Clinical outcomes in psoriatic arthritis: A systematic literature review

Authors


Abstract

Objective

Many outcomes have been proposed in the assessment of psoriatic arthritis (PsA). The Outcome Measures in Rheumatology (OMERACT) core set for PsA evaluation comprises 6 domains: joints, skin, function, pain, patient's global assessment, and quality of life. The objective of this work was to assess reporting of outcomes in PsA, including patient-reported outcomes (PROs) in recent publications.

Methods

A systematic literature search of clinical trials related to PsA and reporting at least 1 clinical outcome between 2006 and 2010 was performed in PubMed, i.e., just before to just after publication of the OMERACT core set. All clinical outcomes were noted and subdivided into domains of health. Data analysis was descriptive.

Results

Fifty-eight articles (12,405 patients) were included in the analysis: 17 (29%) were randomized clinical trials; the patients' mean ± SD age was 48.2 ± 5.4 years and the mean ± SD disease duration was 9.0 ± 3.1 years. Eighty-four different outcomes were reported, with a mean ± SD of 6.9 ± 4.3 per study. Patients were mainly assessed using the 6 core set domains, reported in 37.9% (quality of life) to 55.2% (skin) of articles; however, the core set was rarely completely reported since only 10.3% of the studies reported all 6 core domains. PROs were heterogeneous and in particular there was no consensus regarding the number of joints to assess and instruments for dactylitis and enthesitis. PROs were assessed in more than 75% of publications using 28 different instruments.

Conclusion

There is great heterogeneity in PsA assessment, even since publication of the OMERACT core set. Better consensus on instruments to assess each domain of health and better insight into which outcomes are important for patients is needed.

INTRODUCTION

Psoriatic arthritis (PsA) is a heterogeneous disease, requiring different outcomes and corresponding instruments to evaluate patients' health status, disease activity, and treatment efficacy (1).

Outcome measures in PsA are not standardized and most of the assessment methodologies have been adapted from clinical trials in rheumatoid arthritis (RA), with few disease-specific instruments for PsA currently available (2). Recently, the Outcome Measures in Rheumatology (OMERACT) proposed a core set of 6 domains of health to be included in randomized clinical trials (RCTs) and observational studies regarding PsA: joints, skin, pain, patient global assessment, physical function, and health-related quality of life (3). This consensus was published in May 2007, but the reporting of this core set in trials and in observational studies is unknown. Other clinical domains considered important for OMERACT and the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) are enthesitis, dactylitis, spine, fatigue, nail disease, and physician global assessment (1, 3). However, instruments to assess these domains have not been agreed on and the frequency of their assessment is unknown.

In the last decade, outcome measures based on patients' opinions, i.e., patient-reported outcomes (PROs), have become critical outcomes in both clinical trials and long-term observational studies in rheumatic diseases (4–6). PROs, defined as any measure of a patient's health status that is elicited directly from the patient and assesses how the patient feels or functions with respect to his or her health condition, reflect the patient burden of disease and feeling of wellness more accurately (7). Furthermore, PROs should be assessed because a discrepancy has been found between patients' and physicians' opinions in several diseases (8–10). Instruments for measuring PROs are easy to administer and some, like the Health Assessment Questionnaire (HAQ), have been proven to be reliable, valid, and sensitive to change (11). There are several ways to explore PROs: some are qualitative (12), whereas others are quantitative such as by assessing the frequency of different outcome measures (4, 5).

The objective of the present work was to assess, through a literature review, which are the most frequently assessed domains and clinical outcomes in recent clinical trials addressing PsA. The specific research questions were: 1) is the OMERACT core set being reported? and 2) are other PROs being assessed and what are the instruments used?

Significance & Innovations

  • This work evidenced great heterogeneity in the assessment of psoriatic arthritis (PsA) in recent articles relating to PsA; 84 different outcomes were used.

  • The most frequently assessed outcomes were those proposed in the Outcome Measures in Rheumatology core set; however, there was no consensus regarding the number of joints to assess and instruments for dactylitis and enthesitis.

  • Patient-reported outcomes were assessed in more than 75% of publications; however, instruments used were heterogeneous and some domains, e.g., fatigue, emotional, and esthetic aspects, were rarely reported.

  • This study will serve as a reference for the assessment of PsA.

MATERIALS AND METHODS

This systematic literature review was conducted using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement as a guideline in the development of the study protocol and reporting of the results (13).

Search and selection process.

To obtain all recent published articles reporting clinical trials with any type of clinical measure in PsA, an extensive literature search was performed in the PubMed/Medline database on July 1, 2010. Publications were identified through a search that used the following medical subject headings (MeSH) term: “psoriatic arthritis” (MeSH) with a limitation to “Humans,” “all adults: 19+ years,” “English,” “published in the last 4 years,” i.e., July 2006 to July 2010, and “clinical trials.”

Inclusion criteria.

Clinical trials (14), i.e., interventional clinical trials (randomized or not) and observational clinical trials (cohort, cross-sectional, case–control), including PsA patients and reporting at least 1 clinical outcome were included. This systematic review was not restricted to trials evaluating drug therapy but also included trials evaluating devices, behavioral interventions, diagnostic procedures, etc., if they reported at least 1 clinical outcome.

Exclusion criteria.

Articles not reporting any clinical outcome measures (e.g., articles reporting only laboratory outcomes, radiographic scores, or genetic analysis examination) were excluded. Articles were also excluded if they did not concern PsA or if they were reviews or editorials, since we were interested in obtaining an overview of the use of clinical measures in original research articles. The selection process was performed by 2 authors (PEP and CG-V) based on the titles and abstracts of the articles, and then on full texts.

Clinical outcomes.

All of the clinical outcomes were collected separately by 2 authors (PEP and CG-V). Disagreements were solved by consensus. The Patient-Reported Outcome and Quality of life Instruments Database (ProQolid) was used to confirm the authors' compilation of a predefined list of PROs (15). The reviewers were not blinded to the journal name and the authors, as evidence concerning the effect of masking on assessments of trials is inconsistent (16). Clinical outcomes were defined here as: 1)part of the core set: elements of the core set are clinical outcomes related to joints, pain, skin, patient global assessment, physical function, and quality of life (Figure 1); 2) other domains considered important by the OMERACT group but not in the core set, i.e., enthesitis, dactylitis, spine, fatigue, nail disease, and physician global assessment (3); and 3) other PROs. PROs were defined as any measure of patients' health status that is elicited directly from the patient and assesses how the patient feels or functions with respect to this health condition. PROs were subdivided in the following domains of health by the authors: in the core set, pain, quality of life, physical function, skin, and global assessment; other PROs not in the core set include morning stiffness, fatigue, utility/productivity, coping/self-efficacy, composite scores, and other domains. Composite scores were defined as any measure of a patient's health status elicited directly from the patient and concerning more than 1 domain. For each outcome, the instrument (e.g., the HAQ) was collected and then the instrument was classified into 1 domain of health (e.g., physical function). Results are shown as the frequency of reported domains and the frequency of each clinical instrument in each main category.

Figure 1.

Domains for psoriatic arthritis assessment. MRI = magnetic resonance imaging; CT = computed tomography; US = ultrasound; PGA = physician global assessment. Reproduced, with permission, from ref.8.

General data extraction.

Data were obtained on year of publication, funding sources (public or private, either clearly reported or extrapolated from authors' affiliations), study design (RCT or other clinical trial and observational studies, i.e., case–control, cross-sectional, or retrospective), and number of patients. Demographic data such as sex, mean age, mean disease duration, treatments under evaluation, and maximum duration of followup were recorded for each report. The quality of the publications was determined by use of the Jadad scale (score 0–5), where a high score reflects high quality (17). The Jadad scale evaluates quality of randomization, blinding, and description of withdrawals and dropouts. The Jadad scale is often used to describe clinical trial quality but is more particularly adapted for use in RCTs.

Statistical analysis.

Analysis was mainly descriptive. Comparisons of frequencies of clinical outcomes were performed by the chi-square test or Fisher's exact test, according to the study design; comparisons of frequencies of reporting of core sets according to year of publication (before and after 2008) were also performed. Data analyses involved use of SAS, version 9.1.

RESULTS

Description of recent publications in PsA.

Of the 84 publications identified by the literature search, 58 were included (18–75) in the analysis (Figure 2 and Supplementary Table 1, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). Twenty-six articles were excluded, mainly because they did not report any clinical outcome (n = 17) or were not about PsA (n = 8). The characteristics of the publications and patients evaluated are given in Table 1.

Figure 2.

Flow chart showing the selections of recent publications that reported clinical outcomes in psoriatic arthritis (PsA). MeSH = medical subject headings.

Table 1. Characteristic features of the 58 publications assessing clinical outcomes in psoriatic arthritis*
 All articles (n = 58)RCTs (n = 17)Other studies (n = 41)
  • *

    RCTs = randomized clinical trials; DMARDs = disease-modifying antirheumatic drugs; NSAIDs = nonsteroidal antiinflammatory drugs.

  • Results were calculated on available data.

No. of patients per article, mean ± SD (range)214 ± 272 (7–1,122)213 ± 210 (7–752)214 ± 296 (7–1,122)
Total number of patients12,4053,6278,778
Female sex, no. (%)4,644 (42.0)1,398 (42.5)3,246 (41.8)
Age, mean ± SD years (range of means)48.2 ± 5.4 (26.5–63.0)47.1 ± 6.6 (26.5–60.0)48.6 ± 4.7 (36.0–63.0)
Disease duration, mean ± SD years (range of means)9.0 ± 3.1 (4.4–17.0)7.6 ± 1.7 (5.4–11.0)9.6 ± 3.4 (4.4–17.0)
Duration of followup, mean ± SD weeks (range of means)40.4 ± 33.1 (4.0–144.0)38.0 ± 26.2 (4.0–98.0)41.8 ± 36.8 (12.0–144.0)
Financial support, no. (%)   
 Public2 (3.4)0 (0.0)2 (4.8)
 Totally or partly private40 (69.1)14 (82.4)26 (63.5)
 Not available16 (27.5)3 (17.6)13 (31.7)
Treatment assessed, no. (%)   
 Biologics and/or DMARDs42 (72.4)14 (82.3)28 (68.2)
 NSAIDs1 (1.7)1 (5.8)0 (0.0)
 Steroids1 (1.7)1 (5.8)0 (0.0)
 Other pharmacologic treatment1 (1.7)0 (0.0)1 (2.4)
 Nonpharmacologic treatment2 (3.4)1 (5.8)1 (2.4)
 Study not assessing treatment11 (18.9)0 (0.0)11 (26.8)

Of the 58 articles included in the analysis (18–75), 17 (29%) were RCTs (18–34) and 41 (71%) were nonrandomized interventional clinical trials or observational clinical trials (case–control, cohort, or cross-sectional) (35–75).

Seventy-one percent of RCTs had a Jadad score of ≥3 (see Supplementary Table 1, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). The number of PsA patients evaluated was 12,405, with a mean ± SD age of 48.2 ± 5.4 years and a mean ± SD disease duration of 9.0 ± 3.1 years; 4,644 (42% of available data) were women. The mean ± SD duration of followup was 40.4 ± 33.1 weeks.

Clinical outcomes in recent publications of PsA.

Eighty-four instruments corresponding to 16 domains of health were reported in the 58 articles. A mean ± SD of 6.9 ± 4.3 instruments per article were described. Table 2 lists the health domains, detailed below, considered “core set” or important by OMERACT (3) and their frequency in the 58 analyzed articles, according to the study design.

Table 2. Health domains reported in 58 recent publications of psoriatic arthritis*
 All (n = 58)RCTs (n = 17)Other design (n = 41)P
  • *

    Values are the number (percentage). RCTs = randomized clinical trials.

  • P comparing RCTs to other design.

  • The core set domains and important domains were defined by the Outcome Measures in Rheumatology (3). The domains are shown by decreasing frequency of reporting.

6 core set domains    
 Skin32 (55.2)13 (76.5)19 (46.3)0.046
 Joint30 (51.7)9 (52.9)21 (51.2)1.000
 Physical function28 (48.3)12 (70.6)16 (39.0)0.043
 Pain27 (46.6)10 (58.8)17 (41.5)0.260
 Patient's global assessment23 (39.7)8 (47.1)15 (36.6)0.559
 Quality of life22 (37.9)8 (47.1)14 (34.2)0.387
 At least 1 core set domain53 (91.4)17 (100)36 (87.8)0.308
 All 6 core set domains6 (10.3)4 (23.5)2 (4.9)0.055
Important domains    
 Enthesitis16 (27.6)7 (41.2)9 (22.0)0.197
 Dactylitis15 (25.9)6 (35.3)9 (22.0)0.334
 Fatigue9 (15.5)3 (17.7)6 (14.6)1.000
 Spine3 (5.2)0 (0.0)3 (7.3)0.548
 Nail3 (5.2)1 (5.9)2 (4.9)1.000

Core set domains.

The core set domains were reported in 37.9–55.2% of articles. The most frequently reported domain was skin, evaluated in 32 (55.2%) of the 58 analyzed articles. It was evaluated mainly by the Psoriasis Area and Severity Index (n = 27 [84.3%]) of 32 articles) (76) and the physician global assessment of psoriasis (n = 8 [25.0%] of 32 articles). Joints, the second most reported domain, was assessed mainly by 68 tender joint count and 66 swollen joint count (n = 16 [53.3%] of 30 articles), followed by 28 tender joint count and 28 swollen joint count (n = 7 [23.3%] of 30 articles); other instruments used were 78 tender joint count and 76 swollen joint count (n = 4), number of joints clinically damaged (n = 1), presence of distal interphalangeal joint involvement (n = 1), and presence of symmetric polyarthritis (n = 1). Core set domains were more frequently reported in RCTs (Table 2). There was no difference in the reporting of core sets according to the year of publication (before and after 2008) (data not shown).

Forty-four publications (75.9%) reported at least 1 PRO. Patient-reported domains of health most frequently reported were function/disability, pain, patient's global assessment, and quality of life, which are all part of the OMERACT core set (3). Twenty-eight different PRO instruments were described in 11 domains of health (Table 3).

Table 3. Patient-reported outcomes: domains of health and instruments reported in 58 recent publications of psoriatic arthritis*
 Articles reporting the domain, no. (%)Instrument used to assess the domain of healthArticles reporting the instrument, no. (%)
  • *

    HAQ = Health Assessment Questionnaire; VAS = visual analog scale.

  • Some articles reported a given domain of health by using several instruments; therefore, the total may be >100%.

Core set domains   
 Function/disability28 (48.0)HAQ (11)22 (78.5)
  Modified HAQ (77)5 (17.8)
  Bath Ankylosing Spondylitis Functional Index (78)3 (10.7)
  Revised Leeds Disability Questionnaire (79)2 (7.1)
  Advanced Activities of Daily Living Scale (80)1 (3.5)
  Psoriasis Disability Index (81)1 (3.5)
 Pain27 (46.6)Pain VAS (82)27 (100)
 Patient's global assessment23 (39.6)Patient's global assessment of disease activity VAS (83)22 (95.6)
  Patient's global assessment of health1 (4.3)
 Quality of life22 (37.9)Short Form 36 Health Survey (84)16 (72.7)
  Dermatology Life Quality Index (85)7 (31.8)
  Short Form 12 Health Survey (86)1 (4.5)
  Psoriatic Quality of Life Toll (87)1 (4.5)
 Skin2 (3.4)Patient's skin assessment VAS2 (100)
Other domains   
 Fatigue9 (15.5)Fatigue VAS6 (66.7)
  Functional Assessment of Chronic Illness Therapy (88)3 (33.3)
 Composite scores6 (10.3)Bath Ankylosing Spondylitis Disease Activity Index (89)5 (83.3)
  EuroQol 5-domain (90)2 (33.3)
 Morning stiffness6 (10.3)Morning stiffness presence6 (100)
 Utility/productivity4 (6.9)Short Form 6 dimensions (91)3 (75.0)
  Daily productivity VAS1 (25.0)
 Other2 (3.4)Self-reported number of visits to physicians during the last 6 months2 (100)
  Involvement and satisfaction with care questionnaire (92)1 (50.0)
 Coping/self-efficacy1 (1.7)Rheumatoid Arthritis Self-Efficacy Scale (93)1 (100)
  Arthritis Self-Efficacy Scale (94)1 (100)
  Arthritis Helplessness Index Perceived Control (95)1 (100)
  Arthritis Stages of Change Questionnaire Self-Management (96)1 (100)
  Self-management behaviors in the previous week1 (100)

Domains not in the core set.

Among the other important domains according to OMERACT (3) but not in the core set (Table 2), the only domains frequently reported were enthesitis (n = 16 [27.6%]) and dactylitis (n = 15 [25.9%]). There was great variability in assessing these domains, with 12 and 6 different instruments used to assess enthesitis and dactylitis, respectively. The simple evaluation for the presence of enthesitis or entheseal pain was the most utilized instrument for assessing enthesitis (n = 6 [10.3%]), followed by the Maastricht Ankylosing Spondylitis Enthesitis Score (97) and the presence of enthesitis in the feet and heels, both reported in 5 articles (8.6%). Dactylitis was mainly assessed by the simple number of fingers and toes with dactylitis (n = 14 articles [24.1%]), followed by a scale of severity scoring from 0–3 for each digit of the hand and feet (n = 6 [10.3%]) (21) and the Infliximab Multinational Psoriatic Arthritis Controlled Trial Index for Dactylitis (n = 2 [3.4%]) (98). Nail disease was assessed in only 3 articles (5.1%), mainly by the Nail Psoriasis Severity Index (99).

Composite scores were assessed in 37 articles (63.8%) mainly by the following instruments: American College of Rheumatology response (reported in 26 articles [44.8%]) (100), the 28-joint Disease Activity Score based on erythrocyte sedimentation rate (n = 14 [24.1%]) (101), the Psoriatic Arthritis Response Criteria (PsARC) or modified PsARC (n = 14 [24.1%]) (102), and the European League Against Rheumatism response criteria (n = 6 [10.3%]) (103).

PROs reported in the publications but not in the core set.

Fatigue was reported in 9 articles (15.5%). It was mainly assessed by a fatigue visual analog scale (n = 6 [66.7%]). No articles reported emotional or esthetic aspects, except those assessed in composite measures of quality of life.

DISCUSSION

The present study evidenced great heterogeneity in the assessment of PsA in recently published clinical trials: 84 different instruments were used. Patients with PsA were mainly assessed using the 6 domains decided on by OMERACT (reported in 37.9–55.2% of articles). However, the core set was rarely completely reported since only 10.3% of the studies reported all 6 core domains. Physician-reported outcomes were heterogeneous and in particular there appeared to be no consensus regarding the number of joints to assess or instruments for dactylitis and enthesitis. PROs were assessed in more than 75% of the publications; however, the instruments used were heterogeneous and some domains, i.e., fatigue, emotional, and esthetic aspects, are rarely reported.

This study has strengths and weaknesses. It may not be exhaustive since the one database assessed was PubMed/Medline; however, this is the most important database of biomedical research articles covering more than 5,000 journals published in the US and more than 80 other countries. Only studies published in English were included; our objective was not to perform an exhaustive review but rather to raise awareness on issues around outcomes in PsA. Another potential weakness was the limitation of the analyses to articles published in the last 4 years; however, the reason for limiting the review to the last 4 years was to include articles published after the dissemination of the consensus on the core set of domains to be assessed in PsA by GRAPPA and OMERACT in 2007 (3). Including recent articles can be responsible for some differences in the number of articles found with the same PubMed search criteria, since there is a time lag between a manuscript's publication and its classification in PubMed. However, recent articles give an up-to-date view on outcomes used in PsA trials.

The classification of the instruments into domains of health was not always easy due to the diversity of scores and questionnaires; however, the data extraction was performed separately by 2 investigators and consensus was high (96.8%). In case of unusual instruments, the authors searched for the primary publication to allow classification into domains of health.

Recent publications indicate that obtaining consensus on outcome measures improves the quality of research (104). The heterogeneity of instruments for assessing disease activity evidenced in this work leads to difficulties in comparing studies, and may lead to the use of nonvalidated instruments. This heterogeneity was present in the assessment of both “objective” measures and of PROs.

For physician-reported outcomes, heterogeneity was mainly noticed for some domains, i.e., joints, enthesitis (both evaluated by 12 instruments), and dactylitis (7 instruments). Other domains had good consensus in terms of instruments, e.g., global assessment evaluated by one instrument, a visual analog scale.

The present work showed that the OMERACT core set is incompletely reported in recent studies. This result may represent the consequence of low dissemination of the OMERACT core set. A possible reason for this observation could be that there is necessarily a lag from the publication of recommendations until they can impact reported outcomes. This is particularly true for trials, as the design and execution take years and the reported outcomes cannot be changed after the trial has started. However, it should be noted that no differences were observed for trials published in 2006–2008 and those published in 2008–2010.

There may also be applicability issues due to a lack of consensus regarding instruments to assess each domain. Further work is ongoing to better define instruments within the GRAPPA and OMERACT groups.

In regard to PROs, heterogeneity has also been found in other rheumatic diseases such as RA, where more than 60 different PRO instruments were described (5). In fact, there is an extensive variety of PROs: the ProQolid (15), for example, includes 689 instruments, and the On-Line Guide to Quality of Life Assessment (105) includes thousands of instruments. Even with the large number of PROs available, some patient-reported domains, such as body image, are not being reported in recent publications. There are few studies in PsA aiming to know which domains are considered the most important in patients' opinions (12, 106). Fatigue is a patient-reported domain described in only 15.5% of PsA articles. However, this domain is considered an important issue in other rheumatic diseases; several qualitative studies (107, 108) have pointed out the importance of fatigue for patients with RA. In RA, fatigue is considered intrusive and overwhelming and has consequences on all aspects of quality of life (107, 108). In PsA, however, there is a lack of data regarding the importance of fatigue for patients.

In conclusion, although PsA patients are mainly assessed, in recent published studies, through the OMERACT core set domains, better consensus on instruments to assess each domain is needed. Furthermore, work is necessary to obtain a better insight into which domains of health are relevant in patients' opinions and which scores and/or questionnaires should be used for their assessment. Qualitative studies based on focus groups or individual interviews with PsA patients can be an alternative for enhancing our knowledge in this field, allowing a better assessment of the disease.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Palominos had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Palominos, Dougados, Gossec.

Acquisition of data. Palominos, Gaujoux-Viala, Gossec.

Analysis and interpretation of data. Palominos, Gaujoux-Viala, Fautrel, Dougados, Gossec.

Acknowledgements

We thank Simon Paternotte for statistical analysis.

Ancillary