Postal questionnaire study of disability in the community associated with psoriasisBMJ 1996; 313 doi: https://doi.org/10.1136/bmj.313.7062.919 (Published 12 October 1996) Cite this as: BMJ 1996;313:919
- a Norton Medical Centre, Norton, Stockton on Tees TS20 1AN,
- b Centre for Health and Medical Research, University of Teeside, Middlesborough TS1 3BA
- Correspondence to: Dr O'Neill.
- Accepted 7 August 1996
Objective: To study the disability caused by psoriasis in patients recorded as having psoriasis by their general practitioner.
Design: Postal questionnaire survey using the psoriasis disability index and SF-36.
Setting: Five general practices in Cleveland.
Subjects: Of 767 patients identified, 546 completed the questionnaire and 435 were eligible and gave informed consent.
Main outcome measures: Scores on SF-36 and psoriasis disability index.
Results: The psoriasis disability index score was highly negatively correlated with all eight of the SF-36 health measures (P<0.0001 for each), and the manual social classes scored higher than the non-manual social classes (P<0.0001). The manual social class group scored significantly lower scores than the controls on all the SF-36 scales, and the non-manual group scored significantly lower for physical and mental role limitation (P<0.0004 and P = 0.026), mental health (P<0.0001), energy and vitality (P<0.0004), and health perception (P<0.0001). Also, the manual group had poorer health perception on five of the SF-36 variables when compared with the non-manual group.
Conclusions: Patients with psoriasis have an overall lower perception of their quality of life than healthy controls, and those in the lower social classes suffer a greater degree of disability from their disease than the higher social classes.
In this study people with psoriasis had a greater disability than healthy controls
These patients did not consult about their skin condition frequently, yet most used some treatment
The manual social classes suffer greater disability than non-manual classes
The psoriasis disability index may help measure disability in general practice
A community study in 1976 showed that psoriasis affects 1–2% of the British population,1 but only 3% of people with psoriasis see a dermatologist; 17% are treated by their general practitioner alone, and the other 80% do not consult the medical profession at all.2 Several researchers have shown that doctors probably underestimate the degree of psychological and social morbidity.3 4
A recent study of 369 people with severe psoriasis referred to consultant dermatologists throughout the United Kingdom showed that the disease had greatly affected their quality of life.5 There is evidence that psoriasis is associated with smoking and high alcohol consumption,6 although it is not clear whether these are causal links7 or a reflection of the despair felt by those with severe psoriasis.8 9
These results suggest that the perceived disability of those studied may be greater than that of the rest of the psoriatic population.3 5 People with psoriasis who do not attend consultants may “suffer in silence” or simply not experience serious disability. We therefore set up this study to assess health seeking behaviour of people with psoriasis, to measure knowledge and membership of the Psoriasis Association, to compare the general health of people with psoriasis with that of the general population, and to evaluate the psoriasis disability index as a measure of disability in a community based population.
Subjects and methods
The study took place in five general practices in Cleveland. None of the practices had any special provision for patients with psoriasis (for example, a clinic run by general practitioners or nurses). With the consent of the participating practices we compiled a list of all registered psoriatic patients aged 18–64 years old from the practice morbidity index and patients on repeat prescription lists for preparations used only for psoriasis (for example, dithranol, calcipotriol). Patients in whom the diagnosis was not clear were excluded.
We sent the patients a postal questionnaire. All patients and their general practitioners gave informed consent to the study. The first section started by asking the patients to answer the question “Do you think you have psoriasis?” Patients who answered “no” were excluded from the study. Those who answered “yes” were asked for details about their consulting patterns and use of treatments for psoriasis, membership of the Psoriasis Association, and social class. The second and third sections of the questionnaire comprised the British version of the health survey questionnaire SF-36 and the psoriasis disability index.
The SF-36 is a self administered questionnaire that is easy to use, acceptable to patients, and fulfils stringent criteria of reliability and validity.10 11 We used published normative data based on 9332 adults of working age (18–64 years) from Berkshire, Buckinghamshire, North-amptonshire, and Oxfordshire for comparison.12 The psoriasis disability index was designed by Finlay et al13 and contains 15 points answered on a 1 to 7 linear analogue scale. It covers five main areas, including daily activities, school or work, personal relationships, leisure, and treatment. Comparisons with the sickness impact profile and the psoriasis area and severity index have shown that it is an appropriate and quick method of assessing psoriasis disability in patients attending a hospital dermatological department.8 It has not been validated in any general practice based studies.
We assessed the relation between the psoriasis disability index score and age, and also each of the SF-36 health measures using the correlation coefficient. The effect of social class and sex was assessed by analysis of variance. As SF-36 scores vary with age and sex, the individual scores for each SF-36 variable were converted to a z score, where z = (observed score-mean for that age and sex group) × SD for that age and sex group. The z scores are normally distributed with a mean of 0 (SD = 1). For each variable, the scores for the non-manual and manual groups were compared with those published for healthy controls12 by testing if the mean z score was different from 0 (using the t test); 0 is the expected mean z score if the data were randomly drawn from the control population. We compared mean z scores between the non-manual and manual classes by the two sample t test. SF-36 questionnaires containing missing values were scored strictly according to the instructions provided by the original authors.
We identified 767 patients with psoriasis from the five participating practices. The total list size was 58 257, giving a prevalence of 1.3%. A total of 546 returned completed questionnaires after two postings, a response rate of 71%. Of these, 68 (12%) felt that they did not have psoriasis and 43 (8%) did not wish to take part, leaving 435 patients for analysis.
Fifty six patients had consulted a doctor about their psoriasis within the past month, and a further 183 had consulted in the past year; 196 had not consulted a doctor about their psoriasis in the past year. Only 52 subjects had been seen or treated for psoriasis in hospital or as an outpatient in the past year. However, 274 said that they had used a treatment for their psoriasis in the past week.
Although 170 patients had heard of the Psoriasis Association, only five were members.
When asked, 239 respondents expressed an interest in more information about psoriasis and 300 in advice about treatments and how to use them. Only 39 were interested in support from other people with psoriasis.
Figure 1 shows that for all eight scales of the SF-36, our subjects had generally lower scores than the general population, indicating that they perceive themselves to have poorer health.
Subjects in the manual social group had significantly lower perception of health than controls for all SF-36 variables whereas subjects in the non-manual social group were apparently not impaired in terms of physical function, social function, or pain (table 1).
Table 2 shows that all the SF-36 variables were highly negatively correlated with the psoriasis disability index score. Table 3 shows the breakdown of mean psoriasis disability index scores by social class and sex. Analysis of variance showed a highly significant difference with social class (P = 0.001) but no effect with either sex (P = 0.92) or social class and sex together (P = 0.96).
It is concerning that in this large community based study of psoriasis over two fifths of patients had not been seen by a doctor for over a year, even though almost two thirds continued to use some kind of treatment. Knowledge of the Psoriasis Association was quite high but not reflected in the figures for membership.
Previous studies have looked at selected groups in whom the disability might be expected to be worse.3 4 5 13 14 Nevertheless, even in those selected groups doctors underestimate the degree of disability caused by psoriasis.3 4 Our results suggests that people with psoriasis in the general population suffer serious disability in many aspects of daily life compared with healthy controls. Scores on the psoriasis disability index showed strong negative correlations with all variables on the SF-36 general health questionnaire, and the mean scores were lower than the overall mean reported in Finlay et al's hospital study (38.8%, SD 23.3).5 The consistently lower scores of our patients on the SF-36 questionnaire (fig 1) also support this conclusion. In particular, we found that the perceived social function of psoriatic patients in the manual social class is significantly lower than both the controls and the non-manual patients in most categories of the SF-36 questionnaire.
The high correlation between the SF-36 and the psoriasis disability index suggests that this index is a suitable tool for assessing morbidity among psoriatic patients in the community. However, in view of the effects of social class on the scores, this factor would have to be taken into account in any analyses.
Our control data were derived from the largest published British study using the SF-36 questionnaire. The study was intended to provide normative data for comparison with other populations and samples. The sociodemographic characteristics of their sample mirror closely the characteristics of the general population when compared with 1991 population estimates and the social class distribution in the 1981 census. Although subsequent work has shown that there may be regional variations in normative values for the SF-36, the differences were not found to be due to age, sex, or social class.15 We could have selected controls from the same population as the subjects, but this also leads to problems of selection bias and such controls cannot be considered normal for a geographical area.15
In conclusion, we have found that people with psoriasis perceive themselves to be less healthy than a normal population. The most severely affected are manual workers, and these might be a priority group to be targeted for particular attention. The SF-36 with the psoriasis disability index may be a useful means of identifying those in most need of intervention and of auditing their progress.
We thank Mr Leon Green, Cleveland Medical Audit Advisory Group audit assistant for help in data collection and entry and the five practices and their patients who agreed to take part in the study: Tennant Street Surgery, Stockton; Village Medical Centre, Middlesbrough; Yarm Health Centre; Thornaby and Barwick Medical Group, Thornaby; and Norton Medical Centre, Stockton. We also thank Dr Pali Hungin and other members of the Northern Research Network for help and advice, Dr Andrew Finlay for permission to use the psoriasis disability index, John Ware and colleagues at the Health Institute of the New England Medical Centre for permission to use the SF-36 questionnaire.
Funding Psoriasis Association and Cleveland Primary Care Research Panel.
Conflict of interest None.