Rasch analysis of the dermatology life quality index (DLQI) in patients with mild to moderate-severe psoriasis

Plain Language Summary

Psoriasis is a chronic inflammatory disease that can impact the personal appearance, which may increase susceptibility to psychological disorders impacting the patient’s quality of life. The Dermatology Life Quality Index (DLQI) is one of the primary methods for its good psychometric performance. Our study analysed its reliability and validity in 1329 Colombian adult patients with mild to moderate-sever psoriasis. In general terms, the psychometric findings favor the DLQI’s validity for its use among these patients. However, we identified a significant bias when the DLQI is applied to older adults with psoriasis. We recommend developing a specific version of the DLQI for older adult patients to capture the relevant aspects of quality of life from their perspective, avoiding misclassification of the impact of the disease on their daily living.

Introduction

Psoriasis is a chronic inflammatory disease with significant social and healthcare implications. Its prevalence is unclear due to the heterogeneity of the clinical manifestations and diagnostic and classification criteria; however, it is estimated to affect between 2–3% of the global population. Although the clinical manifestations are variable, in most patients there is a great effect on their quality of life. Psoriasis can be associated with a significant economic burden associated with medical care, occupational costs, and social dependency, which can be compared with the financial burden caused by other diseases such as cancer, melanoma, and asthma.^1,2

As with other dermatological diseases, psoriasis can impact personal appearance, generating greater susceptibility to developing psychological disorders, sleep disorders, and depression that affect work and daily activities.³ In this sense, the health-related quality of life (HRQOL) concept becomes more relevant, considering the disease’s impact on the patient’s social relationships, psychological status, and daily functioning.

The Dermatology Life Quality Index (DLQI) is one of the primary methods for measuring health-related quality of life in patients with dermatologic diseases. This instrument has been analysed in various systematic reviews, highlighting its ease of use due to its brevity and simplicity, with an average response time of 2 minutes. It can be used in approximately 30 dermatologic diseases with empirical support for its external validity. Additionally, it has been established that the DLQI is sensitive to the quality-of-life changes at different stages of the disease.^4,5

Despite its strengths, several studies have reported significant differences in its performance, depending on patient demographic and clinical characteristics. Age, sex, and disease severity are characteristics that condition psychometric performance. Thus, verifying DLQI psychometric performance in local studies is essential to ensure its suitability to support decision-making.^6,7 In Colombia, no specific studies have evaluated the psychometrics of DLQI, by focusing on its potential bias among clinical and demographic-relevant subgroups.

This study uses the Rasch model to analyse the reliability and validity of DLQI to measure the quality of life in patients diagnosed with mild to moderate-severe psoriasis at the Comprehensive Clinic for Immune-Mediated Skin - CLIPSO (Medellin, Colombia).

Materials and Methods

Results

Differential item functioning (DIF)

Table 4 shows the differential item functioning analysis of the DLQI. Type B DIF was found in the item “clothing” for women (male reference). In the older adult age group, type C DIF was detected for the item “work and study”, and type C DIF was detected for items “itching, discomfort, pain or stabbing sensation” and “sexual difficulties”. Additionally, type B DIF was detected for severe PASI in “shopping, taking care of the house or garden”.

Table 4: Differential Item Functioning (DIF) according to sex, age, and psoriasis severity

Ítem	Sex		Age			Psoriasis severity
	Female	Male	Young adult	Middle adult	Older adult	Mild	Moderate	Severe
1. Itching, discomfort, pain, or stinging sensation	–0.3*	0.3	0.2	0.1	-0.6B	0.0	–0.1	0.2
2. Embarrassed, self-conscious	–0.1	0.1	–0.2*	0.1	0.0	0.0	–0.3	0.4
3. Shopping, taking care of the house or garden	–0.2	0.1	0.3*	–0.2	-0.2	–0.1	0.1	0.6B
4. Clothing	–0.6B	–0.3	–0.2*	0.1	0.4*	0.0	–0.1	–0.4
5. Social and recreational activities	0.2	–0.1	–0.1	0.1	0.0	0.1	–0.3	0.0
6. Sports	0.3*	–0.1	0.2	–0.1	-0.1	0.0	0.1	–0.2
7. Work and study	–0.1	–0.1	0.3	–0.4*	1.4C	0.1	0.1	–0.7C
8. Partner, friends, family	0.2	0.3	–0.1	0.1	0.0	–0.0	0.2	0.1
9. Sexual difficulties	0.3*	0.4	–0.2*	0.1	0.6B	–0.0	0.2	–0.2
10. Treatment	0.4*	–0.5	0.2	–0.1	0.0	0.0	0.0	–0.1

B: mild to moderate (DIF ≥0,43 y p-value <0,05); C: DIF moderate to severe (DIF ≥0,64 y p-value <0,05); *: p-value < 0,05.

Figures 1a and 1b show the Bland and Altman limits of agreement for estimating the quality-of-life measure of the original scale vs. the bias-free scale, according to age and the PASI. For the group of older adults [Figure 1a], the DIF items caused a considerable bias in the estimation (bias = 0.12; 95% CI 0.11–0.12); this bias has a differential behaviour according to score level. Concerning the PASI [Figure 1b], the bias in the estimation was not relevant (bias = 0.01; 95% CI 0.00–0.02).

Close

Figure 1a:

Bias for older adults.

Export to PPT

Close

Figure 1b:

Bias for severe psoriasis.

Export to PPT

Unidimensionality, reliability, and item mapping

The measure obtained from the DLQI explained 60.9% of the total variance. No evidence of multidimensionality was found, with the eigenvalue of the first residual contrast = 1.7 (6.7%). The disattenuated correlation between clusters 1 and 3 of the first contrast was 0.70. There was no evidence of autocorrelation between residuals of items with correlations lower than 0.4 in all cases. Reliability was 0.86. Figure 2 shows the item-person map for the best-fit sample, and Figure 3 shows the item-person map for the older adults.

Close

Figure 2:

Item-person map for the best-fit sample of DLQI items in all the samples.

Export to PPT

Close

Figure 3:

Item-person map of DLQI persons in older adults.

Export to PPT

Discussion

The results of this study add evidence to the validity of DLQI in Colombian people living with psoriasis, providing information on the classification and response profile that facilitates its interpretation and usefulness for clinical decision support.

According to the category function results, the patients evaluated can discriminate against their psoriasis-related quality of life impairment. However, it is necessary to consider that the instrument does not adequately capture potentially relevant differences among low scorers due to the floor effect. This implies that DLQI allows the identification of people with lower quality of life, but its performance in characterising less affected patients is inadequate. Consequently, improved characterisation of the HRQOL gains due to the patient’s response to treatment would require including items that reflect daily life functionality better.

The question related to work and study presented some item fit difficulties, mainly among older adults and patients with a severe PASI classification. It should be noted that this item is the only one with different response options in the questionnaire, which could explain in some way the inconsistencies found. Authors such as Mørk et al.¹² in their validation study of DLQI in the Norwegian population, agree with this hypothesis and suggest that a possible solution is to change the response options to ones like the other items.

In this study, DIF was identified in three of the ten items for the older adults subgroup (sexual difficulties, work and study, and skin discomfort), which, according to Bland and Altman’s agreement, induced a relevant bias in the overall DLQI estimation. It is essential to consider that the meaning of quality of life in older adults may have a different perspective. In this sense, for Vera et al.,¹³ the meaning of quality of life for this population focuses mainly on tranquillity, care, protection, affection, and respect, helping them to achieve their goals and allowing them to express themselves freely and make decisions.

Cosco et al.¹⁴ state that the challenges of the young population differ from those of the adult population, so they consider it relevant to adapt or create scales to measure resilience in this population. A study by the National Autonomous University of Mexico on the Rosenberg Scale in the older adult population indicates that self-esteem may be affected by variables such as social, physical, and psychological changes at this stage, which may explain the inconsistencies found in the items on sexual relations and work and study.¹⁵ Another factor to consider is that social isolation increases as we age, and cognitive and physical functions deteriorate, as a Brazilian study on life purpose in older adults shows. This may explain the results in the work and study item.¹⁶

Our results from the Rasch analysis are similar to those previously reported in other populations. Twiss et al.¹⁷ found age, gender, and severity-related differential item functioning in patients from the United Kingdom diagnosed with psoriasis or atopic dermatitis. However, unlike our findings in Colombian patients, they identified more pronounced issues in items 2 (“embarrassed, self-conscious”) and 10 (“treatment”). Poor fit has also been found in patients diagnosed with other specific dermatological conditions such as hand eczema,¹⁸ neurodermatitis,¹⁹ and in patients with various skin diseases,²⁰ leading to recommendations for modifications to the DLQI.

Regarding the specific issues with item 7 (“work and study”), Rencz et al.²¹ also identified this limitation in Hungarian patients with psoriasis and concluded that treating missing responses as a score of 0 is the best alternative to reduce misfit and DIF. It is important to note that item 7 is the only item in the DLQI with a conditional response.”

This study has some limitations. It was performed at a single centre, with most patients presenting a mild PASI score, thus limiting generalisation of the findings. However, our findings mainly refer to the scale psychometric performance and are concordant with previous Rasch analyses, as described before. It is also relevant to highlight that even though the sample size was not estimated a priori for the psychometric analyses, this did not represent a limitation related to statistical power and DIF contrasts were statistically significant. Lastly, the findings related to measuring the quality of life in older adults with psoriasis must be confirmed in additional studies that design and validate adaptations of DLQI to capture relevant effects on their daily living activities.

Conclusion

The validity evidence favours the use of DLQI in moderate-severe psoriasis. However, the instrument has biases in its estimates for older adults. Therefore, for this population group a modified version of the instrument may be considered.