Abstract
Aims: To examine the comorbidity and economic burden among moderate-to-severe psoriasis (PsO) and/or psoriatic arthritis (PsA) patients in the US Department of Defense (DoD) population.
Materials and methods: This retrospective cohort claims analysis was conducted using DoD data from November 2010 to October 2015. Adult patients with ≥2 diagnoses of PsO and/or PsA (cases) were identified, and the first diagnosis date from November 2011 to October 2014 was defined as the index date. Patients were considered moderate-to-severe if they had ≥1 non-topical systemic therapy or phototherapy during the 12 months pre- or 1 month post-index date. Patients without a PsO/PsA diagnosis during the study period (controls) were matched to cases on a 10:1 ratio based on age, sex, region, and index year; the index date was randomly selected. One-to-one propensity score matching (PSM) was conducted to compare study outcomes in the first year post-index date, including healthcare resource utilization (HRU), costs, and comorbidity incidence.
Results: A total of 7,249 cases and 72,490 controls were identified. The mean age was 48.1 years. After PSM, comorbidity incidence was higher among cases, namely dyslipidemia (18.3% vs 13.5%, p < .001), hypertension (13.8% vs 8.7%, p < .001), and obesity (8.8% vs 6.1%, p < .001). Case patients had significantly higher HRU and costs, including inpatient ($2,196 vs $1,642; p < .0016), ambulatory ($8,804 vs 4,642; p < .001), emergency room ($432 vs $350; p < .001), pharmacy ($6,878 vs $1,160; p < .001), and total healthcare costs ($18,311 vs $7,795; p < .001).
Limitations: Claims data are collected for payment purposes; therefore, such data may have limitations for clinical research.
Conclusions: During follow-up, DoD patients with moderate-to-severe PsO and/or PsA experienced significantly higher HRU, cost, and comorbidity burden.
Introduction
Psoriasis (PsO) is a chronic, immune-mediated, and hyper-proliferative dermatological condition that causes red, scaly skin, and may progress and relapse over timeCitation1. It is one of the most prevalent skin conditions in the US, affecting ∼2% of the populationCitation2, and nearly 6.7 million adults have been diagnosed with PsOCitation3. Among PsO patients, ∼30% may develop psoriatic arthritis (PsA), which is characterized by morning stiffness, constant joint pain, and enthesitisCitation4. PsO and PsA share a common pathological mechanism, and both have genetic associations with multiple diseases based on chronic inflammationCitation5.
PsO and PsA are associated with several comorbid conditions such as cancer, cardiovascular disease, depression, diabetes, metabolic syndrome, obesity, Crohn’s disease, and liver disease, according to the National Psoriasis FoundationCitation6. Patients with severe PsO had a higher risk of these comorbidities and a shortened life expectancyCitation7.
PsO and PsA require long-term management and incur cumulative expenses; comorbid conditions will complicate management and increase economic burden. A systematic review showed that total US direct costs of PsO ranged from $51.7 billion to $63.2 billion, the indirect costs were between $23.9 billion and $35.4 billion, and medical comorbidities contributed $36.4 billion annuallyCitation8. Another study examined moderate-to-severe plaque PsO and reported mean total direct costs per patient to be $11,291 and total indirect costs per patient to be $2,101Citation9.
In order to provide a foundation of research and advocacy for policy decisions, it is important to understand the burden on PsO/PsA patients. However, there has been a dearth of recent data for the moderate-to-severe PsO/PsA population’s comorbid and economic burden. Therefore, this study aimed to investigate the economic burden of moderate-to-severe PsO and/or PsA, as well as the epidemiologic patterns of moderate-to-severe PsO comorbidities in a real-world US population.
Methods
Study design and data source
This was a retrospective cohort claims analysis using medical, pharmacy, and enrollment information from November 2010–October 2015 to evaluate the economic burden among moderate-to-severe PsO and/or PsA patients in the US Department of Defense (DoD) population.
The US DoD healthcare system uses the largest computable electronic health record system to provide healthcare to over 9.5 million beneficiaries located all over the world. This data repository includes comprehensive datasets, providing integrated information about the inpatient, outpatient, emergency room (ER), and pharmacy claims from the US DoD facility and civilian/private sector care for eligible beneficiaries. Among several national and commercial databases, the US DoD database is the one with the highest demographical similarity to the US population, capturing fully integrated individual-level health dataCitation10.
Patient selection
Adult patients were included in the moderate-to-severe PsO and/or PsA (cases) cohort if they had ≥2 diagnosis claims for PsO (International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9-CM] codes: 696.1; ICD-10-CM: L40.0–L40.4, L40.8) or PsA (ICD-9-CM: 696.0; ICD-10-CM: L40.54, L40.59) during the study period (November 2010–October 2015), and the first PsO/PsA diagnosis claim date during the identification period (November 2011–October 2014) was designated as the index date. Patients were required to have continuous health plan enrollment for ≥12 months pre- (baseline period) and 12 months post-index date (follow-up period). Patients that had ≥1 non-topical systemic therapy (including biologics and non-biologics) or phototherapy received within 30 days after index date or during the 12-month baseline period were considered moderate-to-severe patientsCitation11,Citation12.
Patients were included in the non-PsO and/or PsA (control) cohort if they had no diagnosis claim for PsO/PsA during the study period (November 2010–October 2015). The index date was randomly selected during the identification period (November 2011–October 2014) to reduce selection bias. Patients were required to have continuous health plan enrollment for ≥12 months pre- (baseline period) and 12 months post-index date (follow-up period), aged ≥18 years as of the index date, and biologic-naïve during the 12-month baseline period. Control patients were matched to moderate-to-severe PsO/PsA patients on a 10:1 ratio based on exact age, sex, geographic region, and index year.
Study variables
Baseline patients’ characteristics
Baseline socio-demographic and clinical characteristics during the 12 months pre-index date were measured.
Baseline characteristics included demographics (age, sex, and geographic region), clinical characteristics (Charlson Comorbidity index [CCI]) score, comorbidities (obesity, diabetes, ankylosing spondylitis, dyslipidemia, hypertension, etc.), and baseline concomitant medication (anxiolytic, anti-depressants, anti-diabetics, and statins).
Outcome variables
Incidence of comorbid conditions was calculated as patients who had a newly-diagnosed comorbid condition of interest during the 1-year follow-up period among those without that specific comorbidity in the baseline. The incidence of medications during follow-up was also examined among those patients who did not have that medication use in the 1-year baseline. The number of inpatient admissions and ambulatory, ER, and pharmacy visits were calculated for the 1-year follow-up period. All-cause healthcare costs were adjusted to 2015 US dollars using the medical care component of the Consumer Price Index, including inpatient, ambulatory visit, ER visit, pharmacy, total medical (inpatient + ambulatory + ER), and total costs (inpatient + ambulatory + ER + pharmacy).
Statistical analysis
Baseline and outcome variables were analyzed descriptively. Numbers and percentages were provided for categorical variables; means and standard deviations were provided for continuous variables. T-tests and Pearson Chi-squared tests were used to test statistically significant differences at the 5% level for continuous and categorical variables, respectively. Standardized differences (STDs) were calculated to distinguish practical from statistical significance, and reported as 100-times the absolute value of the actual STD to allow for easy interpretation. STDs with a value >10 were considered significant.
One-to-one propensity score matching (PSM) was conducted to evaluate healthcare resource utilization (HRU) and costs, as well as comorbidity incidence, comparing cases to controls. The following variables were controlled in the PSM model: age, sex, US geographic region, baseline CCI score, and baseline individual comorbidities including diabetes mellitus, obesity, dyslipidemia, hypertension, coronary artery disease, and baseline anxiolytic use.
Results
Patient demographics and baseline clinical characteristics
A total of 7,249 moderate-to-severe PsO and/or PsA and 72,490 controls were identified after applying the selection criteria (). The mean age of cases and controls was ∼48 years, and 51.8% were women. Mean CCI score was significantly higher in the cases vs controls (0.7 vs 0.4, p < .001) ().
Figure 1. Patient selection flow diagram.
Table 1. Baseline characteristics of moderate-to-severe PsO and/or PsA patients and controls.
Case patients had a significantly higher rate of comorbid conditions than controls. The most frequent comorbidities were dyslipidemia (38.3% vs 31.0%, p < .001), hypertension (38.1% vs 29.2%, p < .001), diabetes (15.0% vs 10.4%, p < .001), and obesity (10.7% vs 6.9%, p < .001). Case patients had greater baseline utilization of anxiolytics (33.1% vs 19.7%, p < .001), anti-depressants (25.0% vs 16.5%, p < .001), and statins (20.8% vs 16.6%, p < .001) compared to controls ().
Outcomes
After PSM, 7,245 patients PsA and/or PsA patients were matched to 7,245 controls. During follow-up, moderate-to-severe PsO and/or PsA patients had higher comorbidity incidence, including dyslipidemia (18.3% vs 13.5%, p < .001), hypertension (13.8% vs 8.7%, p < .001), obesity (8.8% vs 6.1%, p < .001), and diabetes mellitus (3.7% vs 2.3%, p < .001), as compared to control patients. Cases had a higher incidence of anti-depressants (8.2% vs 6.6%, p < .001) and anxiolytics (14.2% vs 9.4%, p < .001) during follow-up, compared to controls ().
Table 2. PSM-adjusted incidence of comorbidities and medications for moderate-to-severe PsO and/or PsA patients and controls in the 12-month follow-up period.
Cases had higher all-cause HRU, including more ambulatory (27.0 vs 16.4 visits, p < .001) and pharmacy visits (22.0 vs 12.7 visits, p < .001) during the follow-up period (). Consequently, cases incurred significantly greater all-cause healthcare costs compared to controls. Specifically, cases incurred greater inpatient ($2,196 vs $1,642; p = 0.002), ambulatory ($8,804 vs 4,642; p < .001), ER ($432 vs $350; p < .001), and pharmacy costs ($6,878 vs $1,160; p < .001). All-cause total costs (ambulatory + inpatient + ER + pharmacy) were significantly higher among cases as compared to controls ($18,311 vs $7,795; p < .001) ().
Figure 2. PSM-adjusted HRU for moderate-to-severe PsO and/or PsA patients and controls in the 12-month follow-up period. Abbreviations. HRU, healthcare resource utilization; PsO, psoriasis; PsA, psoriatic arthritis; PSM, propensity score matching.
Figure 3. PSM-adjusted healthcare costs for moderate-to-severe PsO and/or PsA patients and controls in the 12-month follow-up period. Abbreviations. PsO, psoriasis; PsA, psoriatic arthritis; PSM, propensity score matching.
Discussion
The current study was a real-world retrospective observational claims analysis among US military personnel and their dependents diagnosed with moderate-to-severe PsO and/or PsA. The study examined the incidence of comorbidities and the economic burden among moderate-to-severe PsO and/or PsA patients in the US DoD population.
Moderate-to-severe PsO and/or PsA are burdened with several comorbidities. Prodanovich et al.’sCitation13 study showed that, after controlling for baseline cardiovascular risk factors, there was evidence of ∼>70% greater odds of each of the following: ischemic heart disease, peripheral vascular disease, and atherosclerosis. The incidence rates of those three comorbidities were slightly higher among cases in our study, but were not significant after PSM. This is likely due to the different study population: the Prodanovich et al.Citation13 study used Veterans Administration data, where most individuals were male, with an average age above 65 years old; in our study, the average age was 48, and only newly-diagnosed comorbid conditions were examined.
In previous studies, the odds of being diagnosed with dyslipidemia among PsO patients ranged from 1.04–5.55Citation14. The odds of having hypertension among PsO patients was 37% higherCitation15. In previous studies, the relationship between obesity and PsO has been bi-directionalCitation16: one study showed that the odds for obesity with mild PsO are 46% greater than for not being obese; among severe PsO, the odds are as high as 123%Citation17. After controlling for age, gender, body mass index, hypertension, and hyperlipidemia, PsO increased the risk of incidence of type 2 diabetes mellitus (T2DM) by 14% in a cohort of >100,000 patientsCitation18. In our study, the most frequent comorbidities were dyslipidemia and hypertension, followed by obesity (in order of greater frequency); the cases had a higher incidence of T2DM compared to controls (3.7% vs 2.3%, p < .001), which aligns with prior related studies.
In a survey conducted by Gupta et al.Citation19, the authors’ claim that the visibility of patients’ lesions often result in feelings of embarrassment, social withdrawal, and low self-esteem. In addition, the survey revealed that patients experienced stress from anxiety, mainly due to avoiding others and being evaluated on the basis of their skinCitation19. In our current study, we observed that, in the 12-month follow-up period, moderate-to-severe PsO and/or PsA patients had a significantly higher number of anxiolytics and anti-depressant medications than controls. This finding may support the idea that PsO patients are generally anxious regarding social acceptance, which may have been a factor regarding depressive episodes, and, therefore, the prevalence of their medication-specific behavior. According to the National Psoriasis Foundation, there is a growing body of evidence that suggests a prominent link between psoriasis and mental illnessCitation20. Our findings support the literature that mental health comorbidities such as depression and anxiety may be an important prevalent illness in the sample.
A recent claims database analysis using Optum data showed that moderate-to-severe PsO and PsA patients, compared to controls, had more inpatient admissions (0.17 vs 0.10 visits, p < .001), with a mean length of stay of 1.06 days vs 0.70 days. In addition, moderate-to-severe PsO and PsA patients had more outpatient visits (19.31 vs 8.73 visits, p < .001). Consequently, these patients accrued higher total all-cause costs, of which pharmacy costs were the major driver of the economic burdenCitation21. In a recently-published study among PsA, the authors showed that total direct medical costs per year were $5,108, of which 60% were attributed to hospitalizations. Prescription pharmacy costs were $14,174 during the 12-month follow-up period, of which almost 100% of prescriptions were for biologicsCitation22. Therefore, drug costs contributed a significant burden to patients diagnosed with moderate-to-severe PsO. Our current study aligns with these findings.
In our study, the results show that pharmacy costs contributed to ∼37.5% of total all-cause costs. During follow-up, 79% of the moderate-to-severe PsO and/or PsA patients had injections, with a mean of 5.1 injections over 1 year (Supplementary Table 1), reflecting that moderate-to-severe patients were prescribed either biologic or intravenous (IV) non-biologic therapy during the follow-up period. Moderate-to-severe PsO and/or PsA patients incurred mean PsO and/or PsA-related pharmacy costs of $4,550 during the 12-month follow-up period, representing 66% of all-cause pharmacy costs. Therefore, >50% of the PsO and/or PsA-related burden was also attributed to pharmacy use. Approximately 35.6% of the total all-cause medical costs—and 47% of the total all-cause costs—were moderate-to-severe PsO and/or PsA-related, respectively. Thus, a significant portion of the cost burden was attributed to moderate-to-severe PsO and/or PsA.
With a plethora of comorbidities and higher utilization of biologicals and/or non-biologicals during the follow-up, these patients incurred significant HRU and, consequently, higher all-cause and drug-specific costs. Since pharmacy utilization was the primary burden of the costs, and an average of five injections were used during the 12-month follow-up, it may be inferred that the burden was due to biologic or non-biologics such as IV steroids and disease-modifying anti-rheumatic drugs (DMARDs). Further research may be warranted to study the type of pharmacy use that could have been the driver of these costs in this population, and additional analysis needs to be prioritized to devise strategies to contain these costs among military personnel and dependents.
Strengths
Our analysis was conducted from a large, comprehensive administrative national DoD database that offers a large sample of military personnel and their dependents to evaluate the economic burden and comorbidities of moderate-to-severe PsO and/or PsA patients. The database is composed of a diverse population; 14% consist of those in active duty and the remaining 86% consists of dependents or retireesCitation23. In addition, 78% of the DoD population is aged ≥18 years, compared to 75% of the general US population, which may add to the external validity of this analysisCitation23. To the best of our knowledge, this is the first economic analysis among moderate-to-severe PsO and/or PsA patients within the DoD population.
Limitations
Claims data are collected for payment purposes; therefore, such data may have limitations for clinical research. There is the potential for misclassification errors, as it is possible that some ICD-9-CM or ICD-10-CM codes may have been incorrectly recorded, misused, or not entered at all. In studies using administrative claims data, the true validity of the analyzed variables may not be known with certainty. The administrative databases may also lack information of some critical covariates, and the potential for residual confounding may exist in our modeling results. In addition, the presence of a pharmacy claim may not necessarily mean the prescription was consumed as indicated.
Since this analysis was among DoD patients, study findings may not be generalized to other populations. Access to treatments may vary for patients enrolled in individual commercial plans, as compared to those in the TRICARE program. Therefore, pharmacy costs may be partially reflective of the formulary access protocols. In addition, several limitations for the DoD database were recognized. Of note, we were not able to identify mental health diagnoses in the data, as this information is no longer available in the database for privacy reasons. Also, information on pharmacy costs for drugs distributed inside the military facility was not available.
Conclusions
In conclusion, some important comorbidities identified among psoriasis patients were found to be cardiovascular diseases (such as dyslipidemia and hypertension), obesity, and diabetes mellitus. In addition, moderate-to-severe PsO and/or PsA was significantly associated with increased medical costs, primarily due to pharmacy utilization. Therefore, the increased presence of comorbidities and the resulting pharmacy utilization were the drivers of healthcare costs.
Transparency
Declaration of funding
This study was funded by Janssen Scientific Affairs, LLC.
Declaration of financial/other relationships
SL is an employee and stockholder in Johnson & Johnson, the parent company of Janssen Scientific Affairs, LLC, the study sponsor. LX, YW, and NV are employees of STATinMED Research, which is a paid consultant to Janssen Scientific Affairs, LLC, the study sponsor. OB has no interests to disclose. Peer reviewers on this manuscript have received an honorarium from JME for their review work. A peer reviewer on this manuscript has declared receiving funding from Janssen and other companies, including Abbvie, Celgene, Lilly, and Novartis. The remaining peer reviewers have no relevant financial relationships to disclose.
Previous presentation
This research has been presented at the American Academy of Dermatology (AAD) Annual Meeting in March 2017.
Acknowledgments
The authors wish to thank Chris Haddlesey of STATinMED Research for medical editorial support during manuscript development.
References
- Zhang XJ, Huang W, Yang S, et al. Psoriasis genome-wide association study identifies susceptibility variants within LCE gene cluster at 1q21. Nat Genet 2009;41:205-10
- Langley RG, Krueger GG, Griffiths CE. Psoriasis: epidemiology, clinical features, and quality of life. Ann Rheumatic Dis 2005;64(suppl 2):ii18-ii23
- Helmick CG, Lee-Han H, Hirsch SC, et al. Prevalence of psoriasis among adults in the U.S.: 2003–2006 and 2009–2010 National Health and Nutrition Examination Surveys. Am J Prev Med 2014;47:37-45
- Prey S, Paul C, Bronsard V, et al. Assessment of risk of psoriatic arthritis in patients with plaque psoriasis: a systematic review of the literature. J Eur Acad Dermatol Venereol 2010;24(suppl2):31-5
- Puig L. Cardiometabolic comorbidities in psoriasis and psoriatic arthritis. Int J Mol Sci 2017;19:E58
- National Psoriasis Foundation [Internet]. Comorbidities associated with psoriatic disease. Portland, OR: National Psoriasis Foundation. https://www.psoriasis.org/about-psoriasis/related-conditions
- Khobzey K, Liskova I, Szegedi A, et al. Effectiveness of adalimumab in the treatment of scalp and nail affection in patients with moderate to severe plaque psoriasis in routine clinical practice. Acta Dermatovenerol Alp Pannonica Adriat 2017;26:11-14
- Brezinski EA, Dhillon JS, Armstrong AW. Economic burden of psoriasis in the United States: a systematic review. JAMA Dermatol 2015;151:651-8
- Schaefer CP, Cappelleri JC, Cheng R, et al. Health care resource use, productivity, and costs among patients with moderate to severe plaque psoriasis in the United States. J Am Acad Dermatol 2015;73:585-93.e3
- Military Health System [Internet]. Evaluation of the TRICARE Program: Access, Cost and Quality. Fiscal year 2015. Report to Congress. Falls Church, VA: Defense Health Agency. http://www.health.mil/Military-Health-Topics/Access-Cost-Quality-and-Safety/Health-Care-Program-Evaluation/Annual-Evaluation-of-the-TRICARE-Program
- Malatestinic W, Nordstrom B, Wu JJ, et al. Characteristics and medication use of psoriasis patients who may or may not qualify for randomized controlled trials. J Manag Care Spec Pharm 2017;2:370-81
- Feldman SR, Evans C, Russell MW. Systemic treatment for moderate to severe psoriasis: estimates of failure rates and direct medical costs in a north‐eastern US managed care plan. J Dermatol Treat 2005;16:37-42
- Prodanovich S, Kirsner RS, Kravetz JD, et al. Association of psoriasis with coronary artery, cerebrovascular, and peripheral vascular diseases and mortality. Arch Dermatol 2009;145:700-3
- Ma C, Harskamp CT, Armstrong EJ, et al. The association between psoriasis and dyslipidaemia: a systematic review. Br J Dermatol 2013;168:486-95
- Cohen AD, Weitzman D, Dreiher J. Psoriasis and hypertension: a case-control study. Acta Derm Venereol 2010;90:23-6
- Ni C, Chiu MW. Psoriasis and comorbidities: links and risks. Clin Cosmet Investig Dermatol 2014;7:119-32
- Armstrong AW, Harskamp CT, Armstrong EJ. The association between psoriasis and obesity: a systematic review and meta-analysis of observational studies. Nutr Diabetes 2012;2:e54
- Azfar RS, Seminara NM, Shin DB, et al. Increased risk of diabetes mellitus and likelihood of receiving diabetes mellitus treatment in patients with psoriasis. Arch Dermatol 2012;148:995-1000
- Gupta MA, Gupta AK, Watteel GN. Perceived deprivation of social touch in psoriasis is associated with greater psychologic morbidity: an index of the stigma experience in dermatologic disorders. Cutis 1998;61:339-42
- Leavitt M, National Psoriasis Foundation [Internet]. The link between psoriatic disease and mental illness. Portland, OR: National Psoriasis Foundation; 2015. https://www.psoriasis.org/advance/link-between-psoriatic-disease-and-mental-illness
- Feldman S, Zhao Y, Shi L, et al. Economic and comorbidity burden among moderate-to-severe psoriasis patients with comorbid psoriatic arthritis. Arthritis Care Res (Hoboken) 2015;67:708-17
- Greenberg JD, Palmer JB, Li Y, et al. Healthcare resource use and direct costs in patients with ankylosing spondylitis and psoriatic arthritis in a large US cohort. J Rheumatol 2016;43:88-96
- Francis K, Yu C, Alvrtsyan H, et al. Healthcare utilization and costs associated with dabigatran compared to warfarin treatment in newly diagnosed patients with non-valvular atrial fibrillation. Curr Med Res Opin 2015;31:2189-95