Economic impact of expanded use of biologic therapy for the treatment of rheumatoid arthritis and Crohn’s disease in Argentina, Brazil, Colombia, and Mexico

Abstract

Objectives: To estimate economic impact resulting from increased biologics use for treatment of rheumatoid arthritis (RA) and Crohn’s disease (CD) in Argentina, Brazil, Colombia, and Mexico.

Methods: The influence of increasing biologics use for treatment of RA during 2012–2022 and for treatment of CD during 2013–2023 was modeled from a societal perspective. The economic model incorporated current and projected medical, indirect, and drug costs and epidemiologic and economic factors. Costs associated with expanded biologics use for RA were compared with non-expanded use in Argentina, Brazil, Colombia, and Mexico. A similar analysis was conducted for CD in Brazil, Colombia, and Mexico.

Results: Accounting for additional costs of biologics and medical and indirect cost offsets, the model predicts that expanded use of biologics for patients with RA from 2012 to 2022 will result in cumulative net cost savings of ARS$2.351 billion in Argentina, R$9.004 billion in Brazil, COP$728.577 billion in Colombia, and MXN$18.02 billion in Mexico; expanded use of biologics for patients with CD from 2013 to 2023 will result in cumulative net cost savings for patients with CD of R$0.082 billion in Brazil, COP$502.74 billion in Colombia, and MXN$1.80 billion in Mexico. Indirect cost offsets associated with expanded biologics use were a key driver in reducing annual per-patient net costs for RA and CD.

Limitations: Future economic projections are limited by the potential variance between projected and actual future values of biologic prices, wages, medical costs, and gross national product for each country.

Conclusions: Increasing biologics use to treat RA and CD may limit cost growth over time by reducing medical and indirect costs. These findings may inform policy decisions regarding biologics use in Argentina, Brazil, Colombia, and Mexico.

Keywords:

• Rheumatoid arthritis
• Crohn’s disease
• economic assessment
• Argentina
• Brazil
• Colombia
• Mexico

Introduction

Latin America, like other developing regions, has experienced a shift in the allocation of healthcare resources from acute to chronic diseases, as chronic diseases have become the major contributor to morbidity and mortality^1,2. In particular, chronic inflammatory diseases such as rheumatoid arthritis (RA), a systemic inflammatory disease affecting the synovium of joints, tendons, and some extra-articular sites³, and Crohn’s disease (CD), characterized by focal, asymmetric, transmural inflammation of the gastrointestinal tract⁴, are particularly debilitating and pose a large societal economic burden as populations grow and life expectancy increases.

The economic burden of RA and CD is associated primarily with indirect and direct medical costs⁵. Given the disability associated with RA and a typical age of onset during working years, indirect costs of RA may exceed direct costs in some countries⁵. Indirect costs of RA were estimated to be $10.9 billion in 2005 USD$⁶, with individuals with RA being 53% less likely to be employed and spending 3.6-times more days ill in bed compared with individuals without RA⁷. Direct medical costs associated with RA are also substantial and vary by country⁵. The health burden of RA may vary by socioeconomic factors, with low per-capita income countries having a higher burden of disease⁵. Patients with CD experience flares followed by periods of remission. Flares may require urgent medical attention (e.g. hospitalization and surgery) that contributes to the direct costs of the disease, but periods of remission are not necessarily symptom free and unpredictable episodes of pain and abdominal cramping significantly impair quality-of-life⁸. As with RA, many patients with CD are of workforce age, and the symptom burden of CD may lead some patients to work only part-time or leave the workforce entirely⁸, which contributes to the substantial economic burden of the disease.

The prevalence of RA in Latin America ranges from 0.4–1.6%³, and, although prevalence over time is expected to remain constant, a trend toward an earlier age of onset of RA in Latin American countries may have societal cost implications⁹. The economic burden of RA will also increase as the world’s population ages⁵. The published prevalence of CD is 5.65 cases per 100,000 persons in Brazil; however, reliable prevalence data in other Latin American countries are scarce¹⁰, and the current prevalence of CD is likely much higher in Brazil and other Latin American countries. Worldwide, an increasing prevalence of CD seems to be associated with changes in environment and a “Westernized” lifestyle^4,11; for example, CD is more common in Hispanics living in the US than in Latin America¹².

Both non-biologic and biologic disease-modifying anti-rheumatic drugs (DMARDs) have been demonstrated to reduce signs and symptoms of RA, reduce work disability, and prevent joint damage^13,14. New RA treatment guidelines for Argentina, Brazil, Colombia, and Mexico could contribute to increased utilization of biologics in these countries^15–18. For example, the first early arthritis cohort in Argentina, the Argentine Consortium for Early Arthritis (CONAART), has demonstrated that disease activity is the main disease variable associated with work disability in patients with RA¹⁹, and treat-to-target guidelines have been developed to advocate achievement of remission or low disease activity as the treatment goal for patients with RA²⁰. For patients with moderate-to-severe CD, use of immunomodulators and biologics is recommended to first treat acute disease, then induce clinical remission and ultimately maintain response and remission^4,21. Of note, use of biologics, such as tumor necrosis factor (TNF) antagonists, has been associated with improved clinical outcomes and reduced medical and indirect costs for patients with moderate-to-severe RA and CD^22–26.

The economic environment in developing regions is rapidly changing. In Latin America, gross domestic product (GDP) and total healthcare spending are increasing in Argentina, Brazil, Colombia, and Mexico^27–30. As a percentage of GDP, real average wages have increased more than 50% since 2002 in Argentina, Brazil, Colombia, and Mexico^31–34. Rising medical costs and rising wages will likely result in an increased economic burden of RA and CD in these countries^{29,30,33,35–37}.

Although common in Europe, the US, and Canada, access to care and early intervention to treat inflammatory disorders such as RA and CD can be difficult to achieve in Latin America, which has a fragmented healthcare system³. As populations expand and life expectancy increases, Latin American countries with state-run healthcare programs increasingly face decisions about allocation of scarce resources for the optimal management of debilitating chronic inflammatory diseases. Our objective was to project the economic impact from a societal perspective of expanding the use of biologics for the treatment of RA and CD in Argentina, Brazil, Colombia, and Mexico over a 10-year period.

Methods

An economic model was designed to demonstrate the potential cost offsets of expanding biologics use for the treatment of patients with RA and CD in Argentina, Brazil, Colombia, and Mexico. These countries were selected because they are the four countries with the largest populations in the Latin American region³⁸. Seven individual models were developed using the same Excel-based model framework to project the impact of expanded biologics use among patients with RA in Argentina, Brazil, Colombia, and Mexico and among patients with CD in Brazil, Colombia, and Mexico. The impact of expanded biologics use among patients with CD was not modeled for Argentina because of limitations associated with country-specific economic data that were discovered after developing the RA model³⁹. Model inputs were country- and disease-specific based on available data; if country- and/or disease-specific data were not available, we assumed the closest possible scenario. The RA model included the following biologics that were approved for the treatment of moderate-to-severe RA in Argentina, Brazil, Colombia, and Mexico as of 2012: abatacept, adalimumab, certolizumab pegol, etanercept (including a biosimilar version), golimumab, infliximab, rituximab (including a biosimilar version in Mexico), and tocilizumab. The CD model included the following biologics that were approved for the treatment of moderate-to-severe CD as of 2013: adalimumab, certolizumab pegol, and infliximab in Brazil and Mexico and adalimumab and infliximab in Colombia.

The target population for the model included all patients with RA in Argentina, Brazil, Colombia, and Mexico and all patients with CD in Brazil, Colombia, and Mexico. We used a societal perspective to develop the economic model, which compared costs in two scenarios for both diseases over a 10-year time horizon (2012–2022 for RA and 2013–2023 for CD; model development began in 2012 for RA and 2013 for CD). First, we evaluated an expanded biologic use scenario, in which the biologic penetration rate was greater in 2022 for patients with RA and 2023 for patients with CD than corresponding biologic penetration rates in 2012 (for RA) and 2013 (for CD). Second, we evaluated a non-expanded biologic use scenario, in which the biologic penetration rates in 2022 and 2023 were the same as in 2012 and 2013 for RA and CD, respectively.

The cost elements of the model were selected to reflect potential costs as well as cost offsets associated with the treatment of RA and CD. For all patients, costs included direct (non-drug) medical costs, indirect costs, biologic drug costs, and non-biologic drug costs. Potential cost offsets included reductions in direct medical and indirect costs achieved through effective biologic therapy in patients with RA and CD. The primary model inputs were the base annual medical (non-drug) and indirect costs per patient. Average annual per-patient costs of RA or CD therapy as well as cost offsets were incorporated to estimate a total cost per patient. Drug costs were calculated using prices for 1 year of biologic and non-biologic therapy (based on dosing in product labeling and weighted based on product market share) and an average medication adherence rate (dosing information for single drugs that were used to calculate drug costs for RA and CD are provided in Supplementary Tables 1 and 2, respectively). Biologic adherence rates were calculated using a blended adherence rate based on drug-specific adherence rates for those that were included in the model, where adherence was weighted based on the number of patients in the country using drugs. Annual per-patient costs were aggregated to estimate total annual costs among RA and CD patients in each country using country-specific population size and disease prevalence.

Table 1. Components of an economic model designed from a societal perspective to compare costs associated with expanded vs non-expanded biologics use in patients with rheumatoid arthritis in Brazil.

Model component	2012 Estimate	2012–2022 Estimated annual percentage change
Economic burden
Prevalence per 100,000 (%)	0.46^a	Constant^b
Total medical (non-drug) costs ($)	6,023^c^,^d	↑ 13.09^e
Indirect (workloss) costs ($)	8,931^f^,^g	↑ 5.94^h
Drug costs
Biologics ($)	51,329^i	↓ 5.39^j
Non-biologic DMARDs ($)	1,398^k
Macroeconomic factors
Population (n)	196,526,000^l	↑ 0.79^m
GDP ($billions)	4,456^n	↑ 9.41^o
Treatment use
Current biologic use among patients with RA (%)	4.5^p	↑ 11.66^q
Biologic and non-biologic adherence rate (%)	78.0^r	Constant^s
Treat-to-target implementation rate among patients  with RA (%)	5.0^t	↑ 7.18^u
Treatment effects
Medical cost offsets (%)	1.92 for each additional 1% of patients using biologics^v	Function of percentage of patients on biologics
Indirect cost offsets (%)	2.24 for each additional 1% of patients using biologics^v
Treat-to-target cost offsets (%)	19.32^w	Function of percentage of treat-to-target patients

DMARDs, disease-modifying anti-rheumatic drugs; GDP, gross domestic product; RA, rheumatoid arthritis.

Costs are reported as 2012 calendar-year costs for RA in R$.

a Published prevalence⁵⁷.

b No reliable source was found to indicate that prevalence is expected to change over time; therefore, prevalence is assumed to remain constant.

c Non-drug-related medical costs⁴¹ converted and inflated to 2012 costs based on CPI⁴² and adjusted⁴⁰ to account for hospitalization costs.

d Medical costs include costs of outpatient appointments, hospitalizations, diagnostic tests, intra-articular injection sessions, medical devices and aids, and transportation.

e Average annual increase in total nominal per capita health expenditures²⁹.

f Indirect costs in 2005 USD$⁴⁷ converted and inflated to 2012 R$ based on labor-unit costs⁴⁴ and adjusted⁴⁸ to apply only to employed population.

g Indirect costs include the workloss costs of absenteeism, sick leave, and early retirement due to RA, estimated using the human capital approach.

h From Culpepper Salary Budget Survey Data Tables, 2012–2013.

i Biologic drug cost calculation was the cost required for 1 year of treatment based on approved dosing; prices were taken from the National Agency of Sanitary Surveillance (ANVISA) (AbbVie, data on file) and weighted by market share derived from published drug use rates⁵⁸.

j Drug cost change was weighted by market share derived from drug use⁵⁸; price changes from internal data on file (AbbVie Financial Department, data on file) project an average annual decline in prices of biologic DMARDs of 5.39%. This results in a 42.57% decline over 10 years.

k Non-biologic drug costs were the cost required for 1 year of treatment based on approved dosing with prices taken from the ANVISA (AbbVie, data on file) and weighted by market share in DATASUS and IMS Health Database (AbbVie, data on file).

l Projected 2012 population⁵⁹.

m Average annual growth in projected population 2013–2017⁵⁹.

n Model was conducted in nominal currency values and final results were converted to real values. As such, nominal GDP growth is used in the model⁶⁰.

o Average annual growth in projected nominal GDP 2013–2017⁶⁰.

p Biologics use rate for 2012 was calculated based on change in biologic use rates from 2013 to 2014 (AbbVie, data on file).

q Biologics use rate for 2012 was calculated based on change in biologic use rates from 2013 to 2014 (AbbVie, data on file); this 2012 biologics use rate and the expected 2022 biologics use rate were used to calculate annual percentage change.

r AbbVie internal data on file from DATASUS 2012, Brazil Ministry of Health. Data from Brazil were used in the absence of country-specific data for Argentina and Colombia.

s No source was found to indicate that adherence is expected to change over time; as such, adherence is assumed to remain constant.

t Assumption, no published data available.

u A doubling of the treat-to-target implementation rate (from 5.0% to 10.0% in Brazil) requires 7.18% annual growth.

v Cost offsets are the difference in excess RA costs for the pre- and post-biologic eras, accounting for the percentage of patients on biologics using data from Birnbaum et al.²².

w Based on Santos-Moreno et al.⁶¹.

Table 2. Components of an economic model designed from a societal perspective to compare costs associated with expanded vs non-expanded biologics use in patients with Crohn’s disease in Brazil.

Model component	2013 Estimate	2013–2023 Estimated annual percentage change
Economic burden
Prevalence per 100,000 (%)	0.01^a	Constant^i
Total medical (non-drug) costs ($)	5,669^b^,^c	↑ 11.76^j
Indirect (workloss) costs ($)	11,237^c^,^d^,^e^,^f^,^g^,^h	↑ 6.34^k
Drug costs ($)
Biologics ($)	80,727^l	↓ 1.9^m
Non-biologics ($)	4,209^l	↑ 1.59^n
Macroeconomic factors
Population (n)	198,292,000^o	↑ 0.78^p
GDP ($billions)	4,838^q	↑ 8.25^r
Treatment use
Current biologic use among patients with Crohn’s disease (%)	11.4^s	↑ 7.41^t
Biologic and non-biologic adherence rate (%)	66.0^u	Constant^v
Treatment effects
Medical cost offsets ($)	1.83 for each additional 1% of patients using biologics^w	Function of percentage of patients on biologics
Indirect cost offsets ($)	2.32 for each additional 1% of patients using biologics^w

GDP, gross domestic product.

Costs are reported as 2013 calendar-year costs for Crohn’s disease in R$.

a Based on Victoria et al.¹⁰.

b Based on Araujo & Fonseca⁵⁰.

c Based on Morais et al.⁵¹.

d Based on Loomes et al.⁵².

e Based on Sprakes et al.²⁴.

f Based on Longobardi et al.⁵³.

g Based on Binion et al.²⁶.

h Indirect costs include the workloss costs of absenteeism, presenteeism, and unemployment due to Crohn’s disease, estimated using the human capital approach.

i Assumption, no reliable source was found to indicate that prevalence is expected to change over time; as such, prevalence is assumed to remain constant.

j Average annual increase in total nominal per capita health expenditures²⁹.

k Culpepper Salary Budget Survey Data Tables. 2013–2014³⁶.

l Drug costs were the cost required for 1 year of treatment based on approved dosing with prices taken from the National Agency of Sanitary Surveillance [ANVISA] database (AbbVie, data on file) and weighted by market share in DATASUS and IMS Health Database (AbbVie, data on file).

m Drug costs were estimated to decrease 1.90% per year in the base case for biologics (∼17.43% reduction in 10 years).

n Drug costs were estimated to increase 1.59% per year for non-biologics (∼17.09% increase in 10 years).

o Projected 2013 population²⁸.

p Average annual growth in projected population 2014–2019²⁸.

q Projected 2013 GDP current prices, national currency⁶².

r Average annual growth in projected nominal GDP 2014–2019⁶².

s AbbVie, data on file (rationale based on prevalence in adult population, diagnosed, and treated patients).

t AbbVie, data on file (rationale based on prevalence in adult population, diagnosed, and treated patients) show a penetration rate in 2023 of 23.3%. Given that, percentage of patients using biologics was calculated to increase at a rate of 7.41% per year, reaching a rate of 23.3% in 10 years.

u AbbVie, data on file (DATASUS 2013, Brazil Ministry of Health).

v No source was found to indicate that adherence is expected to change over time; as such, adherence is assumed to remain constant.

w Cost offsets are the difference in excess Crohn’s disease costs for the pre- and post-biologic eras, accounting for the percentage of patients on biologics using data from Feagan et al.⁶³, Hay & Hay⁶⁴, Gibson et al.⁶⁵, Sussman et al.²⁵, and Yu et al.⁶⁶.

Additional elements of the model included epidemiologic, economic, and cost factors. In both the expanded and non-expanded biologic use scenarios, the model incorporated projected changes in population size, GDP, personal income, consumer price index (CPI), disease prevalence, medical costs, indirect costs, and drug costs. The model also incorporated projected changes in treat-to-target (T2T) use in the case of RA. Additional benefits (medical and indirect cost reductions) attributed to biologic use were included in both expanded and non-expanded biologic use scenarios.

The specific elements for each input of the model are summarized in Table 1 for RA and Table 2 for CD, using Brazil as an example. In addition, Figure 1 illustrates the net cost calculation for each year. A full description of these elements for the other countries is provided in Tables 3 and 4. The economic burden of each disease was accounted for by disease prevalence and projected growth in prevalence rates, drug costs and projected drug cost changes, and direct and indirect costs and their associated growth rates. Country-specific macroeconomic factors included the total population of each country and the population growth rate; GDP and projected GDP growth rate and CPI growth rates; and unemployment and projected growth in personal income (indirect costs). Treatment use factors included in the model were biologic use and growth of biologic use as well as a treatment adherence rate. The cost offsets attributed to biologic therapy were considered in the model. Patent expiration and its potential impact on prices were taken into account in the price change in biologics over time, where the prices of biologics are assumed to decline over time.

Figure 1. Calculation of net annual costs per patient.

Table 3. Components of an economic model designed from a societal perspective to compare costs associated with expanded vs non-expanded biologics use in patients with rheumatoid arthritis in Argentina, Colombia, and Mexico.

Model component	2012 Estimate by country			2012–2022 Estimated annual percentage change by country
	Argentina	Colombia	Mexico	Argentina	Colombia	Mexico
Economic burden
Prevalence per 100,000 (%)	0.57^a	0.4^b	1.6^c	Constant^d
Total medical (non-drug)costs ($)	12,214^e^,^f	2,870,643^g^,^h	30,673^i^,^j	↑ 21.65^k	↑ 7.21^l	↑ 6.56^l
Indirect (workloss) costs ($)	18,414^m^,^n	7,750,735^o	25,783^p^,^q	↑ 16.78^r	↑ 4.87^r	↑ 4.33^r
Drug costs
Biologics ($)	114,842^s	37,215,049^t	130,176^u	↓ 2.21^v	↓ 1.00^w	0^x
Non-biologic DMARDs ($)	6,352^y	2,072,993^z	7,433^aa
Macroeconomic factors
Population (n)	41,071,972^bb	46,598,000^cc	114,872,000^cc	↑ 0.8^dd	↑ 1.19^ee	↑ 1.0^ee
GDP ($billions)	2,162^ff	470,947^ff	15,501.974^gg	↑ 21.05^hh	↑ 4.43^hh	↑ 6.74^hh
Treatment use
Current biologic use among  patients with RA (%)	3.53^ii	3.0^jj	6.4^kk	↑ 5.63^ll	↑ 12.69^mm	↑ 3.92^kk
Biologic and non-biologic  adherence rate (%)	78.0^nn	78.0^nn	73.0^oo	Constant^pp
Treat-to-target implementa- tion rate among patients  with RA (%)	2.5^qq	5.5^rr	5.0^qq	↑ 7.18^ss
Treatment effects
Medical cost offsets (%)	1.92 for each additional 1% of patients using biologics^tt			Function of percentage of patients on biologics
Indirect cost offsets (%)	2.24 for each additional 1% of patients using biologics^tt
Treat-to-target cost offsets (%)	19.32^uu			Function of percentage of treat-to-target patients

DMARDs, disease-modifying anti-rheumatic drugs; GDP, gross domestic product.

Costs are reported as 2012 calendar-year costs in ARS$, COP$, and MXN$ for Argentina, Colombia, and Mexico, respectively.

a Prevalence calculated as the average of the lowest⁷⁰ and highest⁷¹ published prevalence rates for Argentina.

b Published prevalence⁷².

c Published prevalence⁷³.

d No reliable source was found to indicate that prevalence is expected to change over time; therefore, prevalence is assumed to remain constant.

e Non-drug-related medical costs⁴⁰ converted and inflated to 2012 costs based on medical consumer price index (CPI)³⁷.

f Medical costs include costs of hospitalization, drugs, medical and allied health professional visits, and outpatient procedures.

g Non-drug-related medical costs⁴³ inflated to 2012 costs based on medical CPI⁴⁴.

h Medical costs include orthopedic surgeries, other surgeries, laboratory tests, hospitalizations, medical consults, diagnostic images, other diagnostic procedures, and physiotherapy.

i Non-drug-related medical costs in 2005 USD$⁴⁵ converted and inflated to 2012 MXN$ using medical CPI⁴⁶.

j Medical costs include alternative therapies, laboratory and radiologic tests, hospitalizations, medical aid devices, third-party help, and transportation.

k Average annual increase in total nominal per capita health expenditures²⁹.

l Estimated as the average yearly percentage increase in total health expenditures per person from 2007 to 2011 based on values taken from WHO²⁹.

m Indirect costs in 2002 USD$⁴⁰ converted and inflated to 2012 ARS$ costs based on labor-unit costs³¹.

n Indirect costs include work absenteeism, permanent work disability, and housework lost for housewives, using the human capital approach.

o Indirect costs were estimated in relation to direct costs from Montoya et al.⁴³ using a ratio of indirect to direct costs of 2.7:1⁴⁹.

p Indirect costs in 2005 USD$⁴⁵ converted and inflated to 2012 MXN$ using labor-unit costs³⁴.

q Indirect costs were the costs due to loss of employment.

r From Culpepper Salary Budget Survey Data Tables, 2012–2013³⁵.

s Biologic drug cost calculation was the cost required for 1 year of treatment based on approved dosing; pricing was based on Kairos Database (AbbVie, data on file) and weighted by market share derived from drug use rates in Camoyte Database and IMS Health Database (AbbVie, data on file).

t Biologic drug costs calculation were the costs in 2012 and 2014 required for 1 year of treatment based on approved dosing; prices were taken from the last law of the National Committee of Prices of Drugs and Devices (maximum price authorized for reimbursement; AbbVie, data on file) and weighted by market share derived from drug use rates based on local business intelligence (AbbVie, data on file).

u Biologic drug costs calculation was based on cost required for 1 year of treatment based on approved dosing with prices taken from IMS Health Database (AbbVie, data on file, 2012–2013) and internal data on file (AbbVie) and weighted by market share derived from drug use rates based on internal data on file (AbbVie).

v Drug cost change was calculated as the exponential growth rate based on percentage change in price by 2022 from internal data on file (AbbVie Financial Department, data on file). This results in a projected average annual decline in prices of biologics of 2.21%. The drug cost change for non-biologic DMARDs was conservatively assumed to be the same as that for biologics.

w Average annual costs of biologics in 2014 were estimated to be COP$26,030,250 based on the last law of the National Committee of Prices of Drugs and Devices (maximum price authorized for reimbursement; AbbVie, data on file) and weighted by market share derived from drug use rates based on local business intelligence (AbbVie, data on file). The average annual decline in prices of biologics from 2014 to 2022 and non-biologic DMARDs from 2012 to 2022 were estimated to be 1.00% in the base case based on internal data on file (AbbVie).

x A long-range planning commercial model projects an average annual decline in prices of biologics for RA of 0.00% (AbbVie, data on file). This results in a 0.00% decline over 10 years. The drug cost change for non-biologic DMARDs was conservatively assumed to be the same as that for biologics.

y Non-biologic DMARD costs were the costs required for 1 year of treatment based on approved dosing with prices taken from the Kairos Database (AbbVie, data on file) and weighted by market share⁷⁴ (BIOBADASAR, AbbVie, data on file).

z Non-biologic drug costs were the cost required for 1 year of treatment based on approved dosing with prices taken from SISMED 2012 (AbbVie, data on file) and estimated by the proportion of market share based on local business intelligence (AbbVie, data on file) and published data⁷⁵.

aa Non-biologic drug costs were the cost required for 1 year of treatment based on approved dosing, with prices taken from the IMS Health Database 2013 (AbbVie, data on file) and Farmacia San Pablo (AbbVie, data on file) and weighted by market share based on data from the Ipsos Health Care Survey (AbbVie, data on file).

bb ECLAC⁷⁶.

cc Projected 2012 population³².

dd Average annual growth in projected population 2013–2020. Projected annual population from 2012–2020⁷⁶; population in 2021 and 2022 were projected based on 2020 value using the average annual growth from 2012–2020⁷⁶.

ee Average annual growth in projected population 2013–2018³².

ff Model was conducted in nominal currency values and final results were converted to real values. As such, nominal GDP growth is used in the model³⁹. This model uses officially reported GDP and CPI-GBA data. However, there is uncertainty regarding these indicators in Argentina. The IMF has issued a declaration of censure and called on Argentina to adopt remedial measures to address the quality of the official GDP and CPI-GBA data³⁹.

gg Model was conducted in nominal currency values and final results were converted to real values. As such, nominal GDP growth is used in the model³⁹.

hh Average annual growth in projected nominal GDP 2013–2018³⁹.

ii The total RA population was calculated using an RA prevalence rate of 0.57% (calculated as the average prevalence estimate from Scublinsky et al.⁷¹ and Spindler et al.⁷⁰, and the number of patients taking biologics (AbbVie, data on file) was divided by this number of RA cases to estimate a biologic use rate.

jj Biologics use rates for 2012 and 2022 were estimated based on local business intelligence (AbbVie, data on file).

kk Published biologic use rates⁷⁷.

ll Biologic use rates for 2012 and 2022 as calculated based on prevalence and penetration were used to calculate annual percentage change.

mm Annual percentage change in the biologic use rate was calculated based on the rates derived using analysis of local business intelligence (AbbVie, data on file).

nn AbbVie internal data on file from DATASUS 2012, Brazil Ministry of Health. Data from Brazil were used in the absence of country specific data for Argentina and Colombia.

oo Internal data on file (AbbVie, 2012).

pp No source was found to indicate that adherence is expected to change over time; as such, adherence is assumed to remain constant.

qq Assumption, no published data available.

rr Based on Resultados analisis basal, PID-T2T adherencia del tratate. Fecha de muestreo 26 March 2012 to 13 April 2012.

ss A doubling of the treat-to-target implementation rate (from 2.5–5.0% in Argentina; from 5.0–10.0% in Mexico; and from 5.5–11.0% in Colombia) requires 7.18% annual growth.

tt Cost offsets are the difference in excess RA costs for the pre- and post-biologic eras, accounting for the percentage of patients on biologics using data from Birnbaum et al.²².

uu Based on Santos-Moreno et al.⁶¹.

Table 4. Components of an economic model designed from a societal perspective to compare costs associated with expanded vs non-expanded biologics use in patients with Crohn’s disease in Colombia and Mexico.

Model component	2013 Estimate by country		2013–2023 Estimated annual percentage change by country
	Colombia	Mexico	Colombia	Mexico
Economic burden
Prevalence per 100,000 (%)	0.1^a	0.06^b	Constant^c
Total medical (non-drug) costs ($)	4,624,821^d	31,443^d	↑ 9.55^e	↑ 6.45^f
Indirect (workloss) costs ($)	7,482,899^g	35,853^g	↑ 4.67^h	↑ 4.32^h
Drug costs ($)
Biologics ($)	31,951,536^i	189,972^i	↓ 9.2^b	↓ 5.0^b
Non-biologics ($)	3,095,581^i	32,447^i	↓ 0.96^b	↓ 5.0^b
Macroeconomic factors
Population (n)	47,151,000^j	118,397,000^j	↑ 1.19^j	↑ 1.0^j
GDP ($billions)	713,542.98^j	16,74.112^j	↑ 7.39^j	↑ 6.63^j
Treatment use
Current biologic use among patients with Crohn’s disease (%)	1.7^b	8.0^b	↑ 21.07^k	↑ 4.97^l
Biologic and non-biologic adherence rate (%)	71.0^b	64.5^b	Constant^m	↑ 0.10%^n
Treatment effects
Medical cost offsets ($)	1.83 for each additional 1% of patientsusing biologics^o		Function of percentage of patientson biologics
Indirect cost offsets ($)	2.32 for each additional 1% of patientsusing biologics^o

a Based on de Jesús Yepes Barreto et al.⁷⁸.

b Data on file, AbbVie.

c No reliable source was found to indicate that prevalence is expected to change over time; therefore, prevalence is assumed to remain constant.

d Based on Araujo and Fonseca⁵⁰ and Morais et al.⁵¹.

e Average annual increase in total nominal per capita health expenditures³⁰.

f Average annual increase in total nominal per capita health expenditures²⁹.

g Based on Longobardi et al.⁵³ and Binion et al.²⁶.

h From Culpepper Salary Budget Survey Data Tables, 2013–2014³⁶.

i Biologic drug costs calculation was based on cost required for 1 year of treatment based on approved dosing with prices taken from internal data on file (AbbVie) and weighted by market share derived from drug use rates based on internal data on file (AbbVie).

j Based on IMF^28,62.

k Calculated from use rates for 2013 and 2023 (internal data on file, AbbVie).

l Calculated from AbbVie internal data on file (AbbVie).

m Assumption.

n Calculated from AbbVie internal data on file (AbbVie).

o Cost offsets are the difference in excess Crohn’s disease costs for the pre- and post-biologic eras, accounting for the percentage of patients on biologics using data from Feagan et al.⁶³, Hay and Hay⁶⁴, Gibson et al.⁶⁵, Sussman et al.²⁵, and Yu et al.⁶⁶.

Outputs of the economic model were as follows: annual total cost per patient (including medical [non-drug] costs, indirect costs, and drug costs for RA and CD in each country) and annual total cost for all RA and CD patients in each country as a percentage of GDP. Different sensitivity analyses were run in different models to evaluate the effect of model variables, such as disease prevalence rate, growth in disease prevalence rate, T2T rate (for RA), biologic penetration rate, and medication adherence rate. For example, in Colombia, medication adherence rates were evaluated, but not T2T rates.

In the analysis assessing costs for RA patients in Brazil, the R$51,329 value in Table 1 represents 2012 biologic drug costs; whereas, R$1,979 in Supplemental Table 3 is the discounted value (to 2012) of biologic drug costs in 2022. To relate these two numbers, it is important to consider the 5.39% assumed biologic price decline over time and the increase in biologic use over time. In more detail, the deflated biologic drug cost in 2022 (R$1,979) is the product of medication adherence, average annual cost of biologic drugs per RA patient, and biologic use among all RA patients. In this example, the medication adherence rate was assumed to be 78% for RA patients, and the average annual cost of biologic drugs per RA patient was calculated to be R$51,329 in 2012 and decreasing at a rate of 5.39% per year. As a result, the average annual cost of biologics per RA patient in 2022 was calculated to be R$29,481, which was in turn used to calculate the deflated value (R$1,979). Similarly, the biologic use rate among all RA patients was calculated to be 4.5% per year, resulting in an average annual cost of biologics per RA patient of R$29,481 in 2022, which was then used to calculate the deflated value (R$1,979).

Cost data are reported in Brazilian reals (R$), Argentine pesos (ARS$), Colombian pesos (COP$), and Mexican pesos (MXN$). For RA, cost data are reported in 2012 calendar-year values, and, for CD, cost data are reported in 2013 calendar-year values. When available, application of country-specific discount rates recommended for health technology assessments were used to evaluate the overall return on investment (ROI) for expanded biologics use in each country.

Results

Total annual societal costs for RA and CD in Argentina, Brazil, Colombia, and Mexico are significant. As a percentage of GDP, the current annual costs range from 0.33–0.71% for RA (2012 costs) in Argentina, Brazil, Colombia, and Mexico and from 0.01–0.09% for CD (2013 costs) in Brazil, Colombia, and Mexico. For patients with RA, biologic drug costs account for 9% of total annual costs in Argentina, 11% in Brazil, 7% in Colombia, and 10% in Mexico. Based on the model, biologic drug costs account for 29% of the total annual costs in patients with CD in Brazil, 2.7% in Colombia, and 11% in Mexico.

Our model demonstrated cumulative net cost savings among patients with RA and CD. For patients with RA, the model predicts that expanded use of biologics from 2012 to 2022 will result in a cumulative net cost savings of ARS$2.351 billion in Argentina (Figure 2(a)) (0.518 billion 2012 USD), R$9.004 billion in Brazil (Figure 2(b)) (4.617 billion 2012 USD), COP$728.577 billion in Colombia (Figure 2(c)) (0.405 billion 2012 USD), and MXN$18.02 billion in Mexico (Figure 2(d)) (1.368 billion 2012 USD), accounting for additional cost of biologics and medical cost, indirect cost, and T2T offsets. For CD, expanded use of biologics from 2013 to 2023 will result in a cumulative net cost savings of R$0.082 billion in Brazil (Figure 3(a)) (0.038 billion 2013 USD), COP$502.74 billion in Colombia (Figure 3(b)) (0.269 billion 2013 USD), and MXN$1.80 billion in Mexico (Figure 3(c)) (0.141 billion 2013 USD), accounting for additional cost of biologics and medical and indirect cost offsets. Expressed as ROI, for every additional dollar spent on biologics, the model projects more than 1 dollar of savings through cost offsets (Table 5). Using country-specific discount rates (Brazilian discount rates were used for Argentina) ranging from 3% to 10%, expanded biologic use maintains an ROI of greater than 1 for both disease states (Table 5).

Figure 2. Cumulative net cost savings from 2012–2022 for patients with rheumatoid arthritis. (a) Argentina. (b) Brazil. (c) Colombia. (d) Mexico. All figures present results of a model and are subject to the assumptions and limitations of that model. [1] No annual discount rate was used for results presented in each figure. [2] Indirect costs include work absenteeism, permanent work disability, and housework lost for housewives, using the human capital approach. Medical costs include costs of hospitalization, drugs, medical and allied health professional visits, and outpatient procedures. [3] Indirect costs include the workloss costs of absenteeism, sick leave, and early retirement due to rheumatoid arthritis, estimated using the human capital approach. Medical costs include costs of outpatient appointments, hospitalizations, diagnostic tests, intra-articular injection sessions, devices and aids, and transportation. [4] Indirect costs were estimated in relation to direct costs using a ratio of indirect-to-direct costs of 2.7:1. Medical costs include orthopedic and other surgeries, laboratory tests, hospitalizations, medical consultations, diagnostic imaging and other tests, physiotherapy, and other costs. [5] Indirect costs include the costs due to job loss. Medical costs include costs of alternative therapies, laboratory/radiological tests, hospitalizations, aid devices, third party help, and transportation.

Figure 3. Cumulative net cost savings from 2013–2023 for patients with Crohn’s disease. (a) Brazil. (b) Colombia. (c) Mexico. [1] No annual discount rate was used for results presented in each figure. [2] All figures present results of a model and are subject to the assumptions and limitations of that model. [3] Indirect costs include the workloss costs of absenteeism, presenteeism, and unemployment due to Crohn’s disease, estimated using the human capital approach. Medical costs include costs of outpatient appointments, hospitalizations, and diagnostic tests.

Table 5. Return on investment from expanded biologic use over a 10-year period in patients with rheumatoid arthritis or Crohn’s disease in Argentina, Brazil, Colombia, and Mexico.

Disease state Country Discount rate	Difference in medical cost, indirect cost, and treat-to-target offsets^a(expanded–non-expanded use) [A]	Difference in spending on biologics (expanded–non-expanded use) [B]	Return on investment [A/B]
Rheumatoid arthritis
Argentina: 5.63% annual expansion of biologic use 2012–2022 (ARS$ B)^b
0.0%	3.639	1.287	2.827
5.0%	2.524	0.945	2.670
10.0%	1.802	0.715	2.520
Brazil: 11.66% annual expansion of biologic use 2012–2022 (R$ B)^b
0.0%	16.815	7.811	2.153
5.0%	11.668	5.646	2.067
10.0%	8.336	4.204	1.983
Colombia: 12.69% annual expansion of biologic use 2012–2022 (COP$ B)^c
0.0%	1,729	1,001	1.73
3.0%	1,386	811	1.71
5.0%	1,203	709	1.70
Mexico: 3.92% annual expansion of biologic use 2012–2022 (MXN$ B)^d
0.0%	41.378	23.362	1.771
5.0%	29.175	16.728	1.744
7.0%	25.583	14.760	1.733
Crohn’s disease
Brazil: 7.41% annual expansion in biologic use 2013–2023 (R$ B)^b
0.0%	0.306	0.224	1.368
5.0%	0.215	0.161	1.329
10.0%	0.155	0.120	1.291
Colombia: 21.07% annual expansion of biologic use 2013–2023 (COP$ B)^c
0.0%	687	185	3.71
3.0%	546	150	3.64
5.0%	471	132	3.57
Mexico: 4.97% annual expansion of biologic use 2013–2023 (MXN$ B)^d
0.0%	3.137	1.341	2.339
5.0%	2.209	0.974	2.269
7.0%	1.937	0.864	2.242

a The treat-to-target offsets apply only to the models developed for patients with rheumatoid arthritis.

b Guidelines for economic evaluation studies of health technologies in Brazil suggest a discount rate of 5% with sensitivity analyses of 0% and 10%⁶⁷. Brazilian data were used in absence of country-specific data for Argentina.

c Guidelines for economic evaluation studies of health technologies in Colombia suggest a discount rate of 3% with sensitivity analyses of 0% and 5%⁶⁸.

d Guidelines for economic evaluation studies of health technologies in Mexico suggest a discount rate of 5% with sensitivity analyses of 0% and 7%⁶⁹.

We also analyzed annual total costs of RA and CD patients as a percentage of GDP. For patients with RA, the model suggests that expanded use of biologics will reduce the percentage of GDP spent on patients with RA from 0.3934% to 0.2749% (vs 0.2877% for non-expanded use in Argentina), from 0.3306% to 0.2639% (vs 0.3050%) in Brazil, and from 0.7117% to 0.6233% (vs 0.6439% for non-expanded use) in Mexico. In Colombia, the model indicates that expanded use of biologics can slow the rate of increase in the percentage of GDP spent on patients with RA from 0.4875% to 0.5019% (vs 0.5470%). The impact of expanded biologic use on the percentage of GDP spent on patients with CD was small in Brazil, Colombia, and Mexico.

The model predicts slower growth in annual costs per RA patient in Argentina (Supplementary Figure 1A), Brazil (Supplementary Figure 1B), Colombia (Supplementary Figure 1C), and Mexico (Supplementary Figure 1D) with expanded use of biologics; the cost savings are driven by medical cost, indirect cost, and T2T offsets. Similarly, growth in annual costs per CD patient in Brazil (Supplementary Figure 2A), Colombia (Supplementary Figure 2B), and Mexico (Supplementary Figure 2C) will be lower with expanded use of biologics, with cost savings driven by medical and indirect cost offsets.

The increased annual cost offsets per patient are summarized in Supplemental Table 3. From 2012 to 2022, the reduction in total costs per RA patient associated with expanded biologics use is mainly due to medical and indirect cost offsets. Similarly, from 2013 to 2023, the reduction in total costs per CD patient with expanded use of biologics is due to the offsets in medical and indirect costs.

Various sensitivity analyses (e.g. increased or decreased disease prevalence, increased growth in RA or CD prevalence over the 10-year horizon, higher biologic penetration rates, lower medication adherence rates, higher treat-to-target rates for patients with RA) did not alter the main conclusions. For example, using a biologic penetration rate of 20.0% in 2022 for patients with RA in Brazil instead of the base-case rate of 13.56% resulted in a reduction in projected annual costs per RA patient from BRL $18,760 (base case) to BRL $16,684 for the expanded biologic use scenario and no change in projected annual costs from the base-case scenario for the non-expanded use scenario. Results of this particular sensitivity analysis followed a similar pattern for Argentina, Colombia, and Mexico, although the reduction in projected annual costs per RA patient in Argentina using a 20.0% biologic penetration rate was of a larger magnitude than the relative reduction in costs in Brazil.

Discussion

RA imposes a significant economic burden in Argentina, Brazil, Colombia, and Mexico. For RA, annual 2012 medical costs per patient were estimated to be ARS$12,214 in Argentina^37,40, R$6,023 in Brazil^40–42, COP$2,870,643 in Colombia^43,44, and MXN$30,673 in Mexico^45,46. Annual indirect costs for patients with RA are estimated to be ARS$18,414 in Argentina^31,40, R$8,931 in Brazil^33,47,48, COP$7,750,735 in Colombia^43,44,49 (based on a 2.7:1 ratio of indirect to direct costs), and MXN$25,783 in Mexico^45,46.

Although the prevalence of CD is less than that of RA, CD also imposes a significant economic burden in Argentina, Brazil, Colombia, and Mexico. For CD, annual 2013 medical costs per patient are estimated to be R$5,669 in Brazil, COP$4,624,821 in Colombia, and MXN$31,443 per patient in Mexico^50,51. Annual indirect costs for patients with CD are R$11,237 per patient in Brazil, COP$7,482,899 per patient in Colombia, and MXN$35,853 per patient in Mexico^{24,26,51–53}.

Our economic model demonstrates that expanding biologic use among patients with RA and CD in Argentina, Brazil, Colombia, and Mexico may lead to a slower growth in total costs of RA and CD over a 10-year horizon as compared with not expanding biologic use. Of course, expanding biologic use will increase drug costs, and medical and indirect costs will increase as the economy of each country grows. However, the treatment benefits associated with expanded biologic use will offset the increase in medical and indirect costs that result from economic growth. Overall, the net reduction in medical and indirect costs will offset the drug costs. The key driving factor in annual net cost savings per patient with RA were the indirect cost offsets associated with expanded biologics use in Argentina, Brazil, and Colombia; in Mexico, medical cost offsets accounted for a greater percentage of the net cost savings. The model predicted that the greatest reduction in annual indirect cost offsets for expanded biologics use among patients with RA will be in Colombia (COP$2,042,126 to COP$618,153). For patients with CD, the magnitude of the difference in net annual cost savings per patient with expanded vs non-expanded biologics use was higher in Colombia and Mexico and lower in Brazil compared with model projections for patients with RA. As with RA, the key factor driving that projected annual net cost savings of expanded biologics use in patients with CD was indirect cost offsets, at least in Brazil and Colombia; in Mexico, the indirect and medical cost offsets were similar. According to the model, the greatest net cost savings percentage-wise with expanded biologics use over a 10-year period will be achieved for patients with CD in Colombia, where annual net costs per patient in 2023 will decrease from COP$18.658 M to COP$15.701 M, and the least annual net cost savings percentage-wise will be achieved for patients with CD in Mexico, where the model projects that annual net costs per patient in 2023 will decrease from MXN$82,170 to MXN$76,353.

This economic model-based evaluation has several strengths. To the authors’ knowledge, this is the first published report of an economic evaluation model from the societal perspective that quantifies the economic value of biologic treatment for the serious autoimmune conditions RA and CD in Argentina, Brazil, Colombia, and Mexico over a 10-year period. The model was comprehensive in that it included epidemiologic and economic disease-related inputs, macroeconomic inputs, treatment-use inputs (i.e. medication adherence, treat-to-target strategy for RA), and treatment effects (i.e. cost offsets from biologic therapy). Also, to the best of our ability, we used local country-specific data for all elements of the model, and country-specific discount rates recommended for health technology assessments to calculate a 10-year ROI associated with expanded use of biologics to treat RA and CD in all countries analyzed. The societal perspective of our model provides the broadest view of the potential economic impact of expanding biologic use among patients with disabling inflammatory disorders, and the conclusions of the study did not change with several sensitivity analyses.

The model included estimations of indirect costs, which are particularly important for developing nations with increasing wages and GDPs. Numerous reports have shown that treatment with biologics in both RA and CD patients has also been shown to reduce indirect costs^{22,26,54–56}. Given that the indirect costs of RA and CD are higher than direct medical costs in many countries, the economic burden associated with reduced work productivity is substantial, and our model confirmed that the cost offsets associated with reduced indirect costs are substantial over a 10-year horizon.

In addition, our model incorporated potential cost offsets from the treat-to-target disease management strategy that has been recommended for patients with RA. The treat-to-target goal is to achieve low disease activity and/or remission, and progress toward this goal may result in more patients receiving biologic medications. We felt that inclusion of the treat-to-target strategy for patients with RA was the best method to model dose optimization of biologics.

Despite the strengths of this model, which projected costs associated with expanding biologics use in patients with RA and CD over a 10-year period, all economic models have various limitations and weaknesses. The model used data and assumptions obtained from the published medical literature; data on file at AbbVie; and data published by various institutions, including the World Bank, the World Health Organization, and the International Monetary Fund. The results, therefore, are subject to the limitations of these data sources and must be viewed in the context of the underlying model assumptions. For Argentina, this model uses officially reported GDP and CPI data. However, there is uncertainty regarding the economic data reported by Argentina. Uncertain inflation rates make precise economic projections difficult, and the conclusions of this model may be impacted by major changes in inflation rate. The model is intended to examine project costs within a range of input values, but may not be applicable, for example, if biologic uptake were to expand greatly beyond assumed uptake. Results of the economic model are also subject to the inherent limitations of future economic projections, such as the potential variance between projected and actual future values of biologic prices, wages, medical costs, and GDP for each country.

Another limitation is that use of data from different published studies of patients with RA and CD assumes that the populations across studies were comparable, and that patients in each country in the model are similar to the populations in the studies. However, for a geographically large country such as Brazil, inflammatory disease prevalence varies between the northern and southern regions of the country. In addition, reliable CD prevalence data for Latin American countries are scarce, and prevalence estimates used in the model may under-estimate actual current and future CD prevalence. Even for countries where disease prevalence rates are reported, such as Brazil, the prevalence in northern and southern portions of the country varies greatly, so the local impact of an economic model may differ somewhat from the results presented here.

It should be noted that Argentina, Brazil, Colombia, and Mexico have different healthcare systems with different proportions of the population being publicly insured, privately insured, or uninsured. For example, in Brazil, the majority of the population is covered by an insurance or health plan, whereas, in Mexico and Argentina, a larger proportion of patients are uninsured or have restricted coverage within their health insurance⁷⁹. This can have an impact on access to care, access to specialist care, and access to biologic treatments. To be as comprehensive as possible in our economic analysis, we used model inputs that were country- and disease-specific based on currently available data, and we evaluated expanded vs non-expanded biologic use scenarios in which the biologic penetration rate was varied.

For the RA model, assumptions were made for current and projected treat-to-target prevalence rates and associated cost offsets because these data were not available for Argentina, Brazil, Colombia, and Mexico. However, sensitivity analyses for Argentina, Brazil, and Mexico showed that the results were not sensitive to these inputs. While the treat-to-target strategy was used to model dose optimization of biologics for the treatment of RA, there is no established parallel strategy to model dose optimization of biologics for the treatment of CD. When applying the model to patients with CD in Colombia and Mexico, assumptions were made for current and projected biologic use and associated offsets because these data were not available. In addition, changes in diagnostic rates and treatment practices may affect the future economic and epidemiologic landscape of CD and RA in all Latin American countries.

Our model used a societal perspective to project future costs of expanded vs non-expanded use of biologics in developing regions. Use of a societal perspective is particularly important given the growing populations of developing nations. Also, chronic diseases increasingly affect populations of Latin American countries as life expectancy increases and acute diseases (infections) become less problematic. The expanding GDP and rising wages of Latin American countries are associated with increasingly significant productivity impacts associated with chronic debilitating immune disorders (absenteeism, presenteeism, etc.). Governments of developing nations will increasingly be forced to balance scarce resources with rising healthcare costs.

The 10-year horizon of our model can provide useful ROI projections for government agencies or other payers who are managing limited healthcare resources, and multiple future applications of the model can be envisioned. Biologic medications used to treat RA and CD are indicated for the treatment of several other rheumatologic, dermatologic, and gastrointestinal inflammatory disorders. Our economic model could be adapted as new biologics become available in Argentina, Brazil, Colombia, and Mexico or other developing regions and to predict the impact of expanding biologic use for other inflammatory disorders in Latin America. In addition, the model inputs could be adapted for countries in other developing nations, with availability of reliable public data for various model inputs being the limiting factor.

Conclusions

Increasing biologics use to treat RA and CD may limit cost growth over time by reducing medical and indirect costs. These findings may inform policy decisions regarding biologics use in Argentina, Brazil, Colombia, and Mexico, both for RA and CD and for other serious autoimmune conditions for which biologics are a treatment option.

Transparency

Declaration of funding

This study and all data analyses were funded by AbbVie, Inc. AbbVie authors participated in the study design, the analysis plan, interpretation of the data, preparation review, and approval of the submitted manuscript.

Declaration of financial/other relationships

PGK is a speaker for AbbVie, Janssen-Cilag, Takeda, and Ferring. JRS is a speaker for Abbvie and Pfizer, and declares honoraria received for participating at symposia and advisory boards organized by AbbVie. NT, MS, LC, and YB are employees and stockholders of AbbVie. JC was an employee and stockholder of AbbVie at the time this research was conducted. DM and JS are employees of Analysis Group, which received payment from AbbVie to conduct this research. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

The authors thank Dana L. Randall, MS, PharmD, of Intuitive Graphite Inc., and Eric Bertelsen, PhD, of Arbor Communications, Inc., Ann Arbor, MI, for assistance with manuscript preparation. AbbVie, Inc., North Chicago, IL, provided financial assistance for medical writing and editorial support.