Work absenteeism and disability associated with psoriasis and psoriatic arthritis in the USA—a retrospective study of claims data from 2009 TO 2020

Psoriasis is a chronic inflammatory skin disorder that is characterized by scaly erythematous skin lesions, resulting in itching, irritation, and stinging for most patients [1]. Psoriasis can profoundly affect patient quality of life (QoL) and can follow a relapsing course with patients fluctuating between periods of remission and inflammation. According to data from the National Health and Nutrition Examination Survey 2010, the prevalence of psoriasis in US adults was estimated to be 3.2% (~ 7.2 million cases) [2]. It can affect all ages and prevalence rates between male and female individuals are similar. Family history as well as environmental factors, such as smoking, infections, stress, and certain medications can contribute to disease susceptibility [3, 4]. Psoriasis is also considered a multi-system disease with significant comorbidities; individuals with psoriasis are at increased risk of developing cerebrovascular disease, cardiovascular disorders, diabetes, metabolic syndrome, and depression [1, 5‐7].

Psoriatic arthritis (PsA) occurs in up to 30–33% of patients with psoriasis, though it is often underdiagnosed and undertreated [8‐10]. PsA is a complex, heterogenous immune-mediated disease characterized by musculoskeletal features, namely, peripheral arthritis, axial disease, enthesitis, and dactylitis. While most patients with PsA have concurrent or prior history of psoriasis, rheumatological manifestations precede the onset of skin lesions in about 20% of patients [11]. During this time, these patients have only a PsA diagnosis (sine psoriasis). PsA can be an aggressive, debilitating disorder that may lead to impaired function, reduced QoL, and irreversible joint damage. The cause of PsA is unknown, but genetic factors as well as environmental factors, such as smoking, stress, and trauma play a role in the susceptibility of developing PsA [12, 13]. In patients with psoriasis, studies have shown that the severity and location of psoriasis, as well as early age of psoriasis onset can also be risk factors for developing PsA [14]. Other risk factors for PsA have been described in patients with psoriasis, such as obesity, infections or genetic history [15]. In addition, the literature describes that patients with PsA have a higher risk of cardiovascular disease [16].

Work disability constitutes a spectrum of states, usually caused by chronic illness, that ranges from reduced productivity while remaining at work, absence from work as sick days (absenteeism), extended work absence with reduced salary or periods of unemployment with receipt of disability benefit (disability leave), to complete cessation of paid employment. Such work disability contributes substantially to the economic burden of psoriasis and PsA. According to a study by the National Psoriasis Foundation (NPF), 49% of people with psoriasis and/or PsA miss work regularly due to illness [7]. Generally, it is estimated that around 20–40% of the total costs of psoriasis are attributable to productivity and work loss [17‐22]. Among patients with PsA, it has been found that work disability rates range from 23 to 39% [23, 24]. Additionally, a study that used an employment disadvantages questionnaire to compare patients with psoriasis and PsA found an increased rate of work problems caused by PsA (32% versus 11% for psoriasis; p < 0.0005) [25]. Similarly, productivity loss has been shown to be greater among patients with PsA compared with those with other inflammatory diseases such as psoriasis, ankylosing spondylitis, rheumatoid arthritis [22, 26].

To date, previous studies that have analyzed work absence (absenteeism and short-term disability leave) have typically looked only at patients with psoriasis or PsA, rather than comparing the two [17, 18, 21, 27‐30]. Fowler et al. conducted a matched analysis between patients with and without psoriasis and found that absenteeism accounts for 40% of the economic burden among patients with psoriasis [21]. When comparing the rates of absenteeism among differing inflammatory diseases, absenteeism was most common among patients with PsA (20.3% versus 14.9%, 8.4% and 10.8% for psoriasis, RA, and AS, respectively) [26]. Data from the NPF survey found that patients with severe psoriasis have 1.8 times greater odds of being unemployed compared with patients with mild psoriasis [7]. Increased severity of psoriasis is also linked to higher costs [19].

Thus far, studies that have analyzed absenteeism are largely based on hospital data or clinical registries [17, 19, 27, 29]. However, there is an unmet need for an up-to-date comparative analysis of absenteeism and short-term disability leave between patients with psoriasis, PsA, and patients without psoriasis and PsA. We used a large administrative claims database to obtain robust estimates among US patients with the most recent data on treatments and standard-of-care.

In this study, evaluations of work absenteeism and short-term disability showed consistently greater associations with PsA and psoriasis than the control group. While work absenteeism outcomes revealed fewer differences between PsA and psoriasis, short-term disability outcomes (the proportion of patients with short-term disability, mean number of days absent, and associated costs) were greater among patients with PsA than patients with psoriasis.

The greater short-term disability incurred by patients with PsA may be linked to greater disease burden and therefore the need for longer work leaves. However, the exact underlying reason for which patients with PsA have greater short-term disability but similar work absenteeism as psoriasis patients cannot be concluded due to the lack of sufficient clinical details from the databases. Little is known from the literature to address this question. Further research is warranted to better understand what is driving work absences and short-term disability leaves among these patient populations.

In a previous study by Fowler et al., analyzing healthcare costs and patient work loss using data from a US-based insurance claims database, mean costs associated with disability per patient per month (PPPM) were $185 for patients with psoriasis and $69 for the control group, and mean costs associated with sick leave PPPM were $43 for patients with psoriasis and $31 for the control group [21]. When extrapolated to yearly values, the corresponding Fowler et al. mean disability costs (PPPY) are considerably higher compared with our findings: $2220 versus $406 for patients with psoriasis and $828 versus $335 for the control group, respectively. However, the mean sick leave costs (PPPY) were lower: $516 versus $941 and $372 versus $974, respectively. The disparities in the disability costs may be due to the assessment of just short-term disability in our study, whereas the Fowler et al. study looked at both short-term and long-term disability claims. Disparities in the sick leave costs may be due to the variations in the calculations: our study used direct sick leave data from employer payroll systems instead of computed sick leaves based on the number of outpatient physician visits (assumed 4 h of sick leave) and inpatient visits (assumed 8 h of sick leave). Furthermore, our study scaled down costs related to short-term disability by 0.6 (typical proportion paid by employers during short-term disability), which was not done in Fowler et al. Finally, our study separately calculated costs related to short-term disability and work absences, whereas Fowler et al. reported a sum of costs. As some patients have data and are eligible for both work absences and short-term disability, the total costs may be underestimated for these patients.

Furthermore, in another previous US-based population study using data from the MarketScan Databases, Feldman et al. found that for patients with psoriasis and PsA, the average adjusted indirect cost from short-term disability was estimated at $1347 PPPY [34]. In contrast, in the present study, the annual short-term disability costs ranged from $406 for patients with psoriasis to $605 for patients with PsA during the 5-year follow-up period. Like the present study, Feldman et al. also assumed an 8-h workday but applied the average hourly wage directly rather than multiplying it by a factor of 0.6 (the average salary proportion paid during a short-term disability leave). This helps explain why greater costs were obtained.

Other arthritic/rheumatic diseases have shown to have a major impact on absenteeism as well. In a retrospective, US-based insurance claims database study of patients with rheumatoid arthritis, short-term disability was associated with adjusted mean annual costs of $466 and corresponding mean sick leave costs were $470 [40]—the latter costs, specifically, are lower than that seen in the present study for both patients with psoriasis and PsA. In a meta-analysis of ankylosing spondylitis studies, annual sick leave costs for patients treated with biologics ranged from €913 to €2336 (approximately equivalent to $1025–$2625) [41]. The greater costs are likely due to the more severe populations, as only patients receiving biologics were considered.

The results of the present study reveal differences in costs associated with work absences among patients with psoriasis and PsA, and sensitivity analyses showed that these findings are robust to the assumptions made.

The study has several limitations. First, biases may arise as a result of the nature of the claims database. While propensity score matching helps reduce bias, certain underlying confounding factors could not be accounted for. For example, disease duration and severity, alcohol, and smoking were not able to be adjusted for given the variables available in the MarketScan Databases [33], which may lead to potential bias or model misspecification. Misclassification may have also occurred given the assumption that patients with psoriasis do not have PsA (as they have no diagnoses in their claims records). The MarketScan HPM Database is a small subset of the Commercial Claims and Encounters Database; male patients and certain employers are more highly represented. In addition, as data were limited to commercially insured patients these findings may not be generalizable to patients who are insured on other plans, uninsured patients, or populations outside of the USA. Furthermore, while this study did not incorporate higher weights associated with more severe comorbidities, matching on the CCI as a sensitivity analysis demonstrated that this assumption did not have a substantial impact on the results. In addition, the control group was matched to patients with psoriasis or PsA, rather than matching separately for each pairwise comparison. As such, the control group more closely resembled the psoriasis group given the greater number of patients with psoriasis than PsA. However, this was the most appropriate approach given the study objectives while limiting the number of matched cohorts. In terms of outcome measurements, long-term disability could not be evaluated due to an insufficient sample size, as less than 0.5% of patients had a long-term disability claim during follow-up. The sample size for absentee-eligible patients with PsA was also small (n = 616), which could introduce bias from the higher sampling variability.

In summary, psoriasis and PsA are associated with significantly higher absenteeism and short-term disability, as compared with individuals without psoriasis and PsA. Short-term disability and associated costs are also higher among patients with PsA than patients with psoriasis. These results contribute to our understanding of the economic burden of work loss in employed Americans with psoriasis and PsA. Future studies linking treatment strategies and work absenteeism are needed.