Healthcare resource utilization and costs among psoriasis patients treated with biologics, overall and by disease severity

Abstract

Aims: To describe healthcare resource utilization (HCRU) and costs among biologic-treated psoriasis patients in the US, overall and by disease severity.

Materials and methods: IQVIA PharMetrics Plus administrative claims data were linked with Modernizing Medicine Data Services Electronic Health Record data and used to select adult psoriasis patients between April 1, 2010 and December 31, 2014. Eligible patients were classified by disease severity (mild, moderate, severe) using a hierarchy of available clinical measures. One-year outcomes included all-cause and psoriasis-related outpatient, emergency department, inpatient, and pharmacy HCRU and costs.

Results: This study identified 2,130 biologic-treated psoriasis patients: 282 (13%) had mild, 116 (5%) moderate, and 49 (2%) severe disease; 1,683 (79%) could not be classified. The mean age was 47.6 years; 45.4% were female. Relative to mild psoriasis patients, patients with moderate or severe disease had more median all-cause outpatient encounters (28.0 [mild] vs 32.0 [moderate], 36.0 [severe]), more median psoriasis-related outpatient encounters (6.0 [mild] vs 7.5 [moderate], 8.0 [severe]), and a higher proportion of overall claims for medications that were psoriasis-related (28% [mild] vs 37% [moderate], 34% [severe]). Relative to mild psoriasis patients, patients with moderate or severe disease had higher median all-cause total costs ($37.7k [mild] vs $42.3k [moderate], $49.3k [severe]), higher median psoriasis-related total costs ($32.7k [mild] vs $34.9k [moderate], $40.5k [severe]), higher median all-cause pharmacy costs ($33.9k [mild] vs $36.5k [moderate], $36.4k [severe]), and higher median psoriasis-related pharmacy costs ($32.2k [mild] vs $33.9k [moderate], $35.6k [severe]).

Limitations: The assessment of psoriasis disease severity may not have necessarily coincided with the timing of biologic use. The definition of disease severity prevented the assessment of temporality, and may have introduced selection bias.

Conclusions: Biologic-treated patients with moderate or severe psoriasis cost the healthcare system more than patients with mild psoriasis, primarily driven by higher pharmacy costs and more outpatient encounters.

Keywords:

• Psoriasis
• biologics
• healthcare costs
• resource utilization
• psoriasis disease severity

JEL Classification Codes:

• I10
• I12

Introduction

Psoriasis is a chronic inflammatory skin disorder, with the plaque type characterized by plaques that are usually painful, itchy, thick, red, and scaly^1,2. Recent studies estimate that the age-adjusted prevalence of psoriasis among adults in the US is 3% (∼6.7 million adults), although estimates as low as 1% have been reported^3,4. Of all psoriasis patients, ∼18% have moderate-to-severe forms of the disease³.

A variety of methods are used to assess psoriasis disease severity^5,6. While the psoriasis area severity index (PASI) is the most commonly used method in clinical trials, it is typically not used in the clinical practice setting^5,7. Other disease severity measures include body surface area (BSA), Physician’s Global Assessment (PGA), and Patient-Reported Global Assessment (PtGA)^6,8. Disease severity, as determined by BSA, affected by psoriasis can be categorized into the following three groups: less than 3% is considered mild, 3–10% moderate, and more than 10% severe⁹. PGA and PtGA, as assessed by a physician and patient, respectively, are used to measure plaque severity and have many variations, including 5-, 6-, or 7-point scales ranking plaques from “clear” to “severe”^8,10.

The burden of psoriasis extends beyond skin manifestations to include associated comorbidities (most notably psoriatic arthritis found in up to 43% of psoriasis patients^11,12), decreased quality-of-life (QoL) and productivity, and increased cost of care^13–15. The burden of cardiovascular and related diseases, some malignancies, and other autoimmune conditions is particularly high among psoriasis patients^13,16–18. Psoriasis is also associated with depression and other mental health conditions¹⁹. Annual psoriasis direct costs in the US have been estimated to range from $52–$63 billion (2013 US dollars)¹⁴, with moderate-to-severe patients accruing disproportionately high annual healthcare cost when compared to mild patients ($10,593 vs $5,011, respectively, using 2007 US dollars)¹⁵.

The use of biologic therapies to treat moderate-to-severe chronic plaque psoriasis was first approved by the US Food and Drug Administration (FDA) in 2003 and has resulted in improved treatment outcomes, higher clearance, acceptable safety profiles, and better health-related quality-of-life^2,20,21. However, the costs associated with biologic treatments are high^22,23, and we have identified no publications of health care resource utilization (HCRU) and costs among biologic-treated psoriasis patients stratified by disease severity in the US. The objective of this study was to describe healthcare resource utilization and costs among psoriasis patients treated with biologics overall and by disease severity.

Methods

Study design

This was a retrospective study of adult psoriasis patients in the US using IQVIA PharMetrics Plus adjudicated medical and pharmacy claims data linked with Modernizing Medicine Data Services, Inc. (MMDS) Electronic Health Record (EHR) data. The study period was from January 1, 2010 to December 31, 2015. Patients were followed from first qualifying biologic prescription medication claim to the end of the 360-day follow-up period. This study was conducted using de-identified data, in compliance with the Health Insurance Portability and Accountability Act of 1996 (HIPAA).

Data sources

The IQVIA PharMetrics Plus database contains adjudicated claims for more than 130 million unique enrollees across the US, with both medical and pharmacy coverage since 2006, and with diverse representation of geography, payers, providers, and therapy areas. PharMetrics Plus was used to provide de-identified patient data on prescription medication use, medical history, healthcare resource utilization and costs, and patient age and sex.

The MMDS EHR contains encounter data from EHRs for more than 35 million patients in the US across eight medical specialties, including dermatology. As of the third quarter of 2017, more than 70 million visits to dermatologists have been recorded. Psoriasis patients were defined on a de-identified basis as those patients with a provider-recorded clinical diagnosis of psoriasis in the MMDS EHR. The MMDS EHR was also used to provide de-identified patient data on psoriasis disease severity and patient race/ethnicity.

These two databases were linked using a proprietary encryption algorithm and a deterministic matching algorithm^24,25. Using a trusted third party, a patented encryption algorithm^26–28 that ensured compliance with HIPAA regulations was configured and deployed at MMDS to generate irreversible, hashed, and concatenated patient tokens. All identifiable EHR data remained at MMDS. At the trusted third party, the patient tokens then underwent a deterministic matching process to be assigned unique and persistent IQVIA patient IDs. The deterministic algorithm uses specific data points (e.g. first name, last name, sex, date of birth, street address, zip code) rather than statistical probability to ensure continuity of patient records across datasets. Once the IQVIA patient IDs were received from the trusted third party, patients with matching IQVIA patient IDs in both data sources were then assembled into a combined database.

Patient selection

Adult (≥18 years of age at any point during the calendar year of the index medication) psoriasis patients with ≥1 prescription medication claim for a biologic for the treatment of psoriasis were selected for study inclusion. The index medication was the first biologic prescription medication claim occurring on or after the date of the patient’s first MMDS EHR visit with continuous medical and pharmacy benefit eligibility in PharMetrics Plus throughout the baseline period (90 days prior to the claim date) and the follow-up period (360 days after the claim date). The index date was the date of the index medication claim. Patients with a single MMDS (or PharMetrics Plus) patient ID linking to multiple PharMetrics Plus (or MMDS) patient IDs and patients with data not meeting quality standards in PharMetrics Plus (e.g. missing year of birth or sex; payer type with high probability of incomplete claims information, suspect enrollment dates) were excluded.

Study measures

Patients were classified into disease severity cohorts based on their first recorded severity measure in the MMDS EHR during the follow-up period using a hierarchical methodology of available measures of PGA, BSA, or PtGA. Mild disease severity was defined as PGA 0–2; if no PGA available, BSA <3%; if no PGA and no BSA available, PtGA 0–2. Moderate disease severity was defined as PGA 3; if no PGA available, BSA 3–10%; if no PGA and no BSA available, PtGA 3. Severe disease severity was defined as PGA 4–5; if no PGA available, BSA >10%; if no PGA and no BSA available, PtGA 4–5. Prescription medications were identified using National Drug Codes (NDC) and the Healthcare Common Procedure Coding System (HCPCS). Phototherapy was identified via HCPCS or the American Medical Association’s Current Procedural Terminology (CPT) codes.

The primary study outcomes included outpatient, emergency department (ED), inpatient, and pharmacy healthcare resource utilization and costs, which were reported from adjudicated medical and pharmacy claims. Outpatient healthcare resource utilization and costs included physician office visits, laboratory, radiology, surgical services, and ancillary/other outpatient services. Psoriasis-related healthcare resource utilization and costs were defined as any claim associated with a diagnosis or medication for psoriasis. Costs were presented overall and by costs to the payer and costs to the patient.

Patient demographics included age, sex, and race/ethnicity. Clinical characteristics were assessed during the 90-day baseline period prior to the index date and included body mass index (BMI), Charlson-Deyo Comorbidity Index (CCI) score²⁹, and top 10 comorbidities (recorded using International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes at the 3-digit level).

Statistical analysis

All study variables, including baseline and outcomes measures, were analyzed descriptively using percentage distributions for categorical variables and descriptive statistics (mean, standard deviation [SD], median, interquartile range [IQR], minimum, and maximum) for continuous and count variables. Unless otherwise specified, all patients were included in the descriptive statistics presented. SAS version 9.2 (SAS Institute Inc., Cary, NC) was used for constructing the analytic dataset and for statistical analyses.

Results

Sample selection and baseline characteristics

Of the 94,682 psoriasis patients present in the MMDS EHR database that linked to the PharMetrics Plus database, 49.5% (n = 46,859) had ≥1 claim during the study period. After applying the study inclusion and exclusion criteria, our final biologic-treated cohort consisted of 2,130 patients: 282 (13.2%) had mild, 116 (5.4%) moderate, and 49 (2.3%) severe psoriasis; 1,683 (79.0%) were unable to be classified. The average length of time from the index date to the date of the first recorded disease severity measurement during the follow-up period was 153.0 (SD =106.2) days for the overall biologic cohort, 161.7 (SD =99.6) days for biologic patients with mild psoriasis, 138.7 (SD =113.3) days for biologic patients with moderate psoriasis, and 137.2 (SD =121.4) days for biologic patients with severe psoriasis.

The mean age of the biologic cohort was 47.6 (SD = 11.7) years; 45.4% were female; and 73.7% were white. The average CCI score was 0.3 (SD =0.6). Of the biologic-treated cohort, 1.6% (n = 35) had ≥1 ICD-9/10 code for comorbid PsA during the study period (1.4%, n = 4 [mild], 2.6%, n = 3 [moderate], and 4.1%, n = 2 [severe]). The top five comorbidities recorded during the baseline period were essential hypertension (18.3%, n = 389), disorders of lipid metabolism (16.3%, n = 348), other and unspecified joint disorders (12.0%, n = 255), general symptoms such as fatigue or sleep disturbances (11.5%, n = 244), and other dermatoses (10.4%, n = 221). The top five comorbidities recorded during the follow-up period were essential hypertension (36.0%, n = 767), disorders of lipid metabolism (34.5%, n = 734), general symptoms such as fatigue or sleep disturbances (27.9%, n = 594), other and unspecified joint disorders (24.7%, n = 526), and respiratory system symptoms and other chest symptoms (24.5%, n = 521) (see Table 1 for a full list of the top 10 comorbidities recorded during the baseline and follow-up periods).

Table 1. Demographic characteristics for the biologic cohort, overall and stratified by disease severity.

	Biologic cohort, All		Biologic cohort, with recorded disease severity		Biologic cohort, Mild disease severity		Biologic cohort, Moderate disease severity		Biologic cohort, Severe disease severity
Demographic characteristics^a	(n = 2,130)		(n = 447)		(n = 282)		(n = 116)		(n = 49)
Years of age (mean, SD)	47.6	11.7	47.4	11.9	47.3	12.1	47.5	11.4	47.9	12.1
Sex (n, %)
Female	968	45.4%	203	45.4%	125	44.3%	54	46.6%	24	49.0%
Race (n, %)
White	1,570	73.7%	340	76.1%	214	75.9%	90	77.6%	36	73.5%
Other/unknown	560	26.3%	107	23.9%	68	24.1%	26	22.4%	13	26.5%
BMI^b (kg/m^2) (mean, SD)	30.2	6.1	31.1	6.6	30.1	6.3	31.9	5.1	34.8	10.0
BMI group (n, %)
Underweight/normal (≤18.5–24.9)	40	1.9%	14	0.7%	13	4.6%	0	0.0%	1	2.0%
Overweight/obese (>25)	168	7.9%	61	2.9%	33	11.7%	20	17.2%	8	16.3%
Unknown	1,922	90.2%	372	83.2%	236	83.7%	96	82.8%	40	81.6%
CCI^c (mean, SD)	0.3	0.6	0.2	0.6	0.2	0.6	0.2	0.6	0.3	0.6
Top 10 most common comorbidities during the baseline period^d (n, %)
Essential hypertension (ICD9: 401)	389	18.3%	73	16.3%	50	17.7%	14	12.1%	9	18.4%
Disorders of lipid metabolism (ICD9: 272)	348	16.3%	77	17.2%	50	17.7%	16	13.8%	11	22.4%
Other and unspecified joint disorders (ICD9: 719)	255	12.0%	50	11.2%	27	9.6%	12	10.3%	11	22.4%
General symptoms such as fatigue or sleep disturbances (ICD-9 780)	244	11.5%	54	12.1%	38	13.5%	11	9.5%	5	10.2%
Other dematoses (ICD9: 702)	221	10.4%	36	8.1%	26	9.2%	7	6.0%	3	6.1%
Benign neoplasm of the skin (ICD9: 216)	221	10.4%	36	8.1%	26	9.2%	8	6.9%	2	4.1%
Diabetes mellitus (ICD9: 250)	188	8.8%	35	7.8%	24	8.5%	5	4.3%	6	12.2%
Respiratory system symptoms and other chest symptoms (ICD9: 786)	179	8.4%	35	7.8%	22	7.8%	9	7.8%	4	8.2%
Other and unspecified back disorders (ICD9: 724)	178	8.4%	34	7.6%	27	9.6%	3	2.6%	4	8.2%
Other disorders of soft tissues (ICD9: 729)	174	8.2%	35	7.8%	26	9.2%	5	4.3%	4	8.2%
Top 10 most common comorbidities during the follow-up period^d (n, %)
Essential hypertension (ICD9: 401)	767	36.0%	166	37.1%	108	38.3%	40	34.5%	18	36.7%
Disorders of lipid metabolism (ICD9: 272)	734	34.5%	167	37.4%	106	37.6%	41	35.3%	20	40.8%
General symptoms such as fatigue or sleep disturbances (ICD-9 780)	594	27.9%	112	25.1%	69	24.5%	32	27.6%	11	22.4%
Other and unspecified joint disorders (ICD9: 719)	526	24.7%	100	22.4%	60	21.3%	25	21.6%	15	30.6%
Respiratory system symptoms and other chest symptoms (ICD9: 786)	521	24.5%	101	22.6%	62	22.0%	26	22.4%	13	26.5%
Other disorders of soft tissues (ICD9: 729)	410	19.2%	73	16.3%	44	15.6%	23	19.8%	6	12.2%
Other and unspecified back disorders (ICD9: 724)	391	18.4%	76	17.0%	58	20.6%	12	10.3%	6	12.2%
Other dematoses (ICD9: 702)	330	15.5%	71	15.9%	46	16.3%	18	15.5%	7	14.3%
Diabetes mellitus (ICD9: 250)	314	14.7%	62	13.9%	44	15.6%	11	9.5%	7	14.3%
Non-specific findings on examination of blood (ICD9: 790)	313	14.7%	74	16.6%	45	16.0%	20	17.2%	9	18.4%

^aDemographic and clinical characteristics were measured as of the index date, unless otherwise specified.

^bBody mass index (BMI) at index was calculated by taking the height and weight recorded (on the same day) either on or closest to the index date (within a window of ±90 days from the index date).

^cCharlson-Deyo Comorbidity Index (CCI) score²⁹.

^dBased on ≥1 International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM, ICD9) diagnosis code at the 3-digit level for the given condition during the respective time period.

Abbreviations. SD, standard deviation; kg, kilogram; m, meter.

Outcomes

Healthcare resource utilization

Relative to patients who had mild psoriasis, patients with moderate or severe disease had more median all-cause outpatient encounters (28.0 [mild] vs 32.0 [moderate], 36.0 [severe]), more median psoriasis-related outpatient encounters (6.0 [mild] vs 7.5 [moderate], 8.0 [severe]), a higher proportion of overall claims for medications that were psoriasis-related (28.2% [mild] vs 37.0% [moderate], 34.4% [severe]), and a higher proportion of overall claims for non-biologic psoriasis medications (6.8% [mild] vs 14.8% [moderate], 17.5% [severe]). Relative to patients with severe psoriasis, patients with mild or moderate disease had a higher proportion of overall claims for biologic psoriasis medications (21.5% [mild], 22.3% [moderate] vs 16.9% [severe]) (Table 2; Figures 1 and 2).

Figure 1. Median number of outpatient encounters (all-cause and psoriasis-related) during the follow-up period, stratified by disease severity.

Figure 2. Proportion of prescription medication claims that were psoriasis-related during the follow-up period, stratified by disease severity.

Table 2. Healthcare resource utilization during the follow-up period, overall and stratified by disease severity.

Healthcare resource utilization^a	Biologic cohort, All		Biologic cohort, with recorded disease severity		Biologic cohort, Mild disease severity		Biologic cohort, Moderate disease severity		Biologic cohort, Severe disease severity
	(n = 2,130)		(n = 447)		(n = 282)		(n = 116)		(n = 49)
All-cause (mean, median)
Total number of outpatient encounters^b	43.9	32	40.2	31	40.1	28	39.2	32	43.8	36
Physician office visits	12.1	8	11.8	8	11.4	8	12.0	9	13.9	9
Total number of emergency department (ED) visits^c	0.3	0	0.2	0	0.2	0	0.3	0	0.3	0
Total number of inpatient visits	0.1	0	0.1	0	0.1	0	0.1	0	0.2	0
Cumulative number of inpatient days^d	6.3	4	10.2	5	7.6	5	4.5	4	26.8	7
Total number of all prescription medication claims^e	36.3	28	34.8	26	35.4	26.5	33.2	26	35.5	24
Psoriasis-related^f (mean, median)
Total number of outpatient encounters^b	7.0	5	8.7	7	8.2	6	9.6	7.5	9.8	8
Physician office visits	2.5	2	3.1	3	2.8	2.5	3.7	3	3.7	3
Total number of emergency department (ED) visits^c	0.0	0	0.0	0	0.0	0	0.0	0	0.0	0
Total number of inpatient visits	0.0	0	0.0	0	0.0	0	0.0	0	0.1	0
Cumulative number of inpatient days^d	4.8	4	5.4	5	5.0	5	3.3	2.5	11.0	11
Total number of all psoriasis prescription medication claims^e	10.7	10	10.8	10	10.0	9	12.3	11	12.2	11
Total number of all biologic psoriasis prescription medication claims^e	7.5	8	7.4	7	7.6	8	7.4	8	6.0	5

^aAs assessed during the follow-up period, excluding the index date.

^bOffice visits were summarized by patient, provider, and date. Therefore, if a patient saw the same provider three times on the same day then that was recorded as one office visit. However, if a patient saw three different providers on the same day then that was recorded as three office visits. All other outpatient categories (i.e. “Lab”, “Pathology”, “Radiology”, “Surgical Services”, and “Ancillary/Other Outpatient Services”) are summarized only by patient and procedure. Therefore, if a patient had three different lab tests drawn on the same day then those three lab tests were counted as three separate events.

^cEmergency department (ED) visits that transitioned to inpatient visits were counted as both ED visits and inpatient visits.

^dAmong the n = 154 patients with ≥1 all-cause inpatient stay and n = 34 patients with ≥1 PsO-related inpatient stay.

^eIncluding medications recorded using J-codes.

^fPsoriasis-related was defined as any claim associated with a diagnosis for psoriasis and/or psoriatic arthritis and/or any NDC/HCPCS/CPT for medications indicated for psoriasis.

Healthcare costs

Considering overall costs (i.e. costs to the payer plus costs to the patient), relative to patients who had mild psoriasis, patients with moderate or severe disease had higher: median all-cause total costs ($37.7k [mild] vs $42.3k [moderate], $49.3k [severe]), median psoriasis-related total costs ($32.7k [mild] vs $34.9k [moderate], $40.5k [severe]), median all-cause pharmacy costs ($33.9k [mild] vs $36.5k [moderate], $36.4k [severe]), and median psoriasis-related pharmacy costs ($32.2k [mild] vs $33.9k [moderate], $35.6k [severe]) (Table 3; Figure 3). Costs to the payer were also substantially higher for patients with moderate or severe psoriasis relative to patients with mild disease (Table 4).

Figure 3. Median total and pharmacy costs (all-cause and psoriasis-related) during the follow-up period, stratified by disease severity.

Table 3. Healthcare resource costs during the follow-up period, overall and stratified by disease severity (payer and patient combined).

	Biologic cohort, All		Biologic cohort, with recorded disease severity		Biologic cohort, Mild disease severity		Biologic cohort, Moderate disease severity		Biologic cohort, Severe disease severity
Healthcare costs^a	(n = 2,130)		(n = 447)		(n = 282)		(n = 116)		(n = 49)
All-cause (mean, median)
Total healthcare costs (overall)	$42,051	$36,293	$46,323	$39,583	$44,830	$37,743	$43,970	$42,270	$60,485	$49,318
Total outpatient costs	$5,070	$2,179	$4,456	$2,134	$4,272	$1,972	$4,060	$2,555	$6,448	$2,568
Physician office visits	$1,423	$850	$1,270	$888	$1,214	$840	$1,247	$989	$1,646	$889
Total emergency department (ED) costs	$288	—	$285	—	$164	—	$513	—	$441	—
Total inpatient costs	$1,998	—	$2,979	—	$2,523	—	$1,580	—	$8,913	—
Total pharmacy costs^b	$34,696	$31,830	$38,604	$34,362	$37,870	$33,912	$37,818	$36,485	$44,682	$36,386
Psoriasis-related^c (mean, median)
Total healthcare costs (overall)	$32,549	$30,446	$37,440	$33,429	$36,394	$32,661	$37,418	$34,867	$43,509	$40,462
Total outpatient costs	$470	$257	$632	$343	$489	$321	$535	$366	$1,684	$371
Physician office visits	$246	$177	$308	$241	$273	$237	$337	$268	$441	$271
Total emergency department (ED) costs	$23	—	$41	—	$10	—	$134	—	$5	—
Total inpatient costs	$411	—	$891	—	$873	—	$712	—	$1,413	—
Total pharmacy costs^b	$31,646	$29,898	$35,876	$32,778	$35,022	$32,161	$36,037	$33,935	$40,406	$35,595

^aAs assessed during the follow-up period, excluding the index date.

^bIncluding medications recorded using J-codes.

^cPsoriasis-related was defined as any claim associated with a diagnosis for psoriasis and/or psoriatic arthritis and/or any National Drug Codes (NDC)/Healthcare Common Procedure Coding System (HCPCS)/American Medical Association's Current Procedural Terminology (CPT) codes for medications indicated for psoriasis.

Outliers have not been removed from the analyses.

Table 4. Healthcare resource costs during the follow-up period, overall and stratified by disease severity (payer only).

	Biologic cohort, All		Biologic cohort, with recorded disease severity		Biologic cohort, Mild disease severity		Biologic cohort, Moderate disease severity		Biologic cohort, Severe disease severity
Healthcare costs^a	(n = 2,130)		(n = 447)		(n = 282)		(n = 116)		(n = 49)
All-cause (mean, median)
Total healthcare costs (overall)	$38,964	$33,435	$43,216	$36,877	$41,779	$34,121	$40,870	$39,214	$57,037	$44,982
Total outpatient costs	$4,085	$1,569	$3,693	$1,565	$3,512	$1,435	$3,282	$1,886	$5,700	$1,915
Physician office visits	$1,000	$594	$963	$645	$915	$597	$942	$710	$1,291	$632
Total emergency department (ED) costs	$234	—	$247	—	$138	—	$451	—	$394	—
Total inpatient costs	$1,883	—	$2,832	—	$2,419	—	$1,348	—	$8,724	—
Total pharmacy costs^b	$32,762	$29,987	$36,444	$32,657	$35,709	$32,093	$35,789	$33,452	$42,219	$33,724
Psoriasis-related^c (mean, median)
Total healthcare costs (overall)	$30,858	$28,538	$35,446	$31,594	$34,441	$30,821	$35,455	$32,264	$41,208	$35,361
Total outpatient costs	$355	$158	$491	$214	$359	$202	$367	$248	$1,544	$285
Physician office visits	$171	$104	$214	$150	$186	$138	$227	$173	$340	$183
Total emergency department (ED) costs	$20	—	$36	—	$8	—	$117	—	$5	—
Total inpatient costs	$385	—	$835	—	$821	—	$649	—	$1,356	—
Total pharmacy costs^b	$30,098	$28,076	$34,084	$31,004	$33,253	$30,409	$34,321	$32,041	$38,303	$33,138

^aAs assessed during the follow-up period, excluding the index date.

^bIncluding medications recorded using J-codes.

^cPsoriasis-related was defined as any claim psoriasis with a diagnosis for psoriasis and/or psoriatic arthritis and/or any National Drug Codes (NDC)/Healthcare Common Procedure Coding System (HCPCS)/American Medical Association’s Current Procedural Terminology (CPT) codes for medications indicated for psoriasis.

Outliers have not been removed from the analyses.

When breaking out the overall costs by costs to the patient, relative to patients who had mild or moderate psoriasis, patients with severe disease had higher: median all-cause total costs ($2.1k [mild], $2.1k [moderate] vs $2.7k [severe]), median psoriasis-related total costs ($0.8k [mild], $0.9k [moderate] vs $1.2k [severe]), median all-cause pharmacy costs ($1.1k [mild], $0.9k [moderate] vs $1.3k [severe]), and median psoriasis-related pharmacy costs ($0.6k [mild], $0.6k [moderate] vs $0.7k [severe]) (Table 5).

Table 5. Healthcare resource costs during the follow-up period, overall and stratified by disease severity (patient only).

	Biologic cohort, All		Biologic cohort, with recorded disease severity		Biologic cohort, Mild disease severity		Biologic cohort, Moderate disease severity		Biologic cohort, Severe disease severity
Healthcare costs^a	(n = 2,130)		(n = 447)		(n = 282)		(n = 116)		(n = 49)
All-cause (mean, median)
Total healthcare costs (overall)	$3,087	$1,948	$3,107	$2,159	$3,051	$2,144	$3,100	$2,140	$3,448	$2,704
Total outpatient costs	$985	$412	$763	$434	$760	$392	$778	$472	$748	$359
Physician office visits	$423	$208	$306	$209	$299	$197	$305	$245	$355	$221
Total emergency department (ED) costs	$53	—	$38	—	$26	—	$61	—	$48	—
Total inpatient costs	$115	—	$147	—	$104	—	$232	—	$189	—
Total pharmacy costs^b	$1,934	$932	$2,160	$1,067	$2,161	$1,104	$2,030	$884	$2,463	$1,258
Psoriasis-related^c (mean, median)
Total healthcare costs (overall)	$1,691	$645	$1,993	$881	$1,952	$842	$1,963	$863	$2,301	$1,186
Total outpatient costs	$116	$63	$140	$95	$129	$88	$168	$117	$140	$93
Physician office visits	$75	$52	$94	$74	$87	$64	$110	$84	$101	$82
Total emergency department (ED) costs	$2	—	$6	—	$2	—	$17	—	—	—
Total inpatient costs	$25	—	$55	—	$52	—	$63	—	$58	—
Total pharmacy costs^b	$1,548	$484	$1,792	$623	$1,769	$614	$1,716	$596	$2,104	$681

^aAs assessed during the follow-up period, excluding the index date.

^bIncluding medications recorded using J-codes.

^cPsoriasis-related was defined as any claim psoriasis with a diagnosis for psoriasis and/or psoriatic arthritis and/or any National Drug Codes (NDC)/Healthcare Common Procedure Coding System (HCPCS)/American Medical Association’s Current Procedural Terminology (CPT) codes for medications indicated for psoriasis.

Note: Outliers have not been removed from the analyses.

Looking at overall costs for the biologic-treated cohort, all-cause pharmacy costs represented 82.5% of all-cause total costs (84.5% [mild], 86.0% [moderate], and 73.9% [severe]) and psoriasis-related pharmacy costs represented 97.2% of psoriasis-related total costs (96.2% [mild], 96.3% [moderate], and 92.9% [severe]) (Figure 4). While <50% of all-cause pharmacy claims were psoriasis-related (Figure 3), 91.2% of all-cause pharmacy costs were psoriasis-related (92.5% [mild], 95.3% [moderate], and 90.4% [severe]). Psoriasis-related pharmacy costs represented 75.3% of all-cause total costs (78.1% [mild], 82.0% [moderate], and 66.8% [severe]). All-cause outpatient costs represented 12.1% of all-cause total costs (9.5% [mild], 9.2% [moderate], 10.7% [severe]), all-cause inpatient costs represented 4.8% of all-cause total costs (5.6% [mild], 3.6% [moderate], 14.7% [severe]), and all-cause ED visits represented 0.7% of all-cause total costs (0.4% [mild], 1.2% [moderate], 0.7% [severe]). Proportions similar to the ones presented above for the overall costs were seen when breaking out the overall costs by costs to the payer (Table 4).

Figure 4. Average total healthcare and pharmacy costs (all-cause and psoriasis-related) during the follow-up period (n = 2,130).

When breaking out the overall costs by costs to the patient, all-cause pharmacy costs represented 62.6% of all-cause total costs (70.8% [mild], 65.5% [moderate], 71.4% [severe]) and psoriasis-related pharmacy costs represented 91.5% of psoriasis-related total costs (90.6% [mild], 87.4% [moderate], 91.4% [severe]). Psoriasis-related pharmacy costs represented 80.0% of all-cause pharmacy costs (81.9% [mild], 84.5% [moderate], 85.4% [severe]) and 50.1% of all-cause total costs (58.0% [mild], 55.4% [moderate], 61.0% [severe] (Table 5).

Discussion

Using a large US administrative claims database linked with dermatology EHR data, this study assessed healthcare resource utilization and costs in a cohort of adult psoriasis patients being treated with biologics, overall and by disease severity. Our findings suggest that the economic burden of patients with moderate or severe psoriasis is higher relative to patients with mild disease. Our descriptive analyses revealed that, relative to patients who had mild psoriasis, patients with moderate or severe disease had more all-cause and psoriasis-related outpatient encounters, higher all-cause and psoriasis-related total costs, and higher all-cause and psoriasis-related pharmacy costs. Previous studies conducted in the US and Europe, using administrative claims, survey, or prospective observational data, have shown similar findings of increased healthcare resource utilization and costs with increased disease severity^15,30–34. However, no prior studies have used both a direct measure of disease severity (i.e. defining disease severity based on clinical measures as opposed to a claims-based algorithm that assigns disease severity based on psoriasis therapies received) and adjudicated administrative claims; ours is the first contribution to the literature to do so.

Additionally, we found that pharmacy-related costs were the main driver of total costs (83%), specifically psoriasis-related pharmacy (75%), followed by costs associated with outpatient visits (12%). Notably, while inpatient stays represented only 5% of total costs in the overall biologic cohort, they represented 15% of total costs for patients with severe psoriasis. Pharmacy-related costs (63% and 84%) and outpatient visit costs (32% and 11%) were also the main drivers of total costs for patient and payer costs, respectively. These drivers of cost are slightly different than what has been reported previously. Major drivers of total healthcare costs in psoriasis patients were reported to be medical resource costs (71%)¹⁵, outpatient services (44%)³⁰, and hospitalizations (30%)³³. Prescription drug costs were the primary driver of overall healthcare costs in one study in Spain (47%)³², and were a minor driver in a study in Italy (18%)³³. These differences in findings could be due to the inclusion of patients who were only treated with non-biologic drugs and/or received no treatment for their psoriasis, differences in type of costs assessed, the operational definitions of costs that were used, and/or differences in the healthcare systems and healthcare costs in Spain and Italy as compared to the US. Al Sawah et al.³⁰ also found that total all-cause healthcare costs for psoriasis patients with moderate-to severe disease (as defined by use of any systemic therapy and/or phototherapy) were 2.5-times the total all-cause healthcare costs of those with mild disease, and that all-cause total healthcare costs for psoriasis patients treated with biologics were 3.2-times the total all-cause healthcare costs of those who did not receive biologics. These differences were primarily driven by outpatient pharmacy costs, which accounted for 52% and 63% of total all-cause costs among psoriasis patients with moderate-to-severe disease and among psoriasis patients treated with biologics, respectively.

There are several potential explanations for the increase in healthcare resource utilization and costs with increasing disease severity found in our study. For example, patients classified as having moderate or severe psoriasis may be more likely to need or seek medical attention due to the severity of the disease itself (as seen across other diseases). Alternatively, moderate or severe psoriasis patients may be more likely to experience an inadequate treatment response due to treatment failure or treatment intolerance, leading to increased healthcare resource utilization and costs. Finally, patients with moderate or severe psoriasis may have a greater number and/or severity of comorbidities, leading to increased healthcare resource utilization and costs. While it did not appear that the prevalence of comorbidities differed appreciably between the disease severity groups in our study, it is possible that some comorbidities, such as psoriatic arthritis, were not captured in the 90-day baseline period or 360-day follow-up period or that chronic conditions were not always recorded on an annual basis. Additionally, since ICD-9-CM diagnosis codes do not capture severity, it is possible that moderate or severe patients may have had more severe comorbidities, leading to increased healthcare resource utilization and/or costs. Feldman et al.³⁵ and Kimball et al.³⁶ found that psoriasis patients with comorbidities had higher healthcare resource utilization and costs than psoriasis patients without comorbidities, and Crown et al.³⁷ found that psoriasis patients with select comorbidities had higher total healthcare costs as compared to non-psoriasis patients with the same comorbidities. This suggests that comorbidities and psoriasis may be independent predictors of the increased healthcare resource utilization and costs observed among psoriasis patients with comorbidities. However, it is important to note that 75% of total costs consisted of psoriasis-related pharmacy costs, most likely driven by the relatively high cost of biologics^22,23, and the role that potentially uncaptured comorbidities played in regards to total costs was most likely small-to-moderate.

Finally, we note here that our study does not address the role that indirect costs plays on the economic burden of psoriasis, overall or by disease severity. Previous studies have shown that indirect costs have a significant impact on total healthcare costs, and that they also can increase with greater disease severity^32,38. As a result, we have most likely under-estimated the true economic burden of psoriasis. However, the strength of our study lies in the ability to assign disease severity using EHR data as opposed to prescription claims for treatment received (or other claims-based algorithm) and report healthcare resource utilization and cost data based on adjudicated claims as opposed to surveys, which are subject to low-response rates, recall bias, and inaccurate reporting of healthcare resource utilization and costs due to patient perception.

Limitations

All retrospective database analyses are subject to certain limitations. For example, claims data are collected primarily for payment purposes and are subject to coding errors. Additionally, the data used for this study came from a primarily commercially insured population of patients, a large proportion of whom were less than 65 years old; therefore, the results of this analysis may not be applicable to other populations, such as patients who are uninsured, younger, or covered through Medicaid/Medicare.

Limitations of this analysis include that the study sample consisted of both biologic-experienced and biologic-naïve patients, that the assessment of psoriasis disease severity may not have necessarily coincided with the timing of biologic use, and that the recording of disease severity was performed on a voluntary basis by healthcare providers. Due to the sparseness of the data, disease severity was defined as the first available measure following the index date. This definition precluded our ability to assess temporality, and may have introduced selection bias, where severe patients could have selectively been non-responders of their biologic.

Additionally, there could have been differences between the disease severity groups that we were unable to capture and/or observe through the claims and EHR data, patient demographics and baseline clinical characteristics were not matched or controlled for between the disease severity groups, some of the disease severity groups were very small, no formal statistical tests were performed to compare strata of disease severity, and outliers were not removed. Psoriasis patients with comorbid psoriatic arthritis were not excluded from our study. These patients can have higher healthcare resource utilization and costs compared to psoriasis patients without comorbid psoriatic arthritis^30,35. While less than 2% of our biologic-treated cohort had a diagnosis of psoriatic arthritis, it is possible that we did not capture all cases due to the limitations associated with claims data. This may have had an impact on our findings if there were a significant number of patients with uncaptured (or undiagnosed) psoriatic arthritis in our cohort, and if the proportion of psoriatic arthritis patients differed appreciably across the disease severity cohorts. Finally, we did not address costs associated with a decrease in quality-of-life or lost productivity; therefore, the full economic burden of psoriasis may be under-estimated.

Conclusions

In this study that included the unique linkage of adjudicated medical and pharmacy claims with EHR data, we found that biologic-treated patients with moderate or severe psoriasis appear to cost the healthcare system more than mild psoriasis patients. The cost was primarily driven by higher pharmacy costs and more outpatient encounters. This study highlights the importance of decreasing disease severity in order to potentially decrease the economic burden of psoriasis. For those patients on biologic therapy who remain with moderate-to-severe disease, newer biologic medications with different mechanisms of action and higher clearance levels have the potential to result in a greater proportion of patients with mild psoriasis and, thus, a potential reduction in the economic and clinical burden of psoriasis on the US healthcare system.

Transparency

Declaration of funding

This study was funded by Eli Lilly and Company (Lilly). MJM, CKO, TMM, and ABA are employees of Lilly. AA, SAO, and DC are employees of IQVIA. JFM is an employee of Brigham and Women’s Hospital.

Declaration of financial/other relationships

MJM, TMM, CKO, and ABA are employees and stock owners of Eli Lilly and Company. JFM is a consultant and/or investigator for Eli Lilly, Biogen IDEC, AbbVie, Eli Lilly, Novartis, Pfizer, Incyte, Janssen, UCB, Samumed, Science 37, Celgene, Sanofi Regeneron, Merck, and GSK. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Previous presentations

AMCP Nexus 2017 abstract and poster.

Acknowledgments

The authors would like to acknowledge Magdaliz Gorritz (IQVIA) for her contributions to the programming and data analysis.