Trends in Patch Testing in the Medicare Part B Fee-for-service Population

Abstract

Background:

Patch testing is a vital component of the work-up for allergic contact dermatitis. There is limited data on changes of patch testing use among Medicare providers, as well as patch testing reimbursement rates.

Objective:

To evaluate trends in the use of patch testing among various Medicare providers and Medicare patch testing reimbursement.

Design:

A longitudinal analysis of patch testing claims was performed with the Medicare part B Physician/Supplier Procedure Summary files from 2010 to 2018. The primary outcomes were the total number and change in number of submitted patch testing services from 2010 to 2018 by three provider groups: dermatology physicians, non-dermatology physicians and non-physician providers. Secondary outcome measures included Medicare reimbursement amounts and changes in reimbursement amounts for patch test services (total and per 1000 enrollees) from 2010 to 2018 for the three provider groups, as well as per patch test service.

Results:

From 2010 to 2018, submitted patch testing services per 1,000 enrollees grew by 89.0%. The annual trend estimate for submitted services relative to 2010 was +10.1% (95% CI 8.1, 12.0) for physicians and +34.1% (95% CI 32.1, 36.0) for non-physician providers (physician assistants and nurse practitioners). Among physicians, the annual trend estimate for submitted services was +5.1% (95% CI −11.3, 21.5) for dermatologists and +31.40% (95% CI 15.00, 47.81) for allergists.

Conclusions:

While patch testing increased in the U.S. Medicare population from 2010–2018, this increase was largely driven by non-physicians providers and allergists.

Introduction

Allergic contact dermatitis (ACD) is prevalent in industrialized nations; a recent systematic review found that 20% of the general population in North America and Europe had at least one identified allergen.¹ Patch testing is an important tool in the diagnosis and subsequent management of patients with ACD. Since the 1995 FDA approval of a prefabricated patch testing kit, Thin-layer Rapid Use Epicutaneous (T.R.U.E.) Test (SmartPractice, Hillerød, Denmark), the process of patch testing has become simplified.^{2, 3} Cheraghlou et al. recently demonstrated the growth of patch test utilization in metro areas of the U.S.⁴ We performed a parallel study to characterize the growth of patch testing in the United States as a whole from 2010 to 2018, comparing trends in patch testing by physician specialty (dermatology, allergy/immunology, and family medicine), trends in patch testing between physicians and non-physician providers (physician assistants and nurse practitioners), and characterizing trends in reimbursement for patch testing.

Methods

Study Design:

This retrospective analysis utilized publicly accessible aggregate information and thus was exempt from institutional review board review. Stemming from anecdotal observations, our primary hypothesis was that the use of patch testing by non-dermatology physicians and non-physician providers (physician assistants and nurse practitioners) increased from 2010–2018 (as compared to dermatologists). Our primary outcome measures were the total number and change in number of submitted patch testing services from 2010 to 2018 by three provider groups: dermatology physicians, non-dermatology physicians and non-physician providers. A patch test service was defined as a bill for a single patch (equivalent to 1 billable unit). Therefore, a patient tested to 36 allergens would equate to 36 submitted services. Unfortunately, patient level data is not available, so it is not possible to determine the number of patches applied per patient in a single patch test session or the number of patients requiring additional testing sessions. Secondary outcome measures included the Medicare reimbursement amounts and changes in reimbursement amounts for patch test services (total and per 1000 enrollees) from 2010 to 2018 for the three provider groups, as well as per patch test service.

Database:

Medicare Part B fee-for-service (FFS) services from 2010–2018 were obtained through the Centers for Medicare and Medicaid Services (CMS) Physician/Supplier Procedure Summary (PSPS) files.⁵ The PSPS file is a summary of Medicare Part B fee-for-service claims for physicians, non-physician providers, and durable medical equipment. It includes aggregated services, charges, and payments based on Healthcare Common Procedure Coding System (HCPCS) codes and provider specialty, as well as other factors (carrier, pricing locality, type of service, and place of service). Level I HCPCS codes are based on the American Medical Association’s Current Procedural Terminology, Fourth Edition (CPT-4). This database has been widely used for procedure utilization analysis in dermatology and other specialties.^6–11

Extracted Data:

We used methodology similar to other studies investigating dermatology procedure utilization in Medicare.⁷ We queried the PSPS files for all services under HCPCS 95044 (patch test), and obtained data on submitted services, denied services, provider code, allowed charges, and amount paid by Medicare (“NCH payments”) from 2010 – 2018. The number of Medicare Part B FFS enrollees is published by CMS.^{12, 13} We excluded denied services from the submitted services to identify the number of CMS-paid reimbursed patch test services in each calendar year. Medicare reimbursement for HCPCS 95044 was defined as the non-facility price amount without additional modifiers in the Medicare Physician Fee Schedule.¹⁴ Payment data (including allowable charges and amount paid by Medicare) were adjusted to 2018 USD using the consumer price index.¹⁵ Submitted services, paid services, and payment data were normalized per 1000 Medicare FFS enrollees for respective years. Provider type was determined by the CMS provider code, which categorizes physicians based on type of specialty (Dermatology-07, Allergy/Immunology-03, Family Practice-08). Nurse practitioners (50) and physician assistants (97) have separate provider codes and are not linked to specialty.¹⁶

Analysis:

We used linear regression to evaluate for the change in submitted services (compared to baseline 2010 values) and the percent of paid reimbursed services (calculated as the ratio of paid reimbursed services to submitted services). Predictor variables included the provider type (dermatology, allergy/Immunology, family practice, physician assistant, nurse practitioner), year (2010–2018), and interaction of these two variables. Pairwise comparisons of the change-over-time for each provider type were corrected with the Tukey method to test if the change over time was different among provider types. Linear regression was also performed for the non-facility price-per-service, using the year as the sole continuous predictor. All P-values were two-sided and considered at the 0.05 level for statistical significance. Analysis was completed in R, Version 3.6.1.¹⁷

Results

Patch Testing Services:

From 2010 through 2018, an average of 1.25 million patch testing services per year were submitted and 1.07 million were reimbursed. Over these nine years, 60.6% of submitted services were from dermatologists, and 39.4% were from non-dermatologist providers (allergy 26.0%, family medicine 0.5%, non-physician providers 9.0%). Submitted services per 1000 Medicare Part B FFS enrollees grew from 26.4 in 2010 to 49.9 in 2018 (an 89.0% increase). There was an increase in submitted services for dermatologists, allergists, and non-physician providers (Figure 1A). However, as a proportion of total submitted services, the share of submitted services from dermatologists decreased from 71.7% in 2010 to 52.9% in 2018. In contrast, the share of submitted services from allergists and non-physician providers increased from 17.6% and 5.8% in 2010 to 32.1% and 11.3% in 2018, respectively (Figure 1B). Given the low proportion of family medicine in relation to other providers, family medicine was not included in our figures.

Figure 1: Submitted services per year from 2010 to 2018.

(A) Submitted services per 1,000 Medicare Part B fee-for-service enrollees for dermatology, allergy, and non-physician providers; (B) Submitted services as a percentage of total submitted services, for dermatology, allergy, and non-physician providers.

Provider Type:

From 2010 to 2018, submitted patch testing services increased significantly for physicians (all specialties combined) as well as non-physician providers. While this increase was observed for both allergy and dermatology (Table 1, Figure 2a), it was greater for allergy as compared to dermatology (31.4 %/year versus 5.1 %/year respectively, p = 0.069). In terms of absolute numbers allergy grew 244% from 2010 to 2018, compared to 39% in dermatology.

Table 1:

Percent change in submitted patch test services relative to 2010. Values are normalized per 1000 Medicare Part B fee-for-service enrollees. Slopes of non-physicians are compared to all physicians, and slopes of family medicine and allergy are compared to dermatology.

Percent Change in Submitted Services (% of 2010 values)
	Annual Trend Estimate (95% CI)	Difference Estimate (95% CI) p-value
Physician vs Non-Physician Providers
All Physicians	10.1 (8.1, 12.0)	NA (Reference)
Non-Physicians (NP/PA)	34.1 (32.1, 36.0)	24.0 (21.3, 26.8) p <0.0001*
Physician Types
Dermatology	5.1 (−11.3, 21.5)	NA (Reference)
Family Medicine	24.3 (7.9, 40.7)	19.2 (−8.9, 47.4) p = 0.220
Allergy	31.4 (15.0, 47.8)	26.3 (−1.8, 54.5) p = 0.069

^“*”superscript indicates statistical significance at the 0.05 level. Pairwise comparisons corrected with Tukey method.

Abbreviations. NA, not applicable; NP, Nurse practitioner; PA, physician assistant.

Figure 2: Percent change in submitted services for patch testing relative to 2010 values, from 2010 to 2018.

(A) Submitted services for physicians including allergy and dermatology; (B) submitted services for all physicians compared to non-physician providers.

The number of submitted services from non-physician providers (including physician assistants and nurse practitioners) increased at a significantly greater rate compared to all physicians (34.1 %/year versus 10.1 %/year respectively, p < 0.0001) (Table 1, Figure 2b). There was no significant increase in submitted patch testing services from family medicine (Table 1).

Payments:

Medicare Part B paid an average of approximately $5.07 million annually for patch testing, of which 61.9% was paid to dermatology, and 38.1% to non-dermatology providers (allergy 26.2%, family medicine 0.4%, non-physician providers 7.5%). Medicare payments for patch testing per 1000 enrollees grew from $123.36 in 2010 to $181.90 in 2018, a 47.5% increase. Non-facility payments for patch testing services decreased from an overall high of $6.83/patch in 2011 to the lowest point of $5.70 in 2014 (Figure 3). The overall decrease from 2010 to 2018 was statistically significant (p = 0.0036). In April 2013, a 2% sequestration of Medicare payments came into effect.¹⁸ This sequestration remained until 2020, following passage of the Coronavirus Aid, Relief, and Economic Security (CARES) Act.¹⁹ Without the sequestration, non-facility payments for patch testing would have been slightly larger (e.g. 2014 payments would have increased to $5.81 per patch), though overall trends have remained (Figure 3).

Figure 3:

Non-facility price-per-service for patch testing services from 2010 to 2018, reported as 2018 USD (value of the U.S. dollar in 2018 to account for inflation). The black line represents actual payments, while the red line represents payments without the 2% sequester from 2013 – 2018. The overall decrease from 2010 to 2018 was statistically significant (p = 0.0036).

Discussion

These results demonstrate that patch testing services overall have increased at a statistically significant and meaningful rate from 2010 to 2018. In that time, patch testing has increased by 89.0% among Medicare Part B FFS enrollees, after controlling for the Medicare Part B population. Factors that may explain this increase include an increasing number of patients being patch tested or an increasing number of patches performed per patient. While the data does not specify which of these factors played a larger role, both may very well account for the increasing numbers of patch testing services provided. Furthermore, with the prevalence of prefabricated T.R.U.E tests, as well as the promotion of ACD education and patch testing by educational institutions, patch testing may be increasing in popularity among providers. While an increase in the aging population may contribute to an increase in the absolute number of patch tests performed in the Medicare population, our analysis normalized services per 1000 enrollees and revealed the above trends, suggesting that additional factors are likely influencing the increasing frequency of patch testing.

While the vast majority of submitted patch testing services were submitted by dermatologists, an increasing proportion came from non-dermatologist providers, especially allergists and non-physician providers, including physician assistants and nurse practitioners. While the difference in growth trends between allergy and dermatology did not meet statistical significance, the confidence interval (−1.8 – 54.5) suggests that a larger sample size may have yielded a statistically significant relationship. In addition, a previous unpublished study demonstrated similar results.²⁰ There are several potential explanations for these trends, including the decreased interest of patch testing by dermatologists, increased interest in patch testing by allergists, as well as financial barriers and time constraints to patch testing for dermatologists.

Although the proportion of medical cases seen by dermatologists have remained largely constant,²¹ there are increasing financial barriers that dermatologists encounter when performing patch testing.²² While the procedure code 95044 has relative value units (RVU) for practice expense fees and malpractice fees, there is no RVU for physician work. This implies an undervaluing of the dermatologist’s time to perform and interpret patch tests, an often time-consuming process.²² In a 2016 survey of members of the American Contact Dermatitis Society (ACDS), 28% of providers reported they were less inclined to perform patch testing due to compensation issues. The prevailing concern was lack of insurance reimbursement, followed by lack of departmental support.²² A previous survey of 2453 dermatologists in 1990 found that 27% of dermatologists did not perform any patch testing, and 54% performed patch testing less than once per week. Reasons cited for not performing patch testing in that study included history being sufficient for diagnosis, testing being too time-consuming, and the lack of adequate reimbursement.²³

The increasing prevalence of non-physician providers including physician assistants and nurse practitioners in dermatology, may also affect trends in patch testing services. An undersupply of dermatologists in the workforce has been evident for over two decades.²⁴ Moreover, dermatologists entering the workforce are heavily centered in urban areas, creating underservice in rural populations.²⁵ This has created opportune roles for non-physician providers, including physician assistants and nurse practitioners. In one study, physician assistants were able to close the dermatologist density gap, especially in less-urban areas.²⁶ The use of physician extenders in dermatology has become more prevalent, increasing from 28% of dermatologists employing physician extenders in 2005, to 46% in 2014. This increase was primarily driven by increased employment of physician assistants.²¹ With the increasing number of non-physician providers in dermatology, the number of patch tests from such providers would be expected to rise concomitantly, as demonstrated by the current study. Unfortunately, in the current study, the dataset used did not allow us to determine whether the non-physician providers, including physician assistants and nurse practitioners, were practicing in a dermatology setting or not. While there are studies of missed diagnoses by non-physician providers compared to dermatologists,²⁷ to our knowledge, no studies have compared the accuracy or outcomes of non-physician providers and dermatologists in the patch test setting.

While patch testing has primarily been considered the domain of dermatologists, our results demonstrate an increasingly greater proportion of patch testing being performed by allergists. These findings are reflected by the number of allergists in the ACDS, which increased from 5% in 2008²⁸ to 15% based on current ACDS data (personal communication, ACDS administrators, January 2020). Furthermore, unlike dermatology residency, patch testing is a procedural component of U.S. allergy fellowships.^{29, 30} It has been theorized that allergists are increasingly assuming the previous role dermatologists provided for patch testing services.³¹ This increase in patch testing by allergists may be related to changes in reimbursement for prick tests (HCPCS 95004), as prick testing has been targeted for reduction by CMS since at least 2014.³² It is likely that providers not trained in dermatology may be less accurate at clinically diagnosing eczematous skin conditions. This could result in an increased use of patch testing for skin conditions for which patch testing is not indicated.³³ Inappropriate use of patch testing is time-consuming, costly, and leads to an increased rate of false-positive results.

The number of increased submitted services over the study period may be due to the increased number of patches on TRUE Test (24 in 1995 to 29 in 2007 to 36 in 2012) and/or the 2013 ACDS recommendation for its expanded Core Series to include 80 allergens.^{34, 35} While there has been an overall increase in both the number of submitted services and Medicare payments per 1000 enrollees, Medicare payments have grown at a slower rate compared to submitted services. As a result, the price-per-patch has decreased from 2010 to 2018. This trend has been reported for other dermatology services in which Medicare paid the same amount or less, despite increased number of services provided.^{7, 36} These trends may also be impacted by the number of medically unlikely edits (MUE; the maximum units of service billable for a given procedure code), currently set at 80 for CPT 95044.³⁷ The MUE limit may also explain the relatively large percentage of submitted services accepted by Medicare (~85.6%), as this would encompass the 36 TRUE test allergens and the 80 ACDS Core Series. While the remaining ~14.4% of submitted services were denied payment by Medicare, it is uncertain why such services were denied. One such possibility is that the number of patches exceeded beyond the 80 MUE limit.

The major limitation of this study was inclusion of only Medicare Part B FFS recipients. While this represents over 33 million enrollees as of 2018 (~10.2% of the United States population), it still only covers a subset of patients.^{12, 38} Additionally, the PSPS dataset only provides the number of individual patches, and does not provide the number of patches tested per session. Future directions include comparing current patch testing trends in the U.S. to other countries, analyzing demographic trends and patch testing utilization in the U.S., or analyzing patch testing trends among non-Medicare beneficiaries.

Conclusion

This analysis of Medicare Part B fee-for-service claims for patch testing from 2010 to 2018 reveals that patch testing services grew in the U.S. Medicare population, while reimbursement declined. While dermatologists have continued to provide the majority of patch testing services, an increasing proportion is being performed by other providers, including allergists and non-physician providers. Several factors may contribute to this trend, including decreased interest in patch testing by dermatologists, increased role of physician assistants and nurse practitioners, and expansion of allergists into the patch testing arena. The changing landscape of diagnosis and management of contact dermatitis underscores the need for comprehensive patch testing training among all providers who perform patch testing.