Quality of life in patients with established inflammatory bowel disease: a UK general practice survey

Authors


Professor G. P. Rubin, Centre for Primary and Community Care, University of Sunderland, Benedict Building, St George's Way, Sunderland SR2 7BW, UK.
E-mail: greg.rubin@sunderland.ac.uk

Summary

Background : The current understanding of quality of life impairment in inflammatory bowel disease has largely been derived from selected populations and may not reflect the experience of patients in the community, where fewer than half are likely to be under specialist care.

Aim : To describe the health-related quality of life in a community-based sample of patients with established inflammatory bowel disease and explore its association with the type and extent of disease, gender, age, material deprivation and other factors.

Methods : Adults with established inflammatory bowel disease were identified systematically from the records of 23 family practices in north-east England. The health-related quality of life was assessed by self-completion of the UK Inflammatory Bowel Disease Questionnaire.

Results : Five hundred and fifty-six patients were sent the questionnaire and 409 (74%) gave usable replies. Lower scores (worse quality of life) were significantly associated with female gender, Crohn's disease, more extensive disease (ulcerative colitis) and being under specialist care. The mean health-related quality of life score was significantly lower in patients resident in more deprived districts, independent of the type and extent of disease.

Conclusions : Most patients with established inflammatory bowel disease showed only minor impairment of their health-related quality of life. On average, women and those with Crohn's disease were relatively more affected. Clinicians responsible for the care of patients with inflammatory bowel disease should be aware of these more vulnerable groups.

Introduction

The health-related quality of life is a critical dimension of any chronic disease. It reflects sensory function, including pain and discomfort, activities of daily living and physical, social, emotional and cognitive function.1 Its measurement quantifies the impact of a disease across a wider psychosocial dimension, rather than taking only the disease-centred perspective. Health-related quality of life measures are useful in the assessment of the quality of health care delivery, enhancement of disease-related knowledge and measurement of treatment efficacy or disease outcome.2

Impairment of the health-related quality of life is a recognized feature of inflammatory bowel diseases (ulcerative colitis and Crohn's disease).3 Associations have been described between a poor quality of life and the type and activity of disease,4 as well as features unrelated to the disease, such as co-morbidity, gender,5 psychological disorders6 and coping strategies.7 Many of these findings have been derived from selected populations, most commonly patients under specialist care. In one community sample of patients with inflammatory bowel disease, however, less than one-half were definitely under specialist care.8 The nature of the differences in health-related quality of life between inflammatory bowel disease patient sub-groups and the validity of previously described associations are uncertain for these more representative populations.

The aim of this study was to describe the quality of life in a community-based sample of patients with established inflammatory bowel disease and to explore its association with the type and extent of disease, gender, age, material deprivation and access to health care. We selected patients who had established inflammatory bowel disease, defined as having been diagnosed at least 2 years previously, in order to describe the sustained rather than the initial impact of inflammatory bowel disease on the health-related quality of life.

Methods

The study was conducted in a large urban area in northern England, UK. The population comprised 178 400, of which 5% were unemployed, 2.8% were from ethnic minorities, 14% were retired and almost 20% had a long-term limiting illness (2001 census). Seventy-one per cent of households were owner-occupied and 29% had no car.

Patients with inflammatory bowel disease were identified from a systematic search of the computer diagnostic and prescribing records of 23 of 25 practices from a single Primary Care Trust. The combined practice population was 163 383. Case definition was based on that described by Garland et al.,9 and has been discussed in detail previously.8 Briefly, a specialist diagnosis supported by confirmatory evidence from histological, endoscopic, radiological or operative findings was essential for inclusion. Patients were considered to be under specialist care if they had attended an out-patient clinic in the previous 12 months, and were considered to be discharged if the most recent letter received from the responsible specialist stated this. Postcodes for registered home addresses were mapped to electoral ward in order to derive the Townsend score for material deprivation (based on the 1991 census).

The quality of life was assessed using the UK version of the Inflammatory Bowel Disease Questionnaire (IBDQ),10 originally developed by Guyatt et al.11 and validated for self-administration.12 The UK instrument is based on 30 items, each scored in the range 1–4, with a summary score between 30 and 120. A low score indicates a poor quality of life. Separate scores were derived for the five sub-domains: bowel I (bowel movements and use of facilities), bowel II (general bowel symptoms), emotional, social and systemic function (mainly fatigue).

The questionnaire was posted with an explanatory letter to patients 16 years and older who had a disease duration of at least 2 years and had not undergone a total colectomy. One reminder was sent to non-responders.

Missing questions were dealt with as follows. The entire questionnaire was ignored when more than half the items of one or more sub-domains were missing. Otherwise, substitution of the mean of the remaining items in the sub-domain was used for missing responses (W. Y. Cheung, 2002).

Scores were analysed as absolute values and as a percentage for comparison across domains. The percentage score for each domain was calculated using (observed score − minimum possible score) × 100/(maximum possible score − minimum possible score). Data were recorded using the EPI-Info database13 and analysed using the Statistical Package for Social Sciences.14 Comparison of proportions was made using the chi-squared test, of means using the t-test and of medians using the Mann–Whitney test and Kruskal–Wallis test. Forward stepwise multivariate regression analysis was used to explore the effects of various factors on the total quality of life score. The 5% level indicated statistical significance. Ethics Committee approval was obtained.

Results

In total, 717 patients fulfilled the inclusion criteria, 425 of whom had ulcerative colitis, 275 had Crohn's disease and 17 had non-specific inflammatory bowel disease. The crude prevalence per 100 000 for ulcerative colitis was 260 (95% confidence interval, 235–285) and for Crohn's disease was 168 (95% confidence interval, 148–188). Postcodes, and hence ward-based Townsend scores, were available for 697 (97%) patients.

Quality of life

Of the 717 patients, 712 were over 15 years of age. Of these, 75 had disease of less than 2 years' duration and a further 81 had undergone total colectomy; these 156 patients were excluded. Questionnaires were mailed to 556 patients (310 women and 246 men). Returned questionnaires were received from 433 patients, 24 of which were unusable due to multiple missing responses. In 371 cases, the questionnaires were fully completed and, in 38 cases, between one and three items had been missed (one question missed in 27 cases, two items missed in nine cases and three items missed in two cases). Hence, usable data were available for 409 (73.6%) patients. The mean age was 53 years (s.d., 16 years) and the mean age at diagnosis was 40 years (s.d., 15.5 years). The mean duration of disease was 13.3 years (s.d., 9.4 years). There were 239 patients with ulcerative colitis [134 (56%) female], 163 with Crohn's disease [98 (60%) female] and seven with indeterminate inflammatory bowel disease [four (57%) female].

The mean age, age at diagnosis and duration of disease did not differ significantly between responders and non-responders (P > 0.3 in all comparisons). The response rates were 76.1% in women and 70.3% in men (P = 0.12), 74.5% in patients with ulcerative colitis, 73.8% in patients with Crohn's disease and 50% in patients with ‘inflammatory bowel disease’ (P = 0.13). The response did not vary with the extent (ulcerative colitis) or site (Crohn's disease) of disease. Two hundred and ten patients were definitely under specialist care, with 346 being under the care of their general practitioner (discharged from specialist care or uncertain). The response rates were 69.9% in those under general practitioner care and 79.5% in those under specialist care (P = 0.013). The response rate did not vary significantly by Townsend score when grouped into quartiles (P = 0.12), but was lower amongst patients resident in the most deprived wards (68%) than in those resident in the most affluent wards (80%).

The overall distributions of total quality of life score and most sub-domain scores were skewed (Figure 1, Table 1). The median IBDQ score for all respondents was 98 (mean, 94.1; interquartile range, 83–108). The total IBDQ score was expressed as a cube function to normalize its distribution and regressed on age as a covariate and dummy variables for gender, diagnosis (ulcerative colitis vs. Crohn's disease) and hospital care (yes/no). The quality of life was worse in younger patients, in women compared with men, in Crohn's disease compared with ulcerative colitis patients and in those under hospital care compared with those under general practitioner care (Table 2).

Figure 1.

Quality of life (QoL): distribution of scores in 409 patients with inflammatory bowel disease.

Table 1.  Inflammatory Bowel Disease Questionnaire (IBDQ) scores: summary statistics in 409 patients with inflammatory bowel disease
Sub-domainMinimumMaximumMeans.d.MedianIQR
  1. IQR, interquartile range; s.d., standard deviation.

Total score3512094.117.09883–108
Emotional93628.54.82925–32
Social72825.04.22723–28
Bowel I62418.44.81914–23
Bowel II41612.03.01310–14
Systemic41610.23.3108–13
Table 2.  Effects of age, gender, disease and responsibility for care on back-transformed total Inflammatory Bowel Disease Questionnaire (IBDQ) score
 Regression coefficient95% CIP value
Age0.10.01, 0.170.038
Gender− 3.7, if female− 7.3, − 0.60.016
Diagnosis− 4.0, if Crohn's disease− 7.8, − 0.90.009
Specialist care− 7.1, if under specialist care− 11.5, − 3.5< 0.001

Respondents were also categorized by disease and sex and, for each group, the top and bottom quartiles were established based on the total quality of life score. In those with the worst quality of life, the impact of the type of disease on the sub-domains varied, with relatively greater impact on systemic symptoms than on bowel symptoms or social functioning (Table 3).

Table 3.  Quality of life sub-domain scores (percentages) for patients in the top quartile and bottom quartile of the distribution for the total score
SexDiseaseQuartilenMedian scores (%)
Bowel IBowel IIEmotionalSocialSystemic
FemaleUlcerative colitisTop35100928910092
Bottom324225546225
FemaleCrohn's diseaseTop2497878510075
Bottom222542525217
MaleUlcerative colitisTop2794929310092
Bottom264242596921
MaleCrohn's diseaseTop1694929310083
Bottom15334252578

Of the patients in the top quartile of the total score, only 25% were under hospital care, compared with 55% of those in the bottom quartile. Hence, 45% of those with the poorest quality of life were not under hospital care.

In patients with ulcerative colitis, women had a significantly worse quality of life than men in the emotional domain (median scores, 29 vs. 31; P = 0.050) and bowel II domain (median scores, 12 vs. 13; P = 0.002). The quality of life scores were related to the extent of colitis, but the differences were not statistically significant (Table 4).

Table 4.  Median values for total quality of life (QoL) score and sub-domain scores in patients with inflammatory bowel disease
 nMedian scores (absolute scores)
Total QoLBowel IBowel IIEmotionalSocialSystemic
  • 54 patients with Crohn's disease at multiple sites excluded.

Extent of colitis
 Proctitis11899.521.013.029.528.011.0
 Left-sided78102.521.013.031.027.512.0
 Extensive2996.020.012.029.028.09.0
 Unclear1491.020.011.026.026.59.0
 P (Kruskal–Wallis test) 0.240.470.0800.120.710.28
Site of Crohn's disease*
 Large bowel only3398.018.013.029.028.010.0
 Small bowel only1992.018.013.027.025.010.0
 Ileal only2987.016.012.027.025.09.0
 Rectal only18100.520.011.528.526.510.0

In patients with Crohn's disease, the quality of life did not differ significantly with site. Women with Crohn's disease had a significantly worse quality of life than men with Crohn's disease in the bowel II domain (median scores, 11 vs. 13; P = 0.003). Women with Crohn's disease had a significantly worse quality of life than women with ulcerative colitis in the bowel I and social domains (Table 5). In men, there were smaller differences in the median scores between diseases.

Table 5.  Median values for total quality of life (QoL) score and sub-domain scores in patients with ulcerative colitis and Crohn's disease: effect of gender
Sex, diseasenMedian scores (absolute scores)
Total QoLBowel IBowel IIEmotionalSocialSystemic
Male, ulcerative colitis105102.021.013.031.028.011.0
Male, Crohn's disease6597.018.013.029.026.010.0
P 0.1010.0660.430.120.130.39
Female, ulcerative colitis13498.520.012.029.027.011.0
Female, Crohn's disease9888.516.011.028.026.010.0
P 0.003< 0.0010.270.200.0010.055

Deprivation and health-related quality of life

After allowance for age, gender, disease and whether under specialist care, patients resident in the most deprived wards (bottom quartile of the Townsend score) had a poorer quality of life than those resident in more affluent wards (mean difference, − 9; P < 0.001). This held true separately for ulcerative colitis and Crohn's disease and was independent of the extent of disease for ulcerative colitis patients. In patients with the poorest quality of life (bottom quartile; total score, < 83), the proportion under specialist care living in the 25% most deprived wards was not significantly different from that living in the remaining 75% of wards (50% vs. 58%, P = 0.39).

Discussion

In this large community sample of patients with established inflammatory bowel disease, impairment of the health-related quality of life was greatest in women, in those with Crohn's disease (compared with ulcerative colitis) and in those under specialist care, with material deprivation having an independent effect. In those with the worst total scores, all domains were affected, but the largest impact appeared to be on the systemic domain (fatigue, lack of energy, feeling generally unwell and sleep disturbance).

Disease may impact on different domains to varying extents, but result in the same total IBDQ score. As a general view, the impact on the quality of life from a total score around the median could arise from a patient experiencing bowel and systemic symptoms on as little as 1 or 2 days over the previous 2 weeks, together with a disruption of social activities and emotional function some or most of this time.

To our knowledge, this is the first large-scale, cross-sectional, community survey of the quality of life in inflammatory bowel disease patients. We are confident that all subjects suitable for inclusion were identified and a high response rate (74%) was achieved. Nevertheless, significantly more patients under specialist care, in whom the mean score was lower, returned questionnaires (79.5% vs. 69.9%), and this may have resulted in bias towards an artificially low score for the entire group (poorer quality of life).

The mean total and domain scores for the whole sample were similar to those described for hospital out-patients with disease in remission.10 We confirmed previous findings of a greater impairment of the health-related quality of life in Crohn's disease compared with ulcerative colitis,4 and in women compared with men.5

It was intuitively correct that patients with the worst quality of life should be under specialist care, although we found that, for 45% of these patients, this was not the case. These patients might benefit from specialist referral, but we were unable to determine whether they were non-consulters or had consulted and were being managed appropriately in primary care.

We sought to examine the impact of material deprivation on the health-related quality of life. We used the ward-based Townsend score from the 1991 census in preference to the more recent Index of Multiple Deprivation, as the latter is less reflective of material deprivation. The health-related quality of life was related independently to ward-based material deprivation, but this may not translate directly to the impact of personal deprivation at an individual level. Further work is needed to evaluate deprivation as an independent determinant of quality of life in inflammatory bowel disease patients. The inverse care law states that the availability of good medical care varies inversely with the need of the population served, and has been shown to operate for a range of clinical conditions.15 Our study does not provide evidence that it applies in this UK population of patients with inflammatory bowel disease, the majority of whom were not under specialist care.

Implications of findings

Quality of life measures can identify patients' problems, facilitate communication and shared decision making and monitor responses to treatment.16 Validated instruments for measuring the health-related quality of life in inflammatory bowel disease include the IBDQ, the Rating Form of Inflammatory Bowel Disease Patient Concerns, the Ulcerative Colitis and Crohn's Disease Health Status Scales and others, although the IBDQ is the most widely used for research purposes. We used the newly published anglicized version of the IBDQ to measure the health-related quality of life. This instrument reflects primarily disease activity, whereas psychological and social aspects of inflammatory bowel disease may have been better reflected by the Rating Form of Inflammatory Bowel Disease Patient Concerns.17 Quality of life measures have not been widely integrated into routine clinical practice, although they are potentially useful in both the clinical encounter and in quality improvement.16 However, the UK version of the IBDQ, based on 30 items, is probably too long for routine use and shortened health-related quality of life instruments have been developed in an attempt to remedy this; the 12-question Short-IBDQ is an example.18

In summary, our findings suggest that, although most patients with established inflammatory bowel disease experience only minor impairment of their quality of life, those with Crohn's disease and women with either type of inflammatory bowel disease are more likely to show greater impairment. Furthermore, a minority of patients with the poorest quality of life are not under hospital care. Clinicians responsible for the care of patients with inflammatory bowel disease should be aware of these more vulnerable groups. One clinical application of a quality of life measure may be to identify those most severely affected by the disease. Our data provide a cut-off point for the worst affected quartile derived from a sample unaffected by selection bias for out-patient attendance.

Acknowledgements

We thank Ms Jane Bates and Mr Leon Green for data collection and Ms Clare Eynon for mapping postcodes to electoral wards.

This work was funded, in part, by Shire Pharmaceuticals. D. J. C. is funded by the Northern Primary Care Research Network (NoReN).

Ancillary