Introduction

Systemic lupus erythematosus (SLE) is an autoimmune disease with a significant impact on physical, social, and psychological health. Currently, disease activity or damage indices composed of laboratory, radiological, or clinical findings emphasizing physicians’ assessments are mostly used in clinical care and research to evaluate health outcomes. However, these have been found to be poorly correlated not only with patient assessments of disease activity [1] but also with quality of life [2] in patients with lupus. In fact, the patient is the best source for assessing the effects of the disease or its treatment on daily life. Therefore, patient-reported outcomes (PRO), used in conjunction with physician assessments of disease activity and damage, are the preferred assessment methods both for patient care and research. Moreover, the use of a patient-reported outcome measure (PROM) as a core outcome measure for SLE is highly encouraged, and the Food and Drug Administration requires its use in clinical trials, especially for licensing and promoting new medications [3].

LupusPRO is a disease-targeted patient-reported outcome measure that was developed and validated from and among US patients with SLE [4]. The LupusPRO includes a comprehensive assessment (health and non-health) with quality of life and pertinent SLE domains that are not yet available in other SLE-specific patient-reported outcome measures. Its Spanish version has also been recently validated [5]. None of the validated disease-specific PROM is currently available in Turkish. Therefore, we aimed to cross-culturally adapt an existent disease-specific PRO tool for use in Turkish SLE patients. We herein report the results of the cross-cultural validation study of the translated version of the LupusPRO in Turkish SLE patients.

Methods

LupusPRO has two constructs: health-related quality of life (HRQoL) and non-health-related quality of life (Non-HRQoL). HRQoL domains are (1) lupus symptoms, (2) lupus medications (3) physical health [themes: physical function and role physical], (4) emotional health [themes: emotional function and role emotional], (5) pain/vitality [themes: fatigue, sleep], and (6) procreation [themes: sexual health and reproduction], (7) cognition, and (8) body image. Non-HRQoL domains are (1) desires/goals, (2) coping, (3) relationship/social support, and (4) satisfaction with care. Individual domain scores, total HRQoL, and total non-HRQoL scores range from 0 to 100, where higher scores indicate better health.

The study was approved by the Institutional Review Board of Gazi University Medical Faculty. Forward and backward translations of the 43-item English LupusPRO were undertaken using standard guidelines [6], to develop the Turkish LupusPRO. The tool was pretested in five individuals and finalized based on their feedback.

Adult patients (≥18 years old) meeting the American College of Rheumatology (ACR) classification criteria for SLE [7] were eligible for enrollment if they were able to read and understand Turkish. The finalized Turkish version was applied to consenting Turkish-speaking SLE patients. Data on the following variables were collected at baseline visit (T1): Demographic information, clinical and serological characteristics, disease activity, and damage assessment. In addition, the Turkish LupusPRO (T1), the Turkish version of the Medical Outcomes Study Short Form (SF-36) [8], was self-administered to all participating patients at that time. Higher scores on SF-36 denote better health. For test–retest validity assessment, a second copy of Turkish LupusPRO questionnaire was provided to the patients to be completed within 2–3 days of T1 (T2) to be mailed or dropped back in person to the study site.

Disease activity was evaluated using physician global assessment (PGA), a modified version of Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) that was developed for the Safety of Estrogens in Lupus Erythematosus National Assessment (SELENA) trial (SELENA-SLEDAI) [9] and the Lupus Foundation of America (LFA)-defined Flare (Yes/No) [10]. Cumulative damage from either the disease or therapies used for it was assessed using the Systemic Lupus International Collaborating Clinics/American College of Rheumatology (SLICC/ACR) damage index (SDI) [11]. ‘SF-6D’ is another parameter which reflects overall health status and was also applied to assess disease activitiy. The SF-6D has been developed subsequent to the introduction of the SF-36 and provides a method to transform responses from select items of SF-36 domains into an index based SF-6D measure that contains six domains [12].

Psychometric properties studied included the following: internal consistency and reliability for each domain were evaluated using Cronbach’s alpha, where an alpha >0.70 is considered acceptable [13]. Test–retest reliability was tested by evaluating the agreement between the patient responses to each domain at two time points. Convergent validity was evaluated based on the strength of correlation of LupusPRO with related domains on SF-36 using Spearman’s correlation coefficient. Criterion validity of the LupusPRO was judged based on its correlation with either health status, measures of disease activity, or damage. Correlations were classified as strong (r > 0.5), moderate (0.3 ≤ r < 0.5), weak (0.1 ≤ r < 0.3), or absent (r < 0.1). The tool was considered responsive if changes in its domain or summary scores correlated with changes in health status and/or disease activity and in the expected direction. The conceptual framework (hypothesized item to scale relationships) of the LupusPRO was evaluated using confirmatory factor analysis (CFA) appropriate for categorical data. CFA was conducted with the LupusPRO item responses using a robust weighted least squares estimator and the software Mplus [14]. Mplus employs a multi-step method for ordinal outcome variables that analyzes a matrix of polychoric correlations rather than covariances. The goodness of fit of the hypothesized item-to-scale relationships (multi-factor) was evaluated with the comparative fit index (CFI) and Tucker–Lewis index (TLI). CFI and TLI are comparative fit indices, which quantify the amount of difference between the examined model and the independence model (i.e., a standard comparison model that asserts none of the components in the model is related), with higher scores indicating larger differences. It is recommended that these two indices be 0.9 or greater as evidence of acceptable model fit [15]. All reported p values are two-tailed, and a p value <0.5 is taken significant.

Results

One hundred two patients with SLE (94 % women) were enrolled in this study. All patients have been followed at the Gazi University Medical Faculty, Department of Internal Medicine, Rheumatology outpatient clinic. The median (IQR; min. to max.) age and mean disease duration (±SD) were 38.5 (18; 18 to 67) years and 60.3 (±56.3) months, respectively. Forty-four percent had less than high school education, and 78.4 % were currently married. Of the 102 patients, 99 (97.1 %) had cutaneous manifestations, 8 (7.8 %) had history of serositis, 27 (26.5 %) had lupus nephritis, 11 (10.8 %) had neuropsychiatric manifestations, 25 (24.5 %) had hematological manifestations, 54 (52.9 %) had musculoskeletal manifestations, and 11 (10.8 %) had thrombo-embolic manifestations. Rates of these manifestations are cumulative occurances for the entire disease duration. The mean ± SD, SELENA-SLEDAI, and SDI were 3.1 ± 3.7 and 0.52 ± 0.75. There were 25 patients with LFA-defined flare status (11 patients had mild, 8 patients had moderate, and 6 patients had severe flare) at the baseline visit (Table 1). PGA scores were 0 for 26 %, 1 for 60 %, 2 for 11 %, and 3 for 5 % of patients. Forty-five percent of patients were taking prednisone, 87 % hydroxychloroquine, 16 % azathioprine, and 6 % mycophenolate mophetil at the time of the study.

Table 1 Description of the study cohort
Full size table

Mean scores on the LupusPRO domains are shown in Table 2. Our patients scored poorly especially in the emotional health and social support domains. Internal consistency reliability (ICR) of the domains ranged from 0.63 to 0.94. The alpha values of ICR tests were over 0.70 for all domains except lupus symptoms domain. Test–retest reliability of the LupusPRO domains ranged from 0.87 to 0.98 (Table 2).

Table 2 Descriptive scores of the LupusPRO domains in Turkish-speaking Lupus patients (0 = worse QoL, 100 = Best QoL) and psychometric properties of Turkish LupusPRO
Full size table

Convergent validity of LupusPRO was confirmed using the corresponding domains of SF-36. As expected, the physical health domain of LupusPRO had good correlations with physical functions (r = 0.65, p = 0.001) and role physical (r = 0.64, p = 0.001) domains of SF-36. Similarly, the emotional health domain of LupusPRO correlated well with mental health (r = 0.52, p = 0.001) and role emotional (r = 0.53, p = 0.001) domains of SF-36. The pain/vitality domain of LupusPRO correlated well, as expected, with the vitality (r = 0.79, p = 0.0001) and bodily pain (r = 0.68, p = 0.001) domains of SF-36 (Table 3).

Table 3 The construct and criterion validity results of Turkish LupusPRO
Full size table

Criterion validity of the HRQOL domains of the LupusPRO was also demonstrated against disease activity measures (PGA, SELENA-SLEDAI, LFA flare, and SF-6D) (Table 3). Significant correlations between patient-reported change in health (since last visit) with cross-sectional domain scores in the anticipated directions were noted in the majority of domains, including lupus symptoms, cognition, physical health, pain/vitality, desires/goals. PGA correlated well with all the parameters tested, except cognition. Statistically significant correlation between SELENA-SLEDAI scores and LupusPRO was found at physical health domain only (r = −0.27, p = 0.006). LFA flare had significant correlation only with emotional health domain of LupusPRO (r = −023, p = 0.02). Interestingly, the best correlations were observed between SF-6D and LupusPRO domains. There were significant correlations between SF-6D with all of LupusPRO domains, except desires/goals domain, which also correlated well but had borderline statistical significance (r = 0.42, p = 0.05).

Results of confirmatory factor analysis lend empirical support for the conceptual framework of the LupusPRO (Table 4). The model fit for the hypothesized item-to-scale relationships was satisfactory (CFI = 0.97, TLI = 0.98). In addition, item-to-factor loadings representing the hypothesized item-to-scale relationships were also satisfactory. In general, items loaded >0.6 with their respective factors.

Table 4 Confirmatory factor analysis of HRQoL and non-HRQoL items of Turkish LupusPRO
Full size table

Discussion

SLE itself or its treatments may profoundly affect the physical, as well as the mental, social, psychological, and sexual well-being aspects of a person’s life. Thus far, patient-reported outcomes in lupus have usually been assessed in longitudinal studies and in some clinical trials using a generic QoL tool [16, 17]. Short Form 20 was found to perform better than Health Assessment Questionnaire [18]; however with the availability of the SF-36 and its Vitality domain, the latter was preferably used in lupus studies. Responsiveness of SF-36 to changes in disease status is controversial. Some studies have reported responsiveness; however, others have revealed it is either not responsive or changes observed are not related to lupus but non-lupus manifestations, such as fibromyalgia [17, 19, 20].

Development of disease-specific or targeted QoL measures is a relatively new effort in lupus and is needed to fill in the gaps in the generic patient-reported health outcome measures [21–23]. During the validation of the LupusQoL in the US, patients noted that some important issues were not captured by LupusQoL, e.g., side effects of lupus drugs, memory and concentration, concerns for pregnancy, effect on vocational aspirations/finances, social supports available, medical care, etc. Furthermore, LupusQoL was derived from female patients with SLE. LupusPRO, a lupus-specific patient-reported outcome measure, was therefore developed from an ethnically heterogenous group of the US SLE patients of both genders (4) with the intent of capturing all pertinent issues affecting their quality of life from their own perspectives. It has fair psychometric properties, and preliminary studies indicate responsiveness to change [24].

The Turkish version of LupusPRO was cross-culturally adapted and validated. The study has included culturally diverse Turkish SLE patients from Ankara (the capital city of Turkey) and surrounding area, attending rheumatology outpatient clinic. LupusPRO exhibited fair reliability and validity. Significant correlations between patient-reported changes in health status since last visit with cross-sectional domain scores were also noted. Confirmatory factor analysis supported the hypothesized item-to-scale structure as the original LupusPRO and had an excellent fit.

The lack of longitudinal data to assess its responsiveness and minimally important difference of the different domains of the LupusPRO is certainly a limitation of our study. Disease activity scores of the study patients were lower than the patients in the US and Spanish validation studies.

Finally, the instrument may need local adaptations before its widespread use in Turkey. Strengths of the study include participation of multicultural Turkish-speaking SLE patients, as well as the inclusion of the LFA definition of flare. Although it was not part of the study methods, spontanous feedback about the LupusPRO questionnaire from the SLE patients in this study has revealed that lupus patients have difficulty differentiating lupus-related and non-lupus-related nature of their symtoms and problems they experience, suggesting that they are in great need of more information about their disease and medications.

In conclusion, the Turkish version of the LupusPRO is valid and appears to perform comparably to the English and Spanish language versions. It can be used as a patient-reported outcome parameter in clinical trials, as well as longitudinal studies for testing responsiveness to change. Using a cross-culturally adapted and validated patient-reported outcome measure might facilitate homogeneity in the methodology of multicultural lupus studies, as well as provide insight into cultural/geographic differences in health outcomes. Moreover, the LupusPRO may provide unique information to understand the feelings of SLE patients and total impact of the disease on their lives.