Quality of Life Measures for Dermatology: Definition, Evaluation, and Interpretation

For the specific purpose of this report, a Medline search was conducted November 2011, including the search terms “patient reported outcomes”, “quality of life”, “patient preferences” and “willingness to pay”. These were cross-referenced with search terms indicating dermatologic disorders in general (e.g. “skin diseases”) and specific diseases (e.g. “psoriasis”). The literature search resulted in 25.450 publications relating to patient reported outcomes in dermatology, including 28 % of publications on HRQoL, 25 % on patient satisfaction, 18 % on psychological factors, and 8 % on others (including willingness to pay and patient-benefit index).

The most frequently focused dermatologic diseases were chronic inflammatory skin diseases such as psoriasis and atopic eczema (24 % and 22 %, respectively), followed by allergies (16 %) and skin cancer (12 %). The publications included reviews (45 %), case reports or other original publications (25 %), and interventional studies (20 %).

The term quality of life is understood very differently in everyday life and in research. Therefore, a prior precise definition of the term is crucial for its scientific use. According to the general scientific notion, QoL is a multidimensional construct that is not directly measured but can only be displayed in its single components [5, 6]. There are differing opinions regarding the areas to be included. Based on a fundamental WHO definition on health, QoL includes the physical, psychological, and social condition of an individual [7]. Several authors point out that QoL encompasses much less the objective availability of material and immaterial things as the degree with which a specifically desired state of physical, psychological, and social condition is achieved. Bullinger [8, 9] and Schipper [10] define HRQoL as the quality of the physical, psychological, social, and role- and function-associated life situation of an individual. Also taken into account in QoL is the degree of concordance between the desired and the real-life situation.

Generic QoL must be distinguished from HRQoL. The latter encompasses all QoL areas that pertain to the relevant dimensions of individual health; however, generic HRQoL must also be distinguished from disease-specific HRQoL. The former pertains to aspects of QoL and how they can appear independent of a specific illness, while the latter focuses on particular characteristics of a certain illness.

Even though an “objective” recording of QoL would be preferable, it is clear from its definition that QoL can only be recorded by self-evaluation from the patient’s perspective [11]. With respect to QoL measurement, basic methodological difficulties result from the fact that QoL isn’t directly observable but can only be quantified by means of a model.

Despite the methodological difficulty of making QoL measurable as a phenomenon, several instruments have proven beneficial in ascertaining generic HRQoL. These are normally made up of standardized questionnaires or inventories that are completed by the patients (self-evaluation) or by examiners or relatives (third-party evaluation or proxy rating). QoL questionnaires that have only a few individual items that are collectively assigned to a scale are termed indexes (eg, disability index).

In addition to generic questionnaires on QoL, numerous inventories that measure specific, disease-related aspects of QoL have been developed. The advantage of disease-specific instruments lies in the precise evaluation of burdens that mainly apply to those affected by the particular illness, but not for the sick in general. Moreover, clinical courses are usually best recorded by means of disease-specific questionnaires (sensitivity to change of questionnaires).

Basic studies show that QoL, despite methodological limitations, can be reasonably measured and reflects the course of a disease in a reproducible manner. The methodology and the implications have been consented in several guidelines: an FDA paper [12], an EMEA recommendation [13], the German national guidelines on QoL assessment in general [14] and in dermatology specifically [15] and a consensus paper of an EADV task force [16]. For the reporting of HRQoL outcomes in trials, a specific CONSORT statement is under development [17]. Taken together, the guidelines give comparable recommendations for clinicians to consider during the planning phase of dermatologic evaluations. These recommendations will be discussed further in the following sections.

Because of its straightforward manageability, standardized questionnaires for patient self-evaluation have proven their value in the assessment of HRQoL [18]. Compared with interviews, they have the advantage of faster data acquisition and analysis that is not dependent on the interviewer. Aside from the greater time efficiency, the methodological mistake that results from the patient’s manipulation by the examiner is also prevented, as compared to third-party interviews. Self-evaluation requires the willingness of patients to provide information about themselves. Based on our clinical experience, this willingness is present in almost all patients, especially if they volunteered to participate in a therapy study.

If the to-be-examined patient group cannot provide personal information (eg, toddlers), third-party evaluations, or proxy ratings, can also be held. These evaluations also require standardized instruments. Proxy ratings are interviews with the patient’s relatives or the therapist, and are to be used if the patient is not able to provide personal information [19]. They are mostly needed in studies with younger children. The age of 8 was determined as the lowest age limit for written personal information according to numerous authors [20]. However, smaller children can also provide information via pictures (eg, “smilies”) or an interview form of the questionnaire, when applicable. Special inventories (ie, different from adult versions) are likewise required for older children [20].

In QoL measurement, multidimensional questionnaires, which contain multiple scales, provide greater methodological robustness than one-dimensional questionnaires or single questions [5]. However, because of organizational reasons, it is not always feasible to use extensive questionnaires in therapy studies. It is possible instead to employ only selected scales of the questionnaire, provided that these scales are valid and that the author of the questionnaire approves of this action. It should, however, be emphasized in the study protocol that QoL would only be measured in part. The use of single questions alone is methodologically insufficient in deriving statements regarding QoL. Similarly, focusing on single symptoms of disease alone (eg, physical signs only) cannot be considered a comprehensive evaluation of QoL.

Despite the emergence of international research on HRQoL assessment at this time, a lot of methodological questions—for example, the election of the optimal HRQoL instrument—have not been conclusively resolved. However, the advantages and disadvantages of the most established HRQoL questionnaire for application in dermatology were described in a comparative German study [21].

Figure 1 breaks down the complete decision-making process for selecting an appropriate HRQoL instrument.

Open image in new window

Preferably, instruments that measure the generic factors of HRQoL, as well as the disease-specific areas, should be used. The advantage of disease-specific questionnaires is their oftentimes higher differentiation capabilities and greater sensitivity to change [22, 23], whereas the advantage of generic questionnaires is better comparability with other disease groups. Given the advantages of each, the majority of health economic and clinical pharmacologic associations recommend the combined use of disease-specific as well as generic HRQoL questionnaires [24, 25].

Internal consistency means that the items of the questionnaire should be homogenous in representing its scale to ensure high-quality data. Today, a minimum value of 0.70 is required internationally for Cronbach’s alpha.

Retest reliability means that the questionnaire should render approximately the same results in repeated measurements on the same patient. For example, the retake is carried out at 1-week intervals without an intermittent clinical intervention.

Convergent validity is achieved when the instrument is in accordance with other instruments that are supposed to measure the same construct. This should be determined in a large patient sample through comparison with validated inventories.

Discriminant validity means that LQ instruments should not considerably correlate with instruments that are supposed to measure different constructs.

Sensitivity to change means that changes in the QoL over time, as well as therapy effect, must coincide with the corresponding changes in the test result.

Objectivity of a HRQoL instrument is given when measurement is independent of the test procedure, the setting and the test administration, and when the results are independent of the method of analysis.

Should a QoL questionnaire be deployed within a clinical study, it should not needlessly disturb the course of the study. Also, the questionnaire’s contents, graphic presentation, mode of questioning, and instructions should ensure good comprehension and high acceptance by the patient. The questionnaire should be of adequate length to ensure both high reliability of the inventory and minimal burden on the patient; should be self-explanatory, and should provide a favorable layout so the patients are able to complete it without help. In addition, the analysis of the questionnaire should be as simple as possible.

All single questions that pertain to a certain dimension of QoL form a subscale. The subscale value can be calculated from the average score or from the cumulative values of the single questions. For reliability reasons, it is reasonable to calculate the subscales of QoL with not just one but several questions each. Thus, outliers of single items and coincidental answers carry less weight.

A large number of studies have shown that the understanding and the results of HRQoL instruments depend on cultural, educational and semantic factors [26, 27]. For this, a very accurate use of any instrument adapted to the specific target group is needed. If questionnaires from other languages are to be used, they should be translated and re-translated in a systematic manner, including independent double-translations into both directions. After that, the resulting translations need to be compared in order to gain a consented final version. Finally, controls for validity in any new language are recommended.

The comparison of averaged results with other patient samples is essential to the interpretation of the QoL results. This can be patient subgroups of the same study or patients of other studies who have the same or a different illness. Therefore, it is desirable that comparative data of the QoL questionnaire can be derived from a reliable data pool.

Should a cost-benefit analysis be conducted within the clinical study, and the QoL be adopted as a benefit, there should be a utilizable score that is calculable for this purpose (eg, a mapping to utility measures would be feasible).

The construction of a new questionnaire is an elaborate process that should only be approached with psychometric help. Hence, it is preferable to use existing questionnaires. The development of a questionnaire is exploratively possible in a study parallel to the use of a standardized procedure, but it must run through several stages: item retrieval (eg, focus groups), formulation/verbal pretest, implementation on at least 100 patients, psychometric analyses for reliability and validity, and retake of measurements after one week (without intervention) for the determination of the retest reliability and after an intervention for the determination of sensitivity to change [5, 28].

The following instruments are for use with children specifically: Children’s Dermatology Life Quality Index [41], Pediatric Symptom Checklist [42], and a version of the Kindl questionnaire for children with atopic dermatitis [43].

QoL questionnaires can mostly be obtained from the particular authors. General overviews can be obtained from Bullinger [5, 26] and Bowling [18], and QoL questionnaires on dermatology from Salek [32], Finlay [45], and Kupfer [44]. Usage rights also lie, in part, with the publishing company (leading in Germany: Hogrefe publishing and Beltz-Test), so that the questionnaires must be obtained and, if applicable, royalties paid.

To better understand the QoL construct and develop strategies for improving HRQoL in dermatology, predictor studies on a variety of indications have been performed. The predictors identified both depended on disease-specific factors and on the predictors included in the model. For example, HRQoL in psoriasis, as measured by the Dermatology Life Quality Index, was most strongly predicted by the time needed for treatment [46]. In another predictor study on patients with atopic eczema, HRQoL was predicted by four major factors: clinical symptoms, emotional distress, problems with social contacts, and helplessness. These differences support the development of disease-specific strategies for the improvement of QoL in patients with skin diseases.