psychosocial factors in ulcerative colitis and Crohn’s disease In: Kirnser, J.B. (Ed.), Inflammatory Bowel Disease, Leah & Febriger, Philadelphia, PA. Drossman, D.A., Leserman, L.Z., 1992. Ulcerative colitis and Crohn’s disease health status scales for research and clinical practice. Journal of Clinical Gastroenterology 15, 104–107. Drossman, D.A., Patrick, D.L., Mitchell, C.M., Zagami, E.A., Applebaum, M.I, 1989. Health related quality of life in inflammatory bowel disease: functional status and patient concerns. Digestive Disease and Science 34, 1379–1380. Drossman, D.A., Mitchell, C.M., Leserman, J., Patrick, D.L., Zagami, E.A., 1990. Psychosocial and disease predictors of utilization in IBD. Gastroenterology 98 (Abs). Drossman, D.A., Leserman, J., Li, Z., 1991a. Health status and health care use in persons with inflammatory bowel disease:
589
a national survey. Digestive Disease and Science 36, 1746– 1749. Drossman, D.A., Leserman, J., Li, Z., 1991b. The rating form of IBD patient concerns: a new measure of health status. Psychosomatic Medicine 53, 701–702. Drummond, M.F., 1987. Resource allocations and decisions in health care: a role for quality of life assessments. Journal of Chronic Disease 40, 605–616. Farmer, R., Whelan, G., Fazio, V.W., 1985. Long term followup of patients with crohn’s disease. Relationship between the clinical pattern and prognosis. Gastroenterology 88, 1818–1825. Farmer, R.G., Easley, K.A., Farmer, J.M., 1992. Quality of life assessment with patients with inflammatory bowel disease. Cleveland Clinical Journal of Medicine 59, 35–52. Fletcher, A., Gore, S., Jones, D., Fitzpatrick, R., Spiegelhalter, D., Cox, D., 1992. Quality of life measures in health care. II: design, analysis, and interpretation. British Medical Journal 305, 1145–1148. Garrett, J.W., Drossman, D.A., 1990. Health status in inflammatory bowel disease. Gastroenterology 99, 90–91. Garratt, A.M., MacDonald, L.M., Ruta, D.A., Russell, I.T., Buckingham, J.K., Krukowski, Z.H., 1993. Towards the measurement of outcomes for patients with varicose veins. Quality of life in Health Care 2, 5–10. Guyatt, G., Mitchell, A., Irvine, J., Singer, J., Williams, N., Goodacre, R., Tompkins, C., 1989a. A new measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology; 96, 804–810. Guyatt, G., Mitchell, A., Irvine, E., Singer, J., Williams, N., Goodacre, R., Tomkins, C., 1989b. A measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology 96, 804–810. Irvine, E.J., 1993. A quality of life index for IBD. Canadian Journal of Gastroenterology 7, 155–159. Irvine, E.J., 1995. Quality of life in inflammatory bowel disease: biases and other factors affecting scores. Scandanavian Journal of Gastroenterology 30, 136–140. Irvine, E.J., Zhou, Q., Thomson, M.B.A., 1996. The short inflammatory bowel disease questionnaire: a quality of life instrument for community physicians managing inflammatory bowel disease. American Journal of Gastroenterology 91, 1571–1578. Kunsebeck, H.W., Korber, J., Freyberger, H., 1990. Quality of life in patients with inflammatory bowel disease. Psychotherapy and Psychosomatic Research 29, 23–31. Love, J.R., Irvine, E.J., Fedorak, R.N., 1992. Quality of life in inflammatory bowel disease. Journal of Clinical Gastroenterology 14, 15–19. McGee, H.M., O’Boyle, C.A., Hickey, A., et al., 1991. Assessing the quality of life of the individual: the seiqol with a healthy and gastroenterology unit population. Psychological Medicine 21, 749–758. McKenna, S.P., 1993. The Nottingham Health Profile. Proceedings of the Fifth European Health Service Research Conference, Maastricht, Netherlands, Unpublished paper. Mitchell, A., Guyatt, G., Singer, J., 1988. Quality of life in patients with inflammatory bowel disease. Journal of Clinical Gastroenterology 10, 306–310. Polit, D.F., Hungler, B.P., 1995. Nursing Research: Principles and Methods. Lippincot, Philadelphia, PA.
590
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
Roger, D., Jarvis, G., Najarian, B., 1993. Detachment and coping: the construction and validation of a new scale for measuring coping strategies. Personality and Individual Differences 15, 619–626. Schipper, H., Clinch, J., Powell, V., 1990. Definitions and conceptual issues. In: Spiker, B. (Ed.), Quality of Life Assessment in Clinical Trials. Raven Press, New York, pp. 5–24. Shin, D.C., Johnson, D.M., 1978. Avowed hapiness as an overall assessment of the quality of life. Social Indicators Research 5, 475–492. Smith, G.D., 1997. Counselling in inflammatory bowel disease. Unpublished PhD thesis, University of Edinburgh.
Talal, A.H., Drossman, D.A., 1995. Psychosocial factors in inflammatory bowel disease. Gastroenterology Clinics of North America 24, 173–181. Turnbull, G.K., Vallis, T.M., Luyendyk, Y., 1996. The assessment of quality of life in inflammatory bowel disease patients: a five item scale for out-patient use. Gut 39, 16–17. Watson, R., Deary, I.J., 1994. Measuring feeding difficulties in patients with dementia: multivariate analysis of feeding problems, nursing intervention and indicators of feeding difficulty. Journal of Advanced Nursing 20, 283– 287.
Inflammatory bowel disease: developing a short disease specific scale to measure health related quality of life G.D. Smitha,*, R. Watsonb, K.R. Palmera a
Gastrointestinal Unit, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, UK b Faculty of Health, School of Nursing, University of Hull, Hull, UK
Received 30 June 2000; received in revised form 16 February 2001; accepted 11 July 2001
Abstract Aim: To report on the preliminary development of a short scale, the Edinburgh inflammatory bowel disease questionnaire (EIBDQ) to measure disease specific aspects of inflammatory bowel disease (IBD) in terms of both physical impact and health related quality of life (HRQoL) consequences. Methods: A survey of individuals with Crohn’s disease (n ¼ 50), ulcerative colitis (n ¼ 50) and psoriatic arthritis (n ¼ 28) was carried out using the EIBDQ. The data were subject to factor analysis in order to investigate underlying dimension of the EIBDQ which were then analysed for internal consistency. Data for disease specific aspects of the EIBDQ were compared between IBD patients and psoriatic arthritis patients using contingency tables and the underlying dimension of the EIBDQ were correlated with measures of quality of life and psychological morbidity and a disease specific measure. Results: There are three underlying dimensions to the EIBDQ: a disease specific factor, a bowel specific factor and an information factor. The disease specific factor and the bowel specific factor are internally consistent and correlate with other measures of disease activity, quality of life and psychological morbidity. The EIBDQ is able to discriminate between IBD and another inflammatory disease: psoriatic arthritis. Conclusion: The EIBDQ is a reliable and valid instrument for measuring disease specific aspects of IBD but further development is required. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Inflammatory bowel disease; Crohn’s disease; Ulcerative colitis; Chronic illness; Health related quality of life; Factor analysis
1. Introduction Inflammatory bowel disease (IBD) is a general term for two idiopathic chronic diseases: Crohn’s disease (CD) and ulcerative colitis (UC) which are distinct, but related, inflammatory disorders of the gastrointestinal tract. IBD has been shown to have physical and psychosocial implications which can adversely effect the health related quality of life (HRQoL) of the sufferer (Drossman, 1994). The aetiology of both CD and UC is *Corresponding author. Tel.:+44-131-537-1756; fax: +44-131-537-1007. E-mail address: [email protected] (G.D. Smith).
largely unknown although genetic, environmental, lifestyle and dietary factors have all been implicated. Disease specific questionnaires are viewed as the most effective type of HRQoL instrument to define fluctuations in health status (Guyatt, 1989). One such tool, the inflammatory bowel disease questionnaire (IBDQ) has been developed and validated and shown to be of value in the assessment of disease outcomes in patients with IBD. The purpose of this paper is to report on the preliminary development of a disease specific scale for the measurement of HRQoL in IBD. The relevance of this study comes from the fact that nurses working in the speciality of gastroenterology are in an ideal position, if they possess the necessary knowledge and skills, to deal
0020-7489/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 7 4 8 9 ( 0 1 ) 0 0 0 4 2 - 6
584
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
with both the physical and psychosocial aspects of chronic illness, such as IBD (Smith, 1997). The development of the scale reported here has been nurseled and it is necessary for nurses to familiarise themselves with the issues surrounding measurement of human attributes and attitudes if they are to implement and gauge the effectiveness of treatment for both the physical and psychosocial aspects of chronic illness. Quality of life is a vague term which is difficult to define. However, it has been described by Shin and Johnson (1978, p. 475) in the following terms: the possession of the resources necessary to the satisfaction of individual needs, wants and desires, participation in activities enabling personal development and self-actualisation and satisfactory comparison between oneself and others. The above provides a very broad definition and one which encapsulates more factors than are probably useful in measuring the impact of a condition such as IBD on the quality of life of the individual. Indeed, it has been demonstrated that only part of quality of life is health related (McGee et al., 1991) and the importance of health aspects increase in those individuals with chronic disease, this has led to the emergence of the concept of HRQoL (Berg, 1975). Following acceptance and acknowledgement by Index Medicus in 1977 there has been an increase in interest of measurement of HRQoL within the medical profession (Bech, 1992). The relevance of HRQoL measurement is that it may provide information on patient population needs, health care delivery requirements, the nature of chronic disease, therapeutic efficacy in clinical trials and as an essential constituent of cost-utility analyses (Fletcher et al., 1992). All quality of life and HRQoL measurement scales require rigorous evaluation prior to their acceptance as valid diagnostic or evaluative tools in disease assessment. Psychometric assessment is the evaluation of the quality of a scale in terms of its reliability, validity, sensitivity and specificity (Bryman and Cramer, 1994; Polit and Hungler, 1995). 1.1. Generic and/or disease specific scales in IBD? Controversy exists over the use of either generic or disease specific questionnaires for the measurement of HRQoL in IBD (Garrett et al., 1990). Generic scales have been seen to have a place in the assessment of chronic disease but they are not always directly applicable and have limited insight into specific conditions (Drummond, 1987). Generic scales do however offer the possibility of making comparisons across disease categories and they are also more likely to fulfil psychometric criteria of reliability and validity.
Despite this, the role of generic scales has received criticism mainly on the grounds of specificity (McKenna, 1993). By their very nature generic instruments will contain irrelevant questions and while the scale may be modified by omitting or ignoring the results of irrelevant questions, generic scales will always lack the specificity of a disease specific scale. Disease specific scales, therefore, have the advantage of specificity because they will have been developed specifically for the condition under study. The value of disease specific questionnaires used in conjunction with generic measures has been proposed by Guyatt et al. (1989a, b).
1.2. Disease specific scales for measuring HRQoL in IBD Disease specific measures used in IBD, such as the Crohn’s disease activity index (CDAI) measure the physical impact of IBD rather than measures of HRQoL (Garrett and Drossman, 1990). The CDAI places emphasis upon the objective markers of disease which may not always be the best indicator of patient well-being. This is not to say that the CDAI is clinically unimportant, disease indices provide the basis for many clinicians’ assessment and management of IBD (Best et al., 1976). It has, however, been recognised in recent years that more attention should be paid to more subjective assessment of psychosocial factors in chronic illness. Indeed there is no clear relationship between laboratory data, such as haematocrit, ESR and overall functional ability. Drossman et al. (1990) highlighted the importance of combining physical disease assessment with psychosocial factors because they define different aspects of medical care. Interest in disease specific HRQoL scales developed as a result of the theory that sufferers’ psychological status influences their response to chronic conditions (Talal and Drossman, 1995) and that sufferers’ psychological and social disability in IBD may be at least as important as physical disability. In part, this is supported by the observation in IBD that objective markers of disease are not clearly related to functional ability (Drossman, 1994). Kunsebeck et al. (1990) supported this theory and concluded that, while objective measures of socio-economic status and disease activity are easy to measure, they lack insight into the personal impact of IBD. Early studies of HRQoL in IBD were methodologically flawed in that they were largely observational, uncontrolled and did not use reliable and validated measures (Mitchell et al., 1988). The outcome of these studies was an indication of the need to quantify psychosocial aspects of HRQoL. Since HRQoL depends upon the sufferer’s perspective, which is subjective and multidimensional, including such aspects as well-being, daily functioning, psychological state, job satisfaction and interpersonal relationships, the necessity for vali-
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
dated questionnaires is indicated. Several scales have therefore been developed to measure HRQoL in IBD. Initial attempts to measure HRQoL in IBD used crude assessment scales asking patients to rate their HRQoL as being ‘good’, ‘fair’ or ‘poor’ (Farmer et al., 1985). Early studies focused on dysfunction following surgery for IBD, these were cross-sectional studies, taking only a ‘snapshot’ of the condition and making no attempt to study the impact of the disease over time using follow-up studies (Irvine, 1995). Drossman et al. (1991a, b) evaluated the health status of a national sample of patients with IBD who were members of the Crohn’s and Colitis Foundation of America. These sufferers were a self-selected group with a range of disease severity who were sufficiently concerned about their condition to join this organisation. CD sufferers had poorer health status and more occupational disability than UC sufferers. This finding was supported by Farmer et al. (1992) who, using the Cleveland Clinic Questionnaire, used a direct-interview technique with IBD sufferers and found HRQoL to be poorer in CD than in those with UC. An attempt has been made to differentiate between the concerns of CD and UC sufferers and this has been carried out using the Rating Form of IBD Patient Concerns (RFIPC) which is a specialised questionnaire identifying the 25 most important worries and concerns of IBD patients (Drossman et al. (1991a, b). The RFIPC contains four dimensions including disease-related, body related, inter/intrapersonal and sex related which are measured on visual analogue scales. The RFIPC was applied by Drossman et al. (1989) and the results indicated that the major problems for IBD sufferers were related to surgery, energy/fatigue and then body image and psychosexual consequences of having a stoma. They concluded that CD and UC patients had similar concerns except that UC patients were more concerned with loss of bowel control and the development of cancer whilst CD patients were most concerned with pain and fertility issues. The results from the RFIPC correlated well with other measures of health status including the clinicians’ global assessment, functional impairment and number of hospitalisations (Drossman, 1992). However, the RFPIC demonstrated flaws related to the content validity of the questionnaire and in relation to lack of reliability. Love et al. (1992) developed the UC/CD Disease Health Status Scale which was designed to discriminate mild from moderate illness and to measure health care use, daily function and psychological distress. Results indicated that quality of life in UC and CD patients was worse than in a group of matched controls. The IBDQ was developed because it was felt that previous questionnaires had not adequately addressed the psychosocial aspects of the disease.
585
The IBDQ examines bowel disturbance, systemic effects, emotional and social function. The psychometric properties of the IBDQ appear to be good in terms of reliability and validity and it correlates well with objective measurement of disease activity. However with 32 items its length has placed practical limitations upon its use within out-patient departments where the vast majority of IBD patients are reviewed. Therefore development of a simpler, shorter instrument, which is less cumbersome to administer may provide benefits in the assessment and management of IBD patients. A shorter ten item IBDQ called the shortened inflammatory bowel disease questionnaire (SIBDQ) was formulated and found to correlate well with both IBDQ and CDAI (Irvine et al., 1996). Further to this Turnbull et al. (1996) have described an even shorter questionnaire containing five items, claiming the instrument which has been designed for use in the out-patient environment to be as reliable and as sensitive as the original 32 item IBDQ. It should also be noted that all the work reported above is North American in origin and a perspective on IBD from outside that context would be useful. 1.3. Development of the EIBDQ Despite the above developments it is reported that an ideal disease specific scale for measurement of HRQoL in IBD in the United Kingdom remains to developed. Such developments would permit more specific assessment of IBD related aspects of HRQoL and also comparison with other chronic diseases whilst retaining sensitivity to treatment effects (Irvine, 1995). Schipper et al. (1990) provide a concise yet adequate HRQoL as ‘‘the functional impact of an illness, and its consequent therapy, upon the patient, as perceived by the patient’’. The EIBDQ was developed, specifically, to encapsulate the above. The EIBDQ contains fifteen questions based upon agreement–disagreement scales (Atkinson, 1991), no copyright is reserved on this scale. The factor structure of the EIBDQ was examined and construct validity was examined using established and validated measures of disease activity and psychological morbidity in IBD. A comprehensive review of the IBD literature and a pilot study of a preliminary version of the EIBDQ resulted in the formulation of the finalised questionnaire which is shown in Table 1. 1.4. Present study The present study was designed to develop and undertake preliminary psychometric testing of a disease specific scale to measure HRQoL in IBD. Turnbull et al. (1996) have indicated the need to develop more user friendly scales to measure facets of HRQoL in IBD and a review of other relevant measures of disease activity
586
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
Table 1 Edinburgh (EIBDQ) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Inflammatory
Bowel
Disease
Questionnaire
Do you feel that sufficient information was given to you about your disease? Are you familiar with the National Association of Colitis and Crohn’s disease (NACC)? Over the past two weeks have you had diarrhoea? Over the past two weeks have you had tummy pain? Over the past two weeks have you been overly tired? Over the past two weeks have you had any joint pain? Do you feel that you have full control over your bowel function? Do you ever have to rush to the toilet in regarding your bowels? Do you ever have accidents regarding your bowels? Does your bowel function affect your relationship at home? Does your bowel function affect your social life? Does your bowel function affect any recreational activities you normally do? Do you feel your recreational activities are restricted by your disease? Do you feel your disease has affected your relationships at home? Do you feel your overall ‘quality of life’ is affected by your disease?
Responses to Items 1, 2, 7, 8, 9, 10, 11 and 12=‘‘Yes’’ and ‘‘No’’. Responses to Items 3, 6, 13, 14, 15 and 16=‘‘Not at all’’, ‘‘A little’’, ‘‘Moderately’’ or ‘‘Severely’’.
and psychological morbidity have been presented. The design of the present study was by means of a survey which utilised the EIBDQ with a sample of IBD sufferers and psoriatic arthritics who were used to represent another group of sufferers of a chronic inflammatory condition.
2. Methods One hundred and twenty-eight subjects, 50 CD (age range 23–82; 38 female), 50 UC (age range 19–78; 37 female) and 28 psoriatic arthritics (age range 17–62; 27 female) were recruited into the study. The CD and UC subjects were recruited from an out-patient gastrointestinal clinic and psoriatic arthritics were recruited from a rheumatology clinic within the same hospital. The following questionnaires were administered to all subjects: a modified CDAI, HAD, SF-36, SS (styles and strategies questionnaire), and the EIBDQ.
medication and also objective clinical measures such as the haematocrit. The modified version used only those aspects which the patients could self-report. The HAD has 14 questions and two scales, one for anxiety and one for depression, it is simple to use and relatively insensitive to physical symptoms. The SF-36 comprises 36 questions measuring eight dimensions of HRQoL including physical function, social function, mental health, role limitation due to physical problems, energy/vitality, role limitations due to emotional problems, pain and health perception. The scale was developed in America but a modified United Kingdom version exists (Brazier et al., 1972). The psychometric properties of the SF-36 were established by Garratt et al. (1993) showing good construct validity and factor analysis supports the underlying dimensions. The SS questionnaire, developed by Roger et al. (1993) is a 16 item questionnaire which measures adaptive and maladaptive styles of coping in terms of being detached and rational, on the one hand, and emotional and avoiding, on the other. The EIBDQ has been described above. Data from the above questionnaires were entered into a database for analysis using SPSS for Windows. The following statistical tests were applied: Pearson’s Product Moment Correlation Coefficient and Cronbach’s alpha to measure reliability using internal consistency of multiple item scales or factors. Principal components analysis with oblique rotation was applied as described by Watson and Deary (1994) to the responses in the EIBDQ, with further manoeuvres to eliminate redundant questions and reduce the number of factors while accounting for as much of the variance as possible as described by Agius et al. (1996). Oblique rotation was selected in preference to Varimax rotation as this obtained the simplest solution with highest loadings on putative factors. Using this application Item 6 of the EIBDQ, which relates solely to joint pain, was omitted and Item 15 was omitted because it may have been misinterpreted due to ambiguity surrounding ‘quality of life’. The omission of both these item from the factor analysis improved the structure of the derived matrix (Table 3). The disease specific nature of selected items from the EIBDQ which were specific to either IBD or psoriatic arthritis (PA) was tested using the chi square test of proportion. The chi-square test of proportions was used to analyse the association between symptoms of IBD in IBD and PA patients.
3. Results 3.1. Disease specificity
2.1. Instruments The CDAI is an eight item instrument including questions on pain, stool consistency and use of
Table 2 shows selected items from the EIBDQ related to symptomatic reporting of diarrhoea, abdominal pain, tiredness and joint pain. Diarrhoea, abdominal pain and
587
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590 Table 2 Response to EIBDQ items specific and non-specific to IBD in IBD and PA patients
Item 3 (diarrhoea) IBD PA Item 4 (abdominal pain) IBD PA Item 5 (tiredness) IBD PA Item 6 (joint pain) IBD PA
Absent
Present
Chisquared
d.f.
31 26
79 2
27
2
o0.05
29 26
71 2
36
2
o0.05
8 13
92 15
66
2
o0.05
39 1
61 27
12
2
o0.05
p
taken to indicate reliability. The intercorrelation between the three rotated factors is very low (Table 3) indicating that, despite the use of oblique rotations, the factors were representing different underlying dimensions of the EIBDQ. 3.4. Validity
d.f.=degrees of freedom. CD and UC sufferers were collapsed into one category of IBD sufferers and responses to the IBDQ were dichotomised into ‘‘absent’’ and ‘‘present’’, the latter category being created by collapsing the ‘‘mild’’, ‘‘moderate’’ or ‘‘severe’’ responses to items.
tiredness which are all recognised as symptoms of IBD, are more prevalent in both CD and UC patients than in those suffering from psoriatic arthritis. Conversely, as shown in Table 2, joint pain is much more prevalent in sufferers of psoriatic arthritis than either group of IBD sufferers. The association between disease state and item responses is statistically significant (po0:05) using the chi square test of proportions.
In order to demonstrate convergent validity, the putative factors were correlated with validated measures of physical and psychological morbidity in IBD. Factor scores were calculated by summing the scores for all items loading on the factors. The correlation matrixes shown in Tables 4 and 5 show the intercorrelation of the bowel specific factor (Factor 2) with the CDAI, the two dimensions of the HAD and one dimension from each of the SS questionnaire (emotional coping) and SF-36 (physical function) for CD and UC respectively. Factor scores for the bowel specific factor were generated by summing the scores for all items loading on Factor 2. Significant correlations (po0:05) above 0.3 are shown in bold as these represent at least a moderate correlation between variables. For both CD and UC Factor 2 correlated with the CDAI score. Factor 2 also correlated with the HAD score for anxiety in the case of CD. The SF-36 dimension related to physical role function correlated with Factor 2 for both CD and UC and the Table 3 Principal components analysis of the EIBDS data followed by oblique rotation Factor 1 (Disease specific)
Facto 2 (Bowel specific)
Factor 3 (Information)
14 13 12 4 5
0.811 0.775 0.702 0.640 0.581
0.030 0.035 0.035 0.113 0.001
0.006 0.006 0.078 0.098 0.032
10 8 9 7 11 3
0.119 0.050 0.023 0.072 0.153 0.362
0.860 0.846 0.833 0.800 0.778 0.408
0.130 0.182 0.032 0.173 0.167 0.114
0.029 0.051 0.79
0.831 0.787 0.55
3.3. Reliability
2 0.193 1 0.205 Cronbach’s a 0.86 Correlation 1 1 2 0.24 3 0.15
Values of Cronbach’s alpha for the three rotated factors are also shown in Table 3, and were acceptable for Factors 1 and 2, a Cronbach’s alpha of 0.8 being
Factor loading greater than 0.400 are shown in bold with the internal consistency of the rotated factors (Cronbach’s a) and the correlation between factors being presented at the foot of the table.
3.2. Factor analysis The results of principal components analysis followed by oblique rotation showed three putative factors, and these factors are, respectively, a non-bowel disease specific factor (Factor 1), a bowel specific factor (Factor 2) and an information related factor (Factor 3). The loadings of variables which are greater than 0.4 are shown in bold in Table 3. The percentage of variance extracted by each factor, before rotation was 39%, 14% and 10%, respectively.
Item
1 0.04
1
588
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
Table 4 Correlation matrix of EIBDS Factor 2 with CDAI, HAD, SS and selected items for SF-36 for Crohn’s disease patients (n=50)
EIBDS Factor 2 Crohn’s activity index HAD (anxiety) HAD (depression) SS (emotional) SF-36 (physical function)
EIBDS Factor 2
Crohn’s activity index
HAD (anxiety)
HAD (depression)
SS (emotional)
SF-36 (physical function)
1 0.52 0.16 0.55 0.23 0.48
1 0.24 0.37 0.04 0.35
1 0.45 0.36 0.06
1 0.07 0.23
1 0.04
1
Correlations greater than 0.3 and po0.05 are shown in bold. See text for explanation of abbreviations used in the table.
Table 5 Correlation matrix of EIBDS Factor 2 with CDAI, HAD, and selected items for SF-36 for ulcerative colitis patients (n=50)
EIBDS Factor 2 Crohn’s activity index HAD (anxiety) HAD (depression) SS (emotional) SF-36 (physical function)
EIBDS Factor 2
Crohn’s activity index
HAD (anxiety)
HAD (depression)
SS (emotional)
SF-36 (physical function)
1 0.30 0.40 0.43 0.30 0.32
1 0.42 0.13 0.45 0.16
1 0.45 0.65 0.06
1 0.25 0.35
1 0.01
1
Correlations greater than 0.3 and po0.05 are shown in bold. See text for explanation of abbreviations used in the table.
SS questionnaire item related to emotional coping with stress correlated with Factor 2 in UC sufferers only.
4. Discussion The purpose of the work described in this paper was the development of a short, user-friendly disease specific scale to measure HRQoL in IBD. Although the IBDQ is widely accepted as the ‘gold standard’ in the measurement of HRQoL in IBD (Irvine, 1993), it is commonly accepted that there is the need for a shorter more practical questionnaire which can be easily administered in the out-patient environment. Whilst acknowledging that generic scales have a place in the measurement of HRQoL in IBD in terms of enabling comparisons to be made across diseases, there is an established requirement for scales which will specifically measure and reflect the impact of HRQoL in patients with IBD. The study was carried out by reviewing the literature for dimensions of HRQoL which appear to be affected by IBD and incorporating these with other aspects of IBD into a new instrument called the EIBDQ. A range of multivariate statistical techniques was employed to establish the specificity, reliability and convergent validity of the EIBDQ.
This study has identified a disease specific dimension of the EIBDQ, as demonstrated by factor analysis, from other aspects of IBD such as bowel activity and the provision of information to IBD patients. It was possible to discriminate between IBD sufferers and sufferers of another inflammatory condition, psoriatic arthritis, by means of individual disease specific items in the EIBDQ and the HRQoL related items of the EIBDQ were more prevalent in IBD patients than in patients with psoriatic arthritis. The reliability of the bowel specific dimension of the EIBDQ (Factor 2) is reliable as demonstrated by Cronbach’s alpha. Factor 2 correlates well with a reliable and valid measure of depression and for UC patients and it also correlates well with a reliable and valid measure of anxiety and depression. In both cases the HAD was used to measure anxiety and depression (Table 4). Likewise, an important dimension of the SF36 correlates with the bowel activity specific as does the CDAI which is an established objective measure of physical disease activity in IBD (Table 5). The correlations suggest that the bowel activity specific scale (Factor 1) has construct validity in terms of other measures of psychosocial, HRQoL and disease activity measures. Correlations greater than 0.3 are taken to represent at least moderate correlations between variables (Bryman and Cramer, 1996).
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
The work reported here is preliminary in the sense that any scale can be improved and other aspects of the questionnaire from which the bowel activity specific scale is derived may also be improved to provide a comprehensive measure of the full impact of IBD upon HRQoL. Further development could proceed by including new items, particularly in the dimension of information provision, but also to the domain of the EIBDS from which the bowel specific factor is derived. The importance of developing such disease specific scales extends beyond that of mere research and surveying the impact of disease upon HRQoL. If such scales are sufficiently sensitive, and further research and development could be aimed at investigating this aspect, then it could be used in further research into the efficacy of treatment in IBD. This could include treatment aimed at alleviating the underlying condition and psychosocial interventions which are aimed at the manifestations of IBD, especially upon HRQoL, such as counselling.
References Agius, R., Blenkin, H., Deary, I., Zealley, H., Wood, R., 1996. Survey of perceived stress and work demands of consultant doctors. Occupational and Environmental Medicine 53, 217–224. Atkinson, F.I., 1991. Survey Design and Sampling. In: Cormack, D.F.S. (Ed.), The Research Process in Nursing, 2nd Edition.. Blackwell Science Publication, Oxford, pp. 196–206 (Chapter 21). Bech, P., 1992. Measuring quality of life: the medical perspective. Nordic Journal of Psychiatry 46, 85–89. Berg, O., 1975. Health and qulaity of life. Acta Sociologica 18, 3–22. Best, W.R., Becktel, J.M., Singleton, J.W., Kem, F., 1976. Development of a crohn’s disease activity index. Gastroenterology 70, 439–444. Brazier, J.E., Harper, R., Jones, N.M.B., O’Cathain, A., Thomas, K.J., Usherwood, T., 1972. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. British Medical Journal 305, 160–164. Bryman, A., Cramer, D., 1994. Quantitative Data Analysis for Social Scientists, Routledge, London, pp. 152–190. Drossman, D.A., 1994. Psychosocial factors in ulcerative colitis and Crohn’s disease In: Kirnser, J.B. (Ed.), Inflammatory Bowel Disease, Leah & Febriger, Philadelphia, PA. Drossman, D.A., Leserman, L.Z., 1992. Ulcerative colitis and Crohn’s disease health status scales for research and clinical practice. Journal of Clinical Gastroenterology 15, 104–107. Drossman, D.A., Patrick, D.L., Mitchell, C.M., Zagami, E.A., Applebaum, M.I, 1989. Health related quality of life in inflammatory bowel disease: functional status and patient concerns. Digestive Disease and Science 34, 1379–1380. Drossman, D.A., Mitchell, C.M., Leserman, J., Patrick, D.L., Zagami, E.A., 1990. Psychosocial and disease predictors of utilization in IBD. Gastroenterology 98 (Abs). Drossman, D.A., Leserman, J., Li, Z., 1991a. Health status and health care use in persons with inflammatory bowel disease:
589
a national survey. Digestive Disease and Science 36, 1746– 1749. Drossman, D.A., Leserman, J., Li, Z., 1991b. The rating form of IBD patient concerns: a new measure of health status. Psychosomatic Medicine 53, 701–702. Drummond, M.F., 1987. Resource allocations and decisions in health care: a role for quality of life assessments. Journal of Chronic Disease 40, 605–616. Farmer, R., Whelan, G., Fazio, V.W., 1985. Long term followup of patients with crohn’s disease. Relationship between the clinical pattern and prognosis. Gastroenterology 88, 1818–1825. Farmer, R.G., Easley, K.A., Farmer, J.M., 1992. Quality of life assessment with patients with inflammatory bowel disease. Cleveland Clinical Journal of Medicine 59, 35–52. Fletcher, A., Gore, S., Jones, D., Fitzpatrick, R., Spiegelhalter, D., Cox, D., 1992. Quality of life measures in health care. II: design, analysis, and interpretation. British Medical Journal 305, 1145–1148. Garrett, J.W., Drossman, D.A., 1990. Health status in inflammatory bowel disease. Gastroenterology 99, 90–91. Garratt, A.M., MacDonald, L.M., Ruta, D.A., Russell, I.T., Buckingham, J.K., Krukowski, Z.H., 1993. Towards the measurement of outcomes for patients with varicose veins. Quality of life in Health Care 2, 5–10. Guyatt, G., Mitchell, A., Irvine, J., Singer, J., Williams, N., Goodacre, R., Tompkins, C., 1989a. A new measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology; 96, 804–810. Guyatt, G., Mitchell, A., Irvine, E., Singer, J., Williams, N., Goodacre, R., Tomkins, C., 1989b. A measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology 96, 804–810. Irvine, E.J., 1993. A quality of life index for IBD. Canadian Journal of Gastroenterology 7, 155–159. Irvine, E.J., 1995. Quality of life in inflammatory bowel disease: biases and other factors affecting scores. Scandanavian Journal of Gastroenterology 30, 136–140. Irvine, E.J., Zhou, Q., Thomson, M.B.A., 1996. The short inflammatory bowel disease questionnaire: a quality of life instrument for community physicians managing inflammatory bowel disease. American Journal of Gastroenterology 91, 1571–1578. Kunsebeck, H.W., Korber, J., Freyberger, H., 1990. Quality of life in patients with inflammatory bowel disease. Psychotherapy and Psychosomatic Research 29, 23–31. Love, J.R., Irvine, E.J., Fedorak, R.N., 1992. Quality of life in inflammatory bowel disease. Journal of Clinical Gastroenterology 14, 15–19. McGee, H.M., O’Boyle, C.A., Hickey, A., et al., 1991. Assessing the quality of life of the individual: the seiqol with a healthy and gastroenterology unit population. Psychological Medicine 21, 749–758. McKenna, S.P., 1993. The Nottingham Health Profile. Proceedings of the Fifth European Health Service Research Conference, Maastricht, Netherlands, Unpublished paper. Mitchell, A., Guyatt, G., Singer, J., 1988. Quality of life in patients with inflammatory bowel disease. Journal of Clinical Gastroenterology 10, 306–310. Polit, D.F., Hungler, B.P., 1995. Nursing Research: Principles and Methods. Lippincot, Philadelphia, PA.
590
G.D. Smith et al. / International Journal of Nursing Studies 39 (2002) 583–590
Roger, D., Jarvis, G., Najarian, B., 1993. Detachment and coping: the construction and validation of a new scale for measuring coping strategies. Personality and Individual Differences 15, 619–626. Schipper, H., Clinch, J., Powell, V., 1990. Definitions and conceptual issues. In: Spiker, B. (Ed.), Quality of Life Assessment in Clinical Trials. Raven Press, New York, pp. 5–24. Shin, D.C., Johnson, D.M., 1978. Avowed hapiness as an overall assessment of the quality of life. Social Indicators Research 5, 475–492. Smith, G.D., 1997. Counselling in inflammatory bowel disease. Unpublished PhD thesis, University of Edinburgh.
Talal, A.H., Drossman, D.A., 1995. Psychosocial factors in inflammatory bowel disease. Gastroenterology Clinics of North America 24, 173–181. Turnbull, G.K., Vallis, T.M., Luyendyk, Y., 1996. The assessment of quality of life in inflammatory bowel disease patients: a five item scale for out-patient use. Gut 39, 16–17. Watson, R., Deary, I.J., 1994. Measuring feeding difficulties in patients with dementia: multivariate analysis of feeding problems, nursing intervention and indicators of feeding difficulty. Journal of Advanced Nursing 20, 283– 287.