Association of Patient Mental Health Status With the Level of Agreement Between Patient and Physician Ratings of Psoriasis Severity

Key Points

Question  Is a patient’s mental health status associated with discordance between patient and physician assessments of psoriasis severity?

Findings  In this cohort study of 502 patients with psoriasis, analysis of repeated cross-sectional data from a tertiary psoriasis service in London found discordance in patient and physician psoriasis severity ratings at 39% of the appointments. Twenty-six percent of the patients underestimated their psoriasis severity, and 13% of patients overestimated their psoriasis severity, compared with physician ratings.

Meaning  Although patient assessments of disease severity remain important for informing appropriate treatment decisions, the data from this study suggest that recognition of anxiety and depression among individuals with psoriasis is important for interpretation of the assessments.

Importance  The emerging paradigm of treat-to-target in psoriasis requires accurate monitoring of treatment response. The commonly used physician global assessment tool does not capture the patient’s perception of their disease. Patient assessments facilitate shared decision-making and foster patient-centered care; however, recent research reports a discordance between patient- and physician-reported psoriasis severity. Understanding the factors underlying this discordance may improve treatment satisfaction and disease outcomes.

Objectives  To evaluate the discordance between patient- and physician-reported measures of psoriasis severity and assess the association with patient mental health status.

Design, Setting, and Participants  A cohort study using repeated cross-sectional analysis of real-world longitudinal data was conducted at a large specialist psoriasis service serving London and Southeast England. A total of 502 patients attending the psoriasis service between May 12, 2016, and November 1, 2018, were included. Data analysis was conducted July 22 to October 22, 2019.

Main Outcomes and Measures  Psoriasis severity was assessed on each visit with identical 5-point physician and patient global assessment scales (clear/nearly clear, mild, moderate, severe, and very severe). Each patient completed validated self-report screens for depression and anxiety on each visit.

Results  Longitudinal data from 502 individuals with psoriasis (1985 total observations) were available. A total of 339 patients (68%) were men, 396 (79%) were White, mean (SD) age was 47 (13) years, and 197 patients (39%) had concurrent psoriatic arthritis, 43 (9%) screened positive for depression, and 49 (10%) screened positive for anxiety. There was discordance between physician and patient measures of disease severity in 768 of 1985 office appointments (39%); on 511 visits (26%) patients rated their psoriasis as less severe and on 257 visits (13%) patients rated their psoriasis as more severe compared with their physician. Individuals who screened positive for depression or anxiety were more likely to overestimate their psoriasis severity compared with their physician (relative risk ratio: depression, 2.7; 95% CI, 1.6-4.5; anxiety, 2.1; 95% CI, 1.3-3.4). These findings remained statistically significant after adjustment for age, ethnicity, sex, body mass index, smoking, number of comorbidities, treatment modality, and presence of psoriatic arthritis.

Conclusions and Relevance  The findings of this cohort study suggest that discordance between patient and physician assessments of psoriasis severity is associated with patients’ mental health status. Recognition of anxiety and depression in individuals with psoriasis appears to be important when interpreting patient-reported outcome measures and informing appropriate treatment decisions.

Psoriasis is a common inflammatory skin condition affecting 2% to 4% of the global population.¹ Psoriasis is associated with reduced quality of life and substantial morbidity, including psoriatic arthritis, cardiovascular disease, and obesity.² The management of psoriasis has been improved by the availability of biologic drugs, such that skin clearance is now a realistic treatment goal.³ This advance has opened avenues for treat-to-target approaches, whereby target disease activity end points are used to support the introduction and modification of treatments, with the ultimate aim of improving clinical and cost-effectiveness outcomes.⁴ Because the introduction and continuation of biologic treatments is informed by clinical assessments of disease severity,^5-7 ensuring current measures are accurate and relevant is vital.

Physician Global Assessment (GA) and Psoriasis Area Severity Index (PASI) are physician-rated disease activity measures commonly used in real-world practice and clinical trials and recommended in treatment guidelines.^5-7 Although data are limited in psoriasis,^8-10 studies have identified discordance between physician and patient measures of disease severity in chronic inflammatory diseases, for example, among one-third of patients with rheumatoid arthritis.^11-13 Because this discordance is associated with patient dissatisfaction, lower treatment adherence, and poorer disease outcomes,^11-13 an improved understanding of the underlying factors is important.

Evidence suggests that pain and fatigue lead patients with arthritis to overestimate their disease activity relative to their physicians.^14-18 Patient-physician discordance in ratings of psoriatic and rheumatoid arthritis has also been associated with poor mental health.^14,16,17 This latter finding warrants further investigation in psoriasis owing to the high prevalence of depression and anxiety, particularly among individuals with worse psoriasis.^17,19 A previous pilot study also showed that psychological interventions, such as mindfulness-based cognitive therapy used as an adjunct to usual psoriasis therapy, may improve self-assessed psoriasis severity.²⁰

This real-world repeated cross-sectional study therefore aimed to (1) evaluate the extent of discordance between physician- and patient-reported measures of psoriasis severity, (2) examine whether discordance is associated with adverse mental health status, and (3) explore the interaction between adverse mental health status and psoriasis disease severity.

All participants were recruited during dermatology outpatient visits between May 12, 2016, and November 1, 2018, in a large specialist psoriasis center serving London and South East England. Longitudinal clinical data were combined with information collected from the same patients as part of the Integrating Mental and Physical Healthcare: Research Training and Services (IMPARTS) screening program.²¹ Participants first completed self-report questionnaires via a tablet device including a patient GA and questions on depression and anxiety. They subsequently underwent clinical assessments by their physician as part of their routine care. Participants attended multiple appointments (between 1 and 8) during the study.

Because IMPARTS is a clinical initiative used in routine care, participants did not require formal consent to take part in the study. Everyone was informed that their anonymized data might be used for research purposes and could opt out at any time. Ethical approval for the present study was given by the IMPARTS Research Ethics Committee (REC reference 11/H0802-7), and ethical approval for IMPARTS was given by NHS Research Ethics Committee. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

In measurement of discordance, psoriasis severity was quantified using identical physician and patient GA scales. This 5-point scale (1, clear/nearly clear; 2, mild; 3, moderate; 4, severe; and 5, very severe) had no specific descriptors relating to redness, thickness, or scaling.^22,23 We elected to use the physician’s assessment as the standard against which to evaluate discordance because it is the standard measure used in clinical guidelines.^5,7 Psoriasis severity discordance was measured by subtracting the patient assessment from the physician assessment and categorizing the ratings as overestimate (the patient reported a greater level of psoriasis severity compared with the physician by at least 1 point on the scale); neutral (the patient and physician reported the same category of psoriasis severity); or underestimate (the patient reported a lower level of psoriasis severity compared with the physician by at least 1 point on the scale). We chose a difference of 1 point following evidence, albeit using varied GA measures showing that differences of 1 point correspond to meaningful differences in physician-assessed outcome measures, such as the PASI.^4,23 Although a difference of 2 or more points resulted in smaller groups of participants in each discordant category, we also assessed this more stringent threshold in a sensitivity analysis. In addition, psoriasis severity was evaluated with the PASI, a clinician-rated measure of lesion area coverage and appearance.²²

Screening for probable major depressive disorder (herein, depression) was conducted with the Patient Health Questionnaire-2 (PHQ-2)²⁴ and PHQ-9.²⁵ Screening for probable generalized anxiety disorder (GAD) (herein, anxiety) was screened with the Generalized Anxiety Disorder scale (GAD-7).²⁶ Details are provided in the eMethods in the Supplement.

Covariates were recorded at each appointment. Demographic information included age, sex, and race/ethnicity. Race/ethnicity was coded into 6 categories (White; Asian or Asian British; Black or Black British; Chinese, Japanese, Korean, or Indochinese; mixed; and other). We further adjusted for clinical factors previously shown to influence treatment outcomes. These included smoking status (current smoker or nonsmoker),²⁷ body mass index (BMI),²⁸ psoriatic arthritis diagnosis (yes or no),²⁹ number of comorbid conditions (0, 1-2, 3-4, and ≥5),³⁰ current treatment (none, topical only, nonbiologic systemic, and biologic systemic).³¹

Psoriasis disease severity (patient GA and physician GA) was summarized as counts and percentages across all appointments over follow-up. Demographic and clinical covariates were summarized for each participant’s first appointment. Discordance was summarized (counts and percentages) for the overall sample, as well as for subsamples defined by the presence of depression and anxiety. The weighted κ value was calculated.³²

Multivariable multinomial logistic regression models were used to assess the association between mental health status and patient-physician psoriasis severity discordance. The outcome was the 3-category measure of the defined discordance (overestimate, neutral, or underestimate), with neutral treated as the base category. Depression and anxiety, as defined by PHQ-9 and GAD-7 screening tools, were entered as binary variables in separate models, given the substantial overlap between these conditions. To investigate whether mental health was more strongly associated with discordance among patients with more severe clinician-assessed psoriasis, we also included a term representing the interaction between probable depression and the continuous PASI score. The statistical significance of the interaction effect (main effects and interaction term) was tested with a 2-sided Wald test.

All models were adjusted for age, sex, race/ethnicity, smoking status, BMI, psoriatic arthritis, current treatment, and number of comorbidities. A sandwich estimator was used to adjust SEs to take into account clustering owing to repeated appointments per participant.³³ Estimates are reported as relative risk ratios with associated 95% CIs and predicted probabilities estimated with other covariates set to their sample mean values. Findings were considered significant at P ≤ .05. Models were estimated in Stata, version 15 (StataCorp).³⁴

After excluding those with missing information on patient GA or physician GA (n = 65) or covariates (n = 5), the analytical sample comprised 502 participants. Compared with those in the analytical sample, excluded participants tended to be younger (mean age, 42.6 vs 46.9 years; P = .02) and had a higher prevalence of probable depression (16% vs 9%; P = .09) and probable anxiety (18% vs 10%; P = .09) but were similar in terms of sex (32% [included] vs 42% [excluded] female; P = .20).

Table 1 summarizes the analytical sample. Participants included 339 men (68%) and 163 women (32%); most were of White race (396 [79%]), nonsmokers (393 [78%]), and overweight (mean [SD] BMI, 29.4 [6.0], calculated as weight in kilograms divided by height in meters squared). The mean (SD) age was 46.8 (13.1) years. Over a third of participants (197 [39%]) had concurrent psoriatic arthritis, and most were receiving systemic treatment (nonbiologic, 257 [51%]; biologic, 170 [34%]). A total of 221 participants (44%) reported having at least 1 comorbid condition (the most prevalent comorbidities were hypertension, autoimmune disorders, and liver disease).

Overall, patient and physician severity assessments tended to be in agreement (Figure 1). At 893 appointments in which the patient rated their psoriasis severity as clear/nearly clear, 76.2% were also assessed by their physician as clear/nearly clear. The weighted κ statistic was 0.58 (95% CI, 0.56-0.60) indicating moderate agreement. Regarding the 3-category measure of discordance, 1217 appointments (61%) resulted in agreement between patient and physician about symptom severity (neutral) (Table 2) and 768 appointments (39%) resulted in disagreements. Where disagreements occurred, patients were more likely to underestimate their psoriasis severity, compared with the physician rating (511 [26%] appointments) rather than overestimate the severity (257 [13%] appointments).

Compared with physician ratings, patients screening positive for depression or anxiety were more likely to overestimate their psoriasis severity (depression, 33 of 129 appointments [26%]; anxiety, 36 of 155 appointments [23%]) than were patients without depression or anxiety (depression, 224 of 1856 appointments [12%]; anxiety, 221 of 1830 appointments [13%]). To investigate the risk of discordance attributable to mental health, we used multivariable multinomial logistic regression analysis. After adjustment for covariates, depression and anxiety were separately associated with an increased risk of patients overestimating their disease severity, compared with their physician, vs those without depression or anxiety (Figure 2) (relative risk ratio for depression, 2.7; 95% CI, 1.6-4.5; for anxiety, 2.1; 95% CI, 1.3-3.4).

We investigated the interaction between mental health status and disease severity on patients’ probability of overestimating or underestimating severity, compared with the physician’s GA (Figure 3; eTable 1 in the Supplement). We found that patients with milder psoriasis (PASI score ≤5) and concurrent depression or anxiety were more likely to overestimate their psoriasis severity. Conversely, patients with moderate to severe psoriasis (PASI score >10) and concurrent depression or anxiety were more likely to underestimate their psoriasis severity. This interaction was statistically significant based on a Wald test (χ²₂ for depression, 8.8; P < .001; χ²₂ for anxiety, 37.0; P < .001). Thus, psoriasis severity may moderate the association of mental health status with patient-physician discordance.

Our findings were consistent when using a more stringent threshold for discordance (a difference of 2 categories between patient and physician GAs). The proportion of patients overestimating or underestimating severity was lower, but the patterns of association with mental health were preserved (eTable 2 and eFigure in the Supplement).

To our knowledge, this represents the first study in psoriasis to characterize the real-world association of both depression and anxiety assessed using validated screening tools with discordance between patient and physician psoriasis severity ratings. These data suggest a substantial (39%) discordance between patients and physicians in their psoriasis severity assessments and highlight patient mental health status as a significant contributor to rating discordance. Patients with depression (major depressive disorder) and anxiety (GAD) were twice as likely to have overestimated their disease severity assessments vs physician rating, compared with patients without depression or anxiety after adjustment for potential demographic and clinical confounders.

These data have important clinical relevance. Shared decision-making between patients and their physicians is necessary for safe and effective care of psoriasis, which often requires lifelong treatment. Decisions about when to start or switch psoriasis therapies in routine practice are currently informed by physician-rated disease severity assessments, such as the Physician Global Assessment and PASI measures.^4-6 However, data from evaluations of other chronic inflammatory diseases have underscored the importance of considering the patient’s perspective through the use of patient-reported measures; agreement between patient and physician assessments has been reported to improve both treatment adherence and outcomes.⁸ Our data highlight an unmet need to align physician and patient perceptions of disease severity, which must be addressed to reach a consensus about optimal treatment strategy.³⁵

Because the prevalence of psoriasis continues to increase in the face of limited health service resources,³⁶ virtual clinical review is increasingly important for the efficient and sustainable delivery of health care. In particular, the implementation of virtual consultations has been rapidly accelerated in recent months in response to the coronavirus disease 2019 pandemic. Virtual consults rely on the accurate interpretation of patient-reported disease severity measures. Our data suggest that coexistent depression and anxiety may lead to the overestimation of psoriasis severity by patients compared with their physician, particularly at the less severe end of the disease severity spectrum. If not considered in the correct context, the patient GA thus has the potential to influence clinical decision-making inappropriately. This possibility underscores the importance of concurrent screening for anxiety and depression to ensure that treatment plans accurately address disease activity in addition to mental health comorbidity. A multidisciplinary approach to psoriasis care is thus necessary.

There is a paucity of real-world data on physician-patient discordance of disease severity ratings in psoriasis and the underlying factors contributing to this discordance,¹⁴ so our study provides evidence to inform clinical practice. Our rate of discordance (39%) is similar to that reported by Griffiths et al⁸ from their multinational surveys of 524 dermatologists and 3821 patients with psoriasis. Their retrospective study found that 45% of physician-patient pairs were in disagreement about psoriasis severity (assessed using a 5-point scale: clear, almost clear, mild, moderate, and severe) but did not examine reasons for the discordance. Two studies in trial settings similarly highlighted discordance between patient and physician assessments of psoriasis, but again did not explore the factors contributing to the discrepancy.^10,11 There are also reports of discordance between physicians and patients regarding their satisfaction with the level of psoriasis control achieved. Two surveys in the US showed patient-physician discordance in satisfaction with psoriasis control among 18% of patients, and this discordance was associated with higher disease severity and reduced quality of life.²

Although limited data exist, the influence of mental health on patient-physician discordance has mostly been explored in individuals with rheumatic disease.^11,15-18 Barton et al¹⁶ reported an association between PHQ-9 depression score and patient-physician disease severity assessment discordance among 223 patients with rheumatoid arthritis. Discordance was found in 36% (compared with 39% in our psoriasis study) and patients overestimated their disease severity compared with physicians’ assessments in 85% of these cases. Consistent with our findings, Barton et al found depressive symptoms (a 5-point increase in PHQ-9 score) to be the strongest predictor of a patient overestimating their psoriasis severity compared with their physician (odds ratio, 1.61; 95% CI, 1.02-2.55).

Eder et al¹⁴ analyzed data from 331 individuals attending a psoriatic arthritis clinic and found lower patient-physician (rheumatologist) discordance (15%) for skin assessments compared with our study. However, the scores of patient and physician GAs ranged from 0 to 10 on a numeric rating scale, and a difference of more than 2 points was considered a clinically relevant discordance. Discordance was attributed to increased pain and levels on the Dermatology Life Quality Index, which accounted for 17% (pain) and 14% (quality of life) of the variation in patient-physician discordance. Unlike our findings, those of Eder et al did not identify an important role for mental health in patient-physician discordance of disease severity ratings. However, they used a simple mental component score derived from the Short Form-36 Mental Component Summary health survey as a surrogate for general psychological distress, which accounted for just 1% of variation in patient-physician discordance. A need to improve physician detection and management of psychological distress in patients with psoriasis has also been highlighted.³⁷

To our knowledge, this represents the largest study to date of patient-physician discordance in psoriasis severity ratings, drawing on real-world clinical data. We were able to adjust for important demographic and clinical confounders and used validated screening tools for depression and anxiety (GAD-7 and PHQ-9), improving on previous studies in which mental health was assessed using the Short Form-36 Mental Component Summary.

In terms of limitations, these data were collected from a single specialist psoriasis clinic comprising patients with moderate to severe psoriasis and so may not be generalizable. The repeated cross-sectional analysis was unable to address questions of causation. There was no information on the stage or duration of treatment, which may help to contextualize the severity of psoriasis (eg, patients experiencing failure of a long-term treatment may perceive their disease severity differently than those starting treatment). Owing to the small sample size, discordance was defined as a difference of 1 category on the patient and physician GAs; however, a difference of 2 categories identified similar findings in sensitivity analyses. Although evidence showed differences of 1 category to be clinically meaningful,⁴ further work is required to understand what constitutes a clinically significant level of discordance. There are also no current standard definitions for physician and patient GAs; there is a paucity of research on patient-reported measures, and 5-, 6-, or 7-point Likert scales have variously been used for physician GAs in earlier studies.²²

Depression and anxiety were assessed using the PHQ-9 and GAD-7 screening tools rather than via clinical interview. However, these screening tools have been shown to be valid and reliable,^25,26,38,39 with sensitivity of 89% and specificity of 76% for the PHQ-2 in primary care samples.⁴⁰ The percentage screening positive for depression (9%) was lower than the 22% reporting a history of depression (ie, current or prior diagnosis) in BADBIR (the UK pharmacovigilance psoriasis registry⁴¹) but consistent with the UK General Practice Research Database (13% of those with severe psoriasis reported a history of depression⁴²), the PsoBest registry in Germany (7%⁴³), and the global PSOLAR study (15%⁴⁴). We were unable to measure pain or fatigue in our sample, which have previously been shown to be predictors of discordance.¹⁴ We were also unable to consider characteristics of the assessing dermatologist, such as age, sex, or level of experience, which may have influenced the extent of discordance.

Taken together, our results suggest that discordance between patient and physician measures of psoriasis severity is associated with patient mental health status. Patient-reported disease severity scores should thus be interpreted in the context of comorbid anxiety and depression, which can be facilitated by the routine use of mental health screening tools. The recognition, monitoring, and management of depression and anxiety in psoriasis by multidisciplinary health care teams (preferably encompassing clinical psychology expertise) has the potential to alleviate the substantial mental health burden in psoriasis while aligning clinician and patient perceptions of disease and treatment goals.

Accepted for Publication: December 28, 2020.

Published Online: March 3, 2021. doi:10.1001/jamadermatol.2020.5844

Corresponding Author: Ewan Carr, PhD, Department of Biostatistics and Health Informatics, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, 16 De Crespigny Park, London SE5 8AF, United Kingdom (ewan.carr@kcl.ac.uk).

Author Contributions: Dr Carr 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. Drs Carr and Mahil contributed equally to the study.

Concept and design: Carr, Mahil, Brailean, Barker, Smith.

Acquisition, analysis, or interpretation of data: Carr, Mahil, Brailean, Dasandi, Pink, Rayner, Turner, Goldsmith, Smith.

Drafting of the manuscript: Carr, Mahil, Brailean, Dasandi, Goldsmith.

Critical revision of the manuscript for important intellectual content: Carr, Mahil, Brailean, Pink, Barker, Rayner, Turner, Smith.

Statistical analysis: Carr, Brailean, Goldsmith.

Administrative, technical, or material support: Carr, Brailean, Dasandi, Pink, Barker, Rayner.

Supervision: Mahil, Barker, Goldsmith, Smith.

Conflict of Interest Disclosures: Dr Mahil reported receiving grants for St John's DermAcademy from AbbVie, Celgene, Eli Lilly, Janssen-Cilag, Novartis, Sanofi, and UCB outside the submitted work. Dr Pink reported receiving personal fees from Sanofi as an advisor and speaker; personal fees from AbbVie as an advisor and speaker; personal fees from Lilly as an advisor and speaker; personal fees from Leo Pharma as an advisor, speaker, and educational support; personal fees from Novartis as an advisor, speaker, and educational support; personal fees from Amgen as an advisor; personal fees from Boehringer as an advisor; personal fees from Almirall as an advisor and speaker; personal fees from Janssen as a lecturer; personal fees from La-Roche Posay as a lecturer; and personal fees from UCB as an advisor and speaker outside the submitted work. Dr Barker reported receiving personal fees from Amgen, Almirall, AbbVie, Janssen, Lilly, Novartis, Samsung, Boehringer-Ingelheim, Sun Pharma, and Bristol Myers Squibb outside the submitted work. Dr Goldsmith reported receiving grants from the National Institute for Health Research (NIHR), National Institutes of Health, Juvenile Diabetes Research Foundation, Stroke Association, and UK Research Innovation outside the submitted work. Dr Smith has received departmental research funding from AbbVie, GSK, Pfizer, Novartis, Regeneron, and Roche, and is coinvestigator on consortia with industry partners (see https://biomap-imi.eu/ and www.psort.org.uk). No other disclosures were reported.

Funding/Support: This article represents independent research part funded by the NIHR Biomedical Research Centre at South London and Maudsley National Health Service (NHS) Foundation Trust, King’s College London, and Guy's and St Thomas' NHS Foundation Trust. Dr Mahil is funded by a Medical Research Council Clinical Academic Research Partnership award (MR/T02383X/1).

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The views expressed herein are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care.