- Email: [email protected]
Characteristics of patients using specialty medications
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Research in Social and Administrative Pharmacy journal homepage: www.elsevier.com/locate/rsap
Characteristics of patients using specialty medications Taehwan Parka,b,∗, Scott K. Griggsa,b, Paul D. Chungc,1 a b c
Pharmacy Administration, St. Louis College of Pharmacy, St. Louis, MO, 63110, USA Center for Health Outcomes Research and Education, St. Louis College of Pharmacy, St. Louis, MO, 63110, USA Anesthesiology and Perioperative Care Services, Veterans Affairs Palo Alto Health Care System, Palo Alto, CA, 94304, USA
A R T I C L E I N F O
A B S T R A C T
Keywords: Patient characteristics Specialty medication Medical expenditure panel survey
Background: Specialty medications include innovative drugs and biologic agents requiring special handling and close monitoring. Although specialty medications have been widely used for various chronic conditions, increased use of these medications has contributed a growing share of total health care expenditures. Objective: The aim of this study was to examine patient characteristics related to specialty medication use. Methods: Using Medical Expenditure Panel Survey (MEPS) data from 2000 through 2013, this study identified U.S. adults using specialty medications. Andersen's Health Services Utilization model was used to identify potential factors related to specialty medication use. Associations between the variables identified by Andersen's model and specialty medication use were analyzed using logistic multilevel modelling. Sampling weights were considered and standard errors were adjusted to account for the complex survey design. Results: A fully adjusted model suggested that older adults, individuals with prescription drug insurance, or those using mail order services were more likely to use specialty medications regardless of whether they used traditional medications concurrently. Behaviors of using specialty medications were positively associated with married and active working status and negatively associated with middle or high income and having a usual source of care (visiting a doctor's office, clinic, or health center when sick) when comparing individuals using traditional medications and those using specialty medications. In addition, when comparing individuals using traditional medications with those using both specialty medications and traditional medications, behaviors of using specialty medications were positively associated with female gender, worse health state, and more comorbidities. Conclusion: This study identified characteristics of patients using specialty medications. Some sociodemographic, economic, and clinical factors were related to specialty medication use among U.S. adults.
1. Introduction In recent years, specialty pharmaceuticals have attracted increasing attention for their substantial contribution to escalating health care costs. Typically, they are novel medications and biologic agents offering advances in the treatment of complex and life-threatening diseases,1 but at the same time imposing a substantial economic burden driven in part by high development costs, unique management requirements, and a need for careful and ongoing monitoring.2 Accordingly, specialty pharmaceuticals are commonly defined by the following characteristics: (1) high cost (generally a monthly price exceeding $600); (2) special handling, administration, preparation, and distribution requirements; and (3) close monitoring requirements (e.g., personalized or frequent dose adjustment requirements).3 Recent rapid growth in the use of specialty pharmaceuticals has
∗
1
contributed an increasing share to total health care costs. In 2015 specialty medications accounted for 33% of total pharmaceutical spending,4,5 and they are expected to constitute about 50% of pharmaceutical expenditures by 2019.6 Between 2012 and 2014, growth rates were estimated to be 13% to 24% for specialty medications compared to −1.7% to 1.5% for non-specialty (i.e., traditional) medications. Moreover, there has been an increased focus on research and development of specialty medications over the past decade. Consequently, a large number of specialty medications have received approval in recent years. Although only 26% of new drug approvals were specialty medications from 2005 to 2009, this percent of specialty new drug approvals increased to 54% from 2010 to 2014.7 Notably, in 2012 and 2013, specialty medications accounted for about 60 percent or more of new drug approvals.8,9 Despite the growing popularity of specialty pharmaceuticals, little is
Corresponding author. 4588 Parkview Place, St. Louis, MO, 63110, USA. E-mail address: [email protected] (T. Park). Research was conducted while the author (Chung, PD) was at the Washington University School of Medicine.
http://dx.doi.org/10.1016/j.sapharm.2017.10.007 Received 7 June 2017; Received in revised form 15 September 2017; Accepted 12 October 2017 1551-7411/ © 2017 Elsevier Inc. All rights reserved.
Please cite this article as: Park, T., Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.10.007
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Component files as poor/near poor/low income (< 200% FPL), middle income (200% to less than 400% FPL), and high income (≥400% FPL). Type of insurance included any private insurance, public insurance, and uninsured, as specified in the Household Component file. To measure prescription medication insurance status, a binary variable was created using 12 expenditure variables in the Prescribed Medicines file which showed the payment sources of prescription medications. This new variable was able to separate respondents without prescription medication insurance who paid all prescriptions out-of-pocket from those with prescription medication insurance who had payment source(s) other than out-of-pocket payments. To identify a usual source of care, a binary variable was created using the Household Component file. This variable indicated whether respondents visited a doctor's office, clinic, health center, or other place they usually go when sick or needing advice about their health. A binary variable indicating any use of mailorder pharmacy for TMUs, SMUs, and BMUs was also created using the Prescribed Medicines file. Finally, medical need factors included self-reported health status and comorbidity conditions. Self-reported health states were categorized as poor, fair, good, very good, and excellent based on the general health variable in the Household Component file. Comorbidities were estimated by computing the Charlson Comorbidity Index, as suggested by D'Hoore.24 Seventeen comorbidity conditions included in the Charlson Index were identified from the Medical Conditions file using the ICD-9 codes.
known about the characteristics associated with specialty medication use. Previous studies have focused largely on the increase in costs and utilization of specialty medications,10–14 clinical and economic outcomes related to a specialty pharmacy program,15–18 and people's willingness to pay for these medications.19,20 Recently, Hosseini Jebeli et al. investigated socioeconomic factors affecting demand for specialty medications using a survey instrument.21 They found that gender, income, education, and job categories significantly influenced user demand for these medications. This study selected four medical conditions (multiple sclerosis, hemophilia, thalassemia, and chronic kidney disease) in which specialty medications are frequently used in Iran. Because this study focused on socioeconomic factors as predictors for demand for specialty medications, it did not include other key variables such as clinical factors (e.g., individuals' health status and comorbidities) that could potentially affect specialty medication use.21 The objective of the current study was to identify characteristics associated with specialty medication use by employing a large, nationally representative sample of patients using specialty medications for their chronic conditions in the U.S. Comprehensive theory-based variables were considered as potential factors related to specialty medication use. 2. Methods 2.1. Data source Data were obtained from the Medical Expenditure Panel Survey (MEPS), a nationally representative survey sponsored by the Agency for Healthcare Research and Quality and the National Center for Health Statistics. In particular, the MEPS data for the years 2000 through 2013 were employed in this study to include as many individuals as possible who had used specialty medications. The MEPS collects comprehensive data on health care utilization, expenditures, insurance coverage, and sources of payment for the civilian, non-institutionalized U.S. population.22 Such data are collected using an overlapping panel design in which each panel is interviewed for a series of five rounds over 30 months.
2.4. Dependent variables Specialty pharmaceutical users were identified using the medication name variable in the MEPS Prescribed Medicines file. Among our study sample, those who used any specialty pharmaceuticals listed in the Zalesak et al. study3 were considered as specialty pharmaceutical users. Consequently, adults 18 or older who used only pharmaceuticals other than specialty medications were considered traditional medication users (TMUs). Because numerous individuals used not only specialty medications but also traditional medications concurrently, they were labeled as both medication users (BMUs) separately from those who used specialty medications only (SMUs). To verify that each specialty medication was associated with its designated condition, the conditions were manually matched with their indicated specialty medications using the International Classification of Diseases, Ninth Revision (ICD9) code in the MEPS Medical Conditions file for a random sample of 100 individuals. As the study objective was to identify characteristics of patients using specialty medications chronically when traditional medications were also available as a treatment option, specialty pharmaceuticals used for acute disease or supportive care were not included in this current study. Accordingly, traditional pharmaceuticals were not included if they were used for conditions where there were no viable specialty medication substitutes. For the same reason, specialty pharmaceuticals (e.g., orphan drugs) were also not included if there were no viable traditional medication options. Because the primary outcome of interest was use of specialty medications, a binary variable indicating a SMU was created to identify characteristics of SMUs compared with TMUs (i.e., those who used traditional medication only). Similarly, a binary variable indicating a BMU was generated to compare BMUs with TMUs.
2.2. Study sample Our study sample was the MEPS respondents aged 18 or older who received at least one medication during the years 2000 through 2013. 2.3. Independent variables Andersen's Health Services Utilization Model guided the selection of independent variables.23 According to this multilevel model, people's use of health services is a function of three major components – their predisposition to use health services (predisposing factors), factors which enable or impede use (enabling factors), and their need for care (need factors).23 As such, these three factors were considered in this study. Predisposing factors included age, gender, race/ethnicity, marital status, employment status, education, and census region of residence. Age was a continuous variable while race/ethnicity was categorized as white, black, Hispanic, and other. Marital status, employment status, and education were measured by creating binary variables, each of which indicated whether an individual was married, employed, and had college or higher post-graduate education, respectively. Census region of residence was categorized into four regional groups: Northeast, Midwest, South, and West. All the predisposing variables were identified from the MEPS Household Component files. Enabling factors included family income level, type of insurance coverage, prescription medication insurance status, usual source of health care, and use of mail-order service. Family income level was categorized based on the Federal Poverty Level (FPL) in the Household
2.5. Statistical analysis For the selected variables, descriptive statistics were examined by computing means, standard error, and proportions. To identify the characteristics of specialty pharmaceutical users compared with traditional pharmaceutical users, each variable between SMUs and TMUs as well as between BMUs and TMUs was compared using bivariate logistic regression. Multivariate logistic regression in a hierarchical fashion was 2
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Table 1 List of specialty medications. Medical condition
Specialty medications used in SMUs
Frequency in SMUs (%)
Specialty medications used in BMUs
Frequency in BMUs (%)
Total
Cancer
Anastrozole, Fluorouracil, Letrozole Ibandronic acid
66 (28.0%)
2225 (37.9%)
2291 (37.5%)
1567 (26.7%)
1638 (26.8%)
Glatiramer acetate, Interferon beta-1a Adalimumab, Certolizumab pegol, Etanercept Boceprevir, Entecavir, Ribavirin – – –
49 (20.8%)
Anastrozole, Capecitabine, Dasatinib, Exemestane, Fluorouracil, Goserelin, Letrozole, Nilutamide Ibandronic acid, Denosumab, Teriparatide, Zoledronic acid Dalfampridine, Glatiramer acetate, Interferon beta1a, Interferon beta-1b Adalimumab, Certolizumab pegol, Etanercept
768 (13.1%)
817 (13.4%)
568 (9.7%)
604 (9.9%)
211 (3.6%)
225 (3.7%)
318 (5.4%) 115 (2.0%) 78 (1.3%)
318 (5.2%) 115 (1.9%) 78 (1.3%)
–
– 236 (100.0%)
27 (0.5%) 5877 (100.0%)
27 (0.4%) 6113 (100.0%)
Osteoporosis Multiple sclerosis Inflammatory conditions Hepatitis HIV infection Pulmonary heart disease Infertility Asthma Total
71 (30.1%)
36 (15.3%) 14 (5.9%) – – –
Adefovir dipivoxil, Boceprevir, Entecavir, Peginterferon alfa-2a, Ribavirin, Telaprevir Enfuvirtide, Lamivudine Sildenafil Chorionic gonadotropin, Follitropin, Ganirelix acetate, Menotropins Omalizumab
SMUs, Specialty Medication Users; BMU, Both Medication Users.
service use with specialty medication use, as well as negative associations of high family income level and having a usual source of care with specialty medication use. The fully adjusted model suggested that the older adults were more likely to use specialty medications than young adults (OR = 1.05, 95% CI: 1.01–1.10). Married people, the employed, individuals with drug insurance, and those using mail-order service were also more likely to use specialty medications compared with unmarried people, the unemployed, individuals with no drug insurance, and those not using mail-order services, respectively (OR = 3.68, 95% CI: 1.56–8.68; OR = 5.82, 95% CI: 1.40–24.22; OR = 4.27, 95% CI: 1.12–16.33; and OR = 3.41, 95% CI: 1.01–11.60, respectively). However, individuals with middle or high income levels were about 70% less likely to use specialty medications than those with a low income level (OR = 0.30, 95% CI: 0.10–0.89 for the middle income level group and OR = 0.33, 95% CI: 0.13–0.88 for the high income level group, respectively). Individuals with a usual source of care were 65% less likely to use specialty medications compared with those without a usual source of care (OR = 0.35, 95% CI: 0.13–0.93). Table 4 shows the characteristics of individuals using specialty medications in addition to traditional medications (i.e., BMUs) compared with those using only traditional medications (i.e., TMUs). Results from bivariate regression showed that specialty medication use by BMU was associated with numerous variables including older age, female gender, black, Hispanic, and other race/ethnicity, high education level, middle income level, public or no health insurance, prescription drug insurance, usual source of care, mail-order service use, worse health status, and more comorbidities. When predisposing variables were added to the regression model, specialty medication use by BMUs was found to be positively associated with older age, female gender, high education level, and residence in the South, but negatively associated with black and other race/ethnicity. After adding enabling variables, high education, race/ethnicity, and residence in the South were no longer significantly associated with specialty medication use. Enabling variables such as prescription drug insurance and mail-order service use were positively associated with specialty medication use. In the fully adjusted model, older adults were more likely to use specialty medications in addition to traditional medications compared with younger adults (OR = 1.03, 95% CI: 1.02–1.04). Females were about 6 times more likely to use specialty medications than males (OR = 5.84, 95% CI: 3.48–9.80). Individuals having drug insurance and using mailorder services were about three times as likely to use specialty medications as those without drug insurance and not using mail-order services, respectively (OR = 3.02, 95% CI: 1.93–4.72 and OR = 2.92, 95% CI = 1.94–4.38). Finally, individuals with worse health states and more comorbidities were more likely to use specialty medications relative to
also conducted to identify associations between predisposing, enabling, and need variables and specialty medication use behavior. To account for the impact of time, the year variable was added as a covariate in multivariate logistic regressions. All analyses were conducted using Stata 14 (StataCorp, College Station, TX) by accounting for the clustered and stratified survey design and sampling weights in MEPS. This study was reviewed and approved by the institution's Institutional Review Board. 3. Results Table 1 shows specialty pharmaceuticals used by SMUs and BMUs. Overall, these pharmaceuticals were used for the treatment of cancer, osteoporosis, multiple sclerosis, inflammatory disorders, and hepatitis among SMUs and BMUs. BMUs also used their specialty pharmaceuticals for treating human immunodeficiency virus (HIV) infection, pulmonary heart disease, infertility, and asthma. The characteristics of the study sample are displayed in Table 2. A total of 204,283 individuals (weighted sample of 150,513,847) were identified as using at least one prescription medication from the year 2000 through 2013; of the study sample identified, the majority used traditional medications only (n = 202,961, weighted sample of 149,539,994). A total of 1267 individuals (weighted sample of 933,214) used specialty medication in addition to traditional medication, and 55 individuals (weighted sample of 40,640) used specialty medication only. Overall, the mean age was 49.9 (standard error, 0.13) and about 58% were female. More than half of the sample were white (64%), married (57%), fully or partially employed (66%), and college or higher post-graduate education (53%). About 73% had private health insurance and 19% had public health insurance, with 59% having prescription drug insurance. The percentage having a usual source of care (87%) was high; however, the percentage using mailorder service (13%) was low. Approximately 18% of the sample reported their health status as fair or poor. The mean Charlson comorbidity score was 0.7 ± 0.01 (range: 0–32). Table 3 presents the unadjusted and adjusted odds ratios (ORs) obtained from bivariate and multivariate logistic regressions when SMUs were compared with TMUs. Results from the bivariate logistic regressions indicated that active employment status and utilization of mail-order services were positively associated with specialty medication use. When predisposing variables were controlled for in the regression model, older age and active employment status were positively associated with specialty medication use. When enabling factors were also added to the model, results revealed additional positive associations of “married” marital status, having drug insurance, and mail-order 3
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Table 2 Description of the study sample (Unweighted n = 204,283; Weighted N = 150,513,847).a Variables
Weighted mean or weighted percent
Predisposing factors Age (year), mean ± standard error (range) Male, % Race/ethnicity, % White, non-Hispanic Black, non-Hispanic Hispanic Others Married, % Employment: full or part-time, % Education: college or post-graduate, % Region of residence, % Northeast Midwest South West Enabling factors Family income level, % Low Middle High Type of health insurance, % Any private Public Uninsured Prescription drug insurance, % Usual source of care, % Use of mail order service, % Medical Need factors Self-reported health status, % Excellent Very good Good Fair Poor Charlson Comorbidity Index, mean ± standard error (range)
49.9 ± 0.13 (18–90) 41.8 64.2 9.0 8.1 18.7 57.0 66.0 52.5 18.9 23.1 36.8 21.2
27.7 29.9 42.4 73.1 18.7 8.2 58.5 87.3 12.7
13.6 34.9 33.4 14.3 3.9 0.69 ± 0.01 (0–32)
a Includes traditional medication users (unweighted n = 202,961; weighted N = 149,539,994), specialty medication users (n = 55; N = 40,640), and both medication users (n = 1267; N = 933,214). Total N is slightly over the sum of these three Ns because of rounding.
medications for their chronic diseases including inflammatory conditions, cancer, and pulmonary arterial hypertension than adults less than 65 years old. Possibly, older adults may have needed a specialty medication with strong clinical efficacy since their health states were not as good as young adults. Alternatively, older adults were more likely to have prescription drug insurance typically through Medicare Part B or Part D, which may have led their specialty medication use. In addition, having prescription drug insurance was found to be positively correlated with using specialty medications in addition to traditional medications. Economic analysis shows that individuals consume more health care services as insurance covers a greater portion of the costs.25 It is expected that patients with prescription drug insurance were less likely to be sensitive to the price of medications, thereby having relatively low resistance to using high-cost specialty medications. In contrast, patients without prescription drug insurance might have faced considerable cost sharing that limited their access to these costly medications. Not surprisingly, prior studies have shown that high cost sharing was associated with reductions in specialty medication use or initiation.26–28 Notably, this finding implies that high cost sharing may restrict using preferred efficacious therapies, which can result in worsening conditions, more comorbidities, and higher overall health care costs. It raises an affordability and accessibility issue for high-cost medications with superior efficacy, suggesting the challenge of managing patients who need these medications but cannot afford them due to a lack of drug insurance. Another finding of more prevalent specialty medication use among individuals using mail-order services can be explained by the reduced
those with a better health state and less comorbidities, with ORs ranging from 1.40 (95% CI: 1.29–1.53) to 2.85 (95% CI: 1.32–6.15).
4. Discussion This study examined the characteristics of patients using specialty pharmaceuticals compared with those using traditional pharmaceuticals. Results from bivariate analysis suggested that active working status and using mail-order services were found to be positively associated with using specialty medications among SMUs. Factors related to the use of specialty medications in addition to traditional medications among BMUs included: older age, female gender, black, Hispanic, or other race/ethnicity, high education, middle income level, having public or no health insurance, having prescription drug insurance, having a usual source of care, using mail-order services, worse health state, and additional comorbidities. After controlling for predisposing, enabling, and need factors in the multivariate analyses, results revealed that three variables (i.e., older age, prescription drug insurance, and mail-order service use) were positively related to using specialty medications in both SMUs and BMUs. In other words, older adults, individuals with prescription drug insurance, or those using mail order pharmacy services were more likely to use specialty medications regardless of whether they used traditional medications concurrently or not. The first finding of more frequent use of specialty medications among the older adults compared with younger adults was also shown by a previous study.11 In this prior study, older adults (aged 65 and older) consumed more specialty 4
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Table 3 Patient attributes associated with specialty medication use: SMUs (unweighted n = 55; weighted N = 40,640) compared with TMUs (n = 202,961; N = 149,539,994). Unadjusted odds ratio [95% CI]
Predisposing factors Age (year) Gender Male Female Race/ethnicity White, non-Hispanic Black, non-Hispanic Hispanic Other Marital status Unmarried Married Employment Unemployed Full or part-time employed Education High school or less College or post-graduate Region of residence Northeast Midwest South West Enabling factors Family income level Low Middle High Type of health insurance Any private Public Uninsured Prescription drug insurance No Yes Usual source of care No Yes Use of mail order service No Yes Medical Need factors Self-reported health status Excellent Very good Good Fair Poor Charlson Comorbidity Index
Adjusted odds ratio [95% CI] Predisposing variables
Predisposing + Enabling variables
Predisposing + Enabling + Need variables
1.01 [0.99–1.03]
1.04 [1.01–1.07]**
1.05 [1.01–1.10]*
1.05 [1.01–1.10]*
Reference 1.34 [0.58–3.09]
Reference 2.22 [0.84–5.86]
Reference 2.23 [0.79–6.35]
Reference 2.78 [0.88–8.76]
Reference 0.38 [0.10–1.42] 0.95 [0.29–3.09] 0.47 [0.20–1.12]
Reference 0.58 [0.13–2.51] 1.24 [0.32–4.83] 0.80 [0.31–2.08]
Reference 0.78 [0.17–3.60] 1.48 [0.44–4.97] 0.89 [0.31–2.50]
Reference 0.33 [0.04–2.78] 1.49 [0.44–5.05] 0.57 [0.20–1.65]
Reference 2.11 [0.95–4.66]
Reference 2.09 [0.88–4.96]
Reference 3.08 [1.35–7.09]**
Reference 3.68 [1.56–8.68]**
Reference 2.40 [1.02–5.65]*
Reference 5.83 [1.63–20.87]**
Reference 5.23 [1.41–19.39]*
Reference 5.82 [1.40–24.22]*
Reference 2.12 [0.96–4.68]
Reference 1.34 [0.60–3.01]
Reference 1.34 [0.52–3.43]
Reference 1.29 [0.49–3.42]
Reference 1.04 [0.27–3.96] 0.73 [0.20–2.64] 2.01 [0.59–6.80]
Reference 0.80 [0.19–3.31] 0.73 [0.17–3.08] 2.27 [0.59–8.72]
Reference 0.77 [0.18–3.20] 0.68 [0.17–2.73] 1.81 [0.48–6.81]
Reference 0.81 [0.18–3.75] 0.64 [0.14–2.90] 1.78 [0.44–7.19]
Reference 0.96 [0.38–2.40] 0.97 [0.41–2.29]
Reference 0.36 [0.13–1.06] 0.36 [0.14–0.88]*
Reference 0.30 [0.10–0.89]* 0.33 [0.13–0.88]*
Reference 0.76 [0.27–2.16] 0.36 [0.11–1.22]
Reference 0.29 [0.07–1.26] 0.34 [0.06–1.83]
Reference 0.28 [0.07–1.11] 0.38 [0.07–2.12]
Reference 5.24 [1.83–15.02]
Reference 3.48 [1.09–11.10]*
Reference 4.27 [1.12–16.33]*
Reference 0.61 [0.27–1.35]
Reference 0.37 [0.15–0.93]*
Reference 0.35 [0.13–0.93]*
Reference 3.27 [1.45–7.38]**
Reference 4.14 [1.37–12.54]*
Reference 3.41 [1.01–11.60]*
Reference 0.66 [0.24–1.83] 0.98 [0.34–2.84] 0.38 [0.11–1.32] 0.12 [0.01–1.03] 1.01 [0.83–1.24]
Reference 1.47 [0.46–4.73] 2.93 [0.71–12.18] 1.26 [0.22–7.13] 1.65 [0.14–18.85] 1.05 [0.72–1.53]
*p < 0.05, **p < 0.01, ***p < 0.001.
threaten the safety of self-administering such medications as was pointed out by Schwartz et al.31 When comparing SMUs with TMUs, specialty medication use was found to be higher among married and actively working people. Given the scope of this study, the reasons for increased specialty medication use among married and actively working individuals are not fully understood. Future research may further elucidate such reasons. Individuals with higher incomes tended to use specialty medications less, which is consistent with the finding in the Hosseini Jebeli et al. study showing that an increase in income decreased the demand for specialty medications.21 Hosseini Jebeli et al. argued that high income might serve as a proxy for excellent health by reducing stress, concerns, or fear of being unable to afford high-cost medications. Their rationale could likewise apply to the current study. Post-hoc analysis showed that individuals with higher income were in better health states: the proportions in a good, very good, or excellent state (versus a fair or poor
copayment of patients using mail-order services. Mail-order pharmacies have been associated with lower copayments compared to community pharmacies.29 Another possible explanation is that the pharmacy benefit managers (PBMs) regularly negotiate contracts with larger financial discounts for the plan sponsors/employers to channel their business to the PBM-managed mail-order rather than retail pharmacies. Accordingly, the PBMs design their specialty prescription benefit plans with strong incentive (or requirement) for plan participants (i.e., patients) to utilize their mail-order facilities.30 Some specialty pharmaceuticals could have been mail-ordered if patients were able to take their oral medications with no difficulty or had low resistance to self-administer injectables due to their past experience of self-administration. Of note, this finding implies that particular attention should be paid to patients using mail-order services to ensure safety of specialty pharmaceuticals given their limited stability and short shelf life. For example, lack of temperature control of specialty medications during shipping can 5
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Table 4 Patient attributes associated with specialty medication use: BMUs (unweighted n = 1267; weighted N = 933,214) compared with TMUs (n = 202,961; N = 149,539,994). Unadjusted odds ratio [95% CI]
Predisposing factors Age (year) Gender Male Female Race/ethnicity White, non-Hispanic Black, non-Hispanic Hispanic Other Marital status Unmarried Married Employment Unemployed Full or part-time employed Education High school or less College or post-graduate Region of residence Northeast Midwest South West Enabling factors Family income level Low Middle High Type of health insurance Any private Public Uninsured Prescription drug insurance No Yes Usual source of care No Yes Use of mail order service No Yes Medical Need factors Self-reported health status Excellent Very good Good Fair Poor Charlson Comorbidity Index
Adjusted odds ratio [95% CI] Predisposing variables
Predisposing + Enabling variables
Predisposing + Enabling + Need variables
1.04 [1.04–1.04]***
1.03 [1.03–1.04]***
1.04 [1.03–1.05]***
1.03 [1.02–1.04]***
Reference 4.05 [3.25–5.03]***
Reference 4.44 [3.49–5.64]***
Reference 4.51 [2.82–7.22]***
Reference 5.84 [3.48–9.80]***
Reference 0.57 [0.46–0.71]*** 0.47 [0.35–0.63]*** 0.29 [0.21–0.39]***
Reference 0.62 [0.49–0.79]*** 0.78 [0.57–1.05] 0.50 [0.37–0.69]***
Reference 1.02 [0.62–1.68] 1.49 [0.89–2.49] 0.58 [0.30–1.13]
Reference 0.92 [0.55–1.55] 1.43 [0.82–2.52] 0.50 [0.24–1.03]
Reference 0.99 [0.84–1.17]
Reference 1.17 [0.98–1.40]
Reference 1.00 [0.70–1.43]
Reference 1.02 [0.70–1.48]
Reference 0.40 [0.37–0.47]
Reference 0.86 [0.68–1.08]
Reference 1.10 [0.70–1.73]
Reference 1.48 [0.88–2.50]
Reference 1.24 [1.04–1.47]*
Reference 1.34 [1.11–1.61]**
Reference 1.08 [0.75–1.56]
Reference 1.23 [0.84–1.78]
Reference 0.93 [0.71–1.22] 1.23 [0.98–1.53] 1.11 [0.88–1.40]
Reference 0.95 [0.74–1.24] 1.25 [1.00–1.56]* 1.17 [0.93–1.48]
Reference 1.18 [0.70–1.98] 1.06 [0.68–1.68] 0.90 [0.53–1.52]
Reference 1.16 [0.68–1.97] 1.00 [0.63–1.58] 0.82 [0.48–1.39]
Reference 0.81 [0.67–0.98]* 1.03 [0.87–1.24]
Reference 1.28 [0.87–1.89] 1.40 [0.89–2.21]
Reference 1.32 [0.86–2.03] 1.60 [0.96–2.65]
Reference 1.29 [1.08–1.53]** 0.30 [0.19–0.48]***
Reference 1.39 [0.89–2.16] 1.48 [0.67–3.23]
Reference 1.20 [0.74–1.95] 1.29 [0.54–3.11]
Reference 3.87 [2.56–5.86]***
Reference 3.27 [2.14–5.01]***
Reference 3.02 [1.93–4.72]***
Reference 3.04 [2.24–4.14]***
Reference 1.60 [0.88–2.91]
Reference 1.43 [0.74–2.74]
Reference 3.57 [3.02–4.23]***
Reference 2.88 [1.91–4.33]***
Reference 2.92 [1.94–4.38]***
Reference 1.36 [0.99–1.87] 1.95 [1.43–2.66]*** 2.53 [1.83–3.49]*** 3.98 [2.79–5.66]*** 1.33 [1.30–1.37]***
Reference 1.37 [0.75–2.51] 2.44 [1.20–4.98]* 2.85 [1.32–6.15]** 2.77 [1.05–7.33]* 1.40 [1.29–1.53]***
*p < 0.05, **p < 0.01, ***p < 0.001.
specialty medications.21 A possible interpretation for this finding is that many specialty pharmaceuticals identified in this study were used for conditions related to women. Females have conditions such as breast cancer and osteoporosis which require chronic use of specialty medications for these conditions. An inclusion of these conditions in this study could have resulted in more prevalent specialty medication use among females. Patients with worse health states or more comorbidities (regardless of their ages) were highly likely to use more specialty medications in addition to traditional medications. These patients may have used specialty medications with superior therapeutic efficacy to achieve treatment goals that had not been achieved using traditional medications alone. This study has several strengths. To the authors' knowledge, this is the first study to employ a large, nationally-representative dataset to identify patient characteristics associated with specialty medication
state) were 68.0%, 83%, and 91% in the low income, middle income, and high income groups, respectively (p < 0.0001). Given their excellent health state, high income patients might not need more expensive specialty medications as much as those with a worse health state. Patients having a usual source of care were less likely to use specialty medications compared with those without a usual source of care. Although having a usual source of care has been related to the receipt of preventive care or screening services in earlier studies,32,33 its association with prescription drug use has not been explored. Future studies are needed to determine a provider's roles in the use of specialty medications in their patients. When comparing BMUs with TMUs, female gender was found to be a significant predictor for using specialty medications in addition to traditional medications. Hosseini Jebeli et al. have also shown a positive association between female gender and the increased use of 6
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
References
use. Findings from this study underscore the need for appropriate management strategies in using specialty medications with a high incidence of adverse events and compliance issues. For example, older adults or individuals with worse health states and more comorbidities who are more likely to use specialty medications should be closely monitored to ensure that they take these medications as directed. Risks of potential drug-drug interactions need to be avoided or minimized. Pharmacists/clinicians can provide education regarding drug dosing, missed-dose management, frequency of dosing, administration (e.g., self-injection technique), storage and disposal, expected therapeutic effect, adverse events, and strategies for adverse events in using specialty medications.34 Empirical evidence has shown that better management of specialty medication use improves patient health-related outcomes such as increased adherence and reduced health care costs.15–18 In addition, this study employed Andersen's health services utilization model as a theoretical framework. This model has been widely used elsewhere in health service research.35−38 As in these previous studies, the model guided the selection of comprehensive factors potentially associated with health services use – i.e., specialty medication use in this study. Several limitations of this study should be noted. First, because the MEPS data file did not include all specialty pharmaceuticals used chronically, the study sample might not be generalizable to all patients using specialty medications. However, this study included most therapeutic classes, including oncology, osteoporosis, multiple sclerosis, inflammatory conditions, hepatitis, and immune deficiency disorders encompassing the majority of specialty pharmaceutical use. Therefore, there would likely be little impact on the study findings from including additional new samples to the current data. Second, this study had a small sample of SMUs, making the study underpowered to detect significant differences in some characteristics between individuals using specialty medications alone and those using traditional medications. A small sample size of SMUs along with a much larger sample size of BMUs indicates that most patients with chronic diseases were likely to use a combination of specialty and traditional medications in their treatment regimen. To make the findings related to SMUs more robust, more patients using only specialty medications would be needed in a future study. Finally, although covariates were adjusted for in the regression models, residual confounding may nonetheless exist by unmeasured factors. This is always a concern from using secondary data inherent in observational studies. To the extent that any unmeasured characteristics affect outcome measures, the results will be biased. However, this bias is likely to be small, if any, since all the variables addressed in Andersen's model were comprehensively controlled for in this study.
1. Navarro RP, Johnson KA. Opportunities and challenges of specialty pharmaceuticals. J Manag Care Pharm. 2013;19:70–71. 2. Hirsch BR, Balu S, Schulman KA. The impact of specialty pharmaceuticals as drivers of health care costs. Health Aff Millwood. 2014;33:1714–1720. 3. Zalesak M, Greenbaum JS, Cohen JT, et al. The value of specialty pharmaceuticals - a systematic review. Am J Manag Care. 2014;20:461–472. 4. Schumock GT, Li EC, Suda KJ, et al. National trends in prescription drug expenditures and projections for 2014. Am J Health Syst Pharm. 2014;71:482–499. 5. Express Scripts Inc. 2014 Drug Trend Report. St. Louis: Express Scripts; 2015. 6. Express Scripts Inc. 2011 Drug Trend Report. St. Louis: Express Scripts; 2012. 7. PriceWaterhouseCoopers Health Research Institute. Medical Cost Trend: Behind the Number 2016. Available at: https://www.pwc.com/us/en/health-industries/behindthe-numbers/assets/pwc-hri-medical-cost-trend-chart-pack-2016.pdf. Accessed April, 2017. 8. Tharaldson A. Specialty Drug Approvals in 2013. Available at: http://lab.expressscripts.com/lab/insights/drug-options/specialty-drug-approvals-in-2013. Accessed May, 2017. 9. American Journal of Managed Care. The Growing Cost of Specialty Pharmacy - Is it Sustainable? Available at: http://www.ajmc.com/payer-perspectives/0213/thegrowing-cost-of-specialty-pharmacyis-it-sustainable. Accessed May, 2017. 10. Managed Care. Biggest Increases in Specialty Drug Costs Seen in Rheumatologic, Cancer, and Hemophilia Agents. Available at: https://www.managedcaremag.com/ linkout/2005/2/68. Accessed May, 2017. 11. Mullins CD, DeVries AR, Hsu VD, Meng F, Palumbo FB. Variability and growth in spending for outpatient specialty pharmaceuticals. Health Aff Millwood. 2005;24:1117–1127. 12. Jacobs MS, Johnson KA. Curbing the costly trend: exploring the need for a progressive approach to the management of specialty pharmaceuticals under the medical benefit. Am Health Drug Benefits. 2012;5:280–289. 13. Trish E, Joyce G, Goldman DP. Specialty drug spending trends among medicare and medicare advantage enrollees, 2007-11. Health Aff Millwood. 2014;33:2018–2024. 14. Gleason PP, Alexander GC, Starner CI, et al. Health plan utilization and costs of specialty drugs within 4 chronic conditions. J Manag Care Pharm. 2013;19:542–548. 15. Hlubocky JM, Stuckey LJ, Schuman AD, Stevenson JG. Evaluation of a transplantation specialty pharmacy program. Am J Health Syst Pharm. 2012;69:340–347. 16. Murphy P, Cocohoba J, Tang A, Pietrandoni G, Hou J, Guglielmo BJ. Impact of HIVspecialized pharmacies on adherence and persistence with antiretroviral therapy. AIDS Patient Care STDS. 2012;26:526–531. 17. Tschida S, Aslam S, Khan TT, Sahli B, Shrank WH, Lal LS. Managing specialty medication services through a specialty pharmacy program: the case of oral renal transplant immunosuppressant medications. J Manag Care Pharm. 2013;19:26–41. 18. Henderson RR, Visaria J, Bridges GG, Dorholt M, Levin RJ, Frazee SG. Impact of specialty pharmacy on telaprevir-containing 3-drug hepatitis C regimen persistence. J Manag Care Spec Pharm. 2014;20:1227–1234. 19. Goldman DP, Jena AB, Lakdawalla DN, Malin JL, Malkin JD, Sun E. The value of specialty oncology drugs. Health Serv Res. 2010;45:115–132. 20. Romley JA, Sanchez Y, Penrod JR, Goldman DP. Survey results show that adults are willing to pay higher insurance premiums for generous coverage of specialty drugs. Health Aff Millwood. 2012;31:683–690. 21. Hosseini Jebeli SS, Barouni M, Orojloo PH, Mehraban S. Estimating the marginal effect of socioeconomic factors on the demand of specialty drugs. Glob J Health Sci. 2014;7:28–37. 22. Cohen JW, Cohen SB, Banthin JS. The medical expenditure panel survey: a national information resource to support healthcare cost research and inform policy and practice. Med Care. 2009;47(7 Suppl 1):S44–S50. 23. Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav. 1995;36:1–10. 24. D'Hoore W, Bouckaert A, Tilquin C. Practical considerations on the use of the charlson comorbidity index with administrative data bases. J Clin Epidemiol. 1996;49:1429–1433. 25. Pauly MV. The economics of moral hazard: Comment. Am Econ Rev. 1968;58:531–537. 26. Gleason PP, Starner CI, Gunderson BW, Schafer JA, Sarran HS. Association of prescription abandonment with cost share for high-cost specialty pharmacy medications. J Manag Care Pharm. 2009;15:648–658. 27. Doshi JA, Li P, Ladage VP, Pettit AR, Taylor EA. Impact of cost sharing on specialty drug utilization and outcomes: a review of the evidence and future directions. Am J Manag Care. 2016;22:188–197. 28. Kim YA, Rascati KL, Prasla K, Godley P, Goel N, Dunlop D. Retrospective evaluation of the impact of copayment increases for specialty medications on adherence and persistence in an integrated health maintenance organization system. Clin Ther. 2011;33:598–607. 29. Valluri S, Seoane-Vazquez E, Rodriguez-Monguio R, Szeinbach SL. Drug utilization and cost in a medicaid population: a simulation study of community vs. mail order pharmacy. BMC Health Serv Res. 2007;7:122. 30. Pharmaceutical Care Management Association. The Management of Specialty Drugs. Available at: https://www.spcma.org/wp-content/uploads/2016/06/sPCMA_The_ Management_of_Specialty_Drugs.pdf. Accessed April 2017. 31. Schwartz RN, Eng KJ, Frieze DA, et al. NCCN task force report: specialty pharmacy. J Natl Compr Canc Netw. 2010;8(Suppl 4):S1–S12. 32. Ettner SL. The timing of preventive services for women and children: the effect of having a usual source of care. Am J Public Health. 1996;86:1748–1754. 33. DeVoe JE, Fryer GE, Phillips R, Green L. Receipt of preventive care among adults:
5. Conclusions Given the growing popularity of using specialty pharmaceuticals, understanding the characteristics of patients using these medications is essential. This study identified sociodemographic, economic, and clinical factors associated with specialty medication use among U.S. adults. As a next step, appropriate management strategies are needed in using specialty medications with a high incidence of side effects and compliance problems. Attention should be paid to ensure that patients requiring specialty medications actually have access to these medications and use them in an appropriate and safe way. Once identified, individuals receiving specialty medication can receive education and interventions specifically tailored for them by clinicians, thereby improving health outcomes.
Conflicts of interest All authors have nothing to declare. 7
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
2006;17:395–404. 37. Deshpande M, Chewning B, Mott D, Thorpe JM, Young HN. Asthma medication use among U.S. adults 18 and older. Res Soc Adm Pharm. 2014;10:e113–e123. 38. Look KA. Patient characteristics associated with multiple pharmacy use in the U.S. population: findings from the medical expenditure panel survey. Res Soc Adm Pharm. 2015;11:507–516.
insurance status and usual source of care. Am J Public Health. 2003;93:786–791. 34. Mullican KA, Francart SJ. The role of specialty pharmacy drugs in the management of inflammatory diseases. Am J Health Syst Pharm. 2016;73:821–830. 35. Aparasu RR, Mort JR, Brandt H. Polypharmacy trends in office visits by the elderly in the United States, 1990 and 2000. Res Soc Adm Pharm. 2005;1:446–459. 36. Farley JF, Cline RR, Gupta K. Racial variations in antiresorptive medication use: results from the 2000 medical expenditure panel survey (MEPS). Osteoporos Int.
8
Contents lists available at ScienceDirect
Research in Social and Administrative Pharmacy journal homepage: www.elsevier.com/locate/rsap
Characteristics of patients using specialty medications Taehwan Parka,b,∗, Scott K. Griggsa,b, Paul D. Chungc,1 a b c
Pharmacy Administration, St. Louis College of Pharmacy, St. Louis, MO, 63110, USA Center for Health Outcomes Research and Education, St. Louis College of Pharmacy, St. Louis, MO, 63110, USA Anesthesiology and Perioperative Care Services, Veterans Affairs Palo Alto Health Care System, Palo Alto, CA, 94304, USA
A R T I C L E I N F O
A B S T R A C T
Keywords: Patient characteristics Specialty medication Medical expenditure panel survey
Background: Specialty medications include innovative drugs and biologic agents requiring special handling and close monitoring. Although specialty medications have been widely used for various chronic conditions, increased use of these medications has contributed a growing share of total health care expenditures. Objective: The aim of this study was to examine patient characteristics related to specialty medication use. Methods: Using Medical Expenditure Panel Survey (MEPS) data from 2000 through 2013, this study identified U.S. adults using specialty medications. Andersen's Health Services Utilization model was used to identify potential factors related to specialty medication use. Associations between the variables identified by Andersen's model and specialty medication use were analyzed using logistic multilevel modelling. Sampling weights were considered and standard errors were adjusted to account for the complex survey design. Results: A fully adjusted model suggested that older adults, individuals with prescription drug insurance, or those using mail order services were more likely to use specialty medications regardless of whether they used traditional medications concurrently. Behaviors of using specialty medications were positively associated with married and active working status and negatively associated with middle or high income and having a usual source of care (visiting a doctor's office, clinic, or health center when sick) when comparing individuals using traditional medications and those using specialty medications. In addition, when comparing individuals using traditional medications with those using both specialty medications and traditional medications, behaviors of using specialty medications were positively associated with female gender, worse health state, and more comorbidities. Conclusion: This study identified characteristics of patients using specialty medications. Some sociodemographic, economic, and clinical factors were related to specialty medication use among U.S. adults.
1. Introduction In recent years, specialty pharmaceuticals have attracted increasing attention for their substantial contribution to escalating health care costs. Typically, they are novel medications and biologic agents offering advances in the treatment of complex and life-threatening diseases,1 but at the same time imposing a substantial economic burden driven in part by high development costs, unique management requirements, and a need for careful and ongoing monitoring.2 Accordingly, specialty pharmaceuticals are commonly defined by the following characteristics: (1) high cost (generally a monthly price exceeding $600); (2) special handling, administration, preparation, and distribution requirements; and (3) close monitoring requirements (e.g., personalized or frequent dose adjustment requirements).3 Recent rapid growth in the use of specialty pharmaceuticals has
∗
1
contributed an increasing share to total health care costs. In 2015 specialty medications accounted for 33% of total pharmaceutical spending,4,5 and they are expected to constitute about 50% of pharmaceutical expenditures by 2019.6 Between 2012 and 2014, growth rates were estimated to be 13% to 24% for specialty medications compared to −1.7% to 1.5% for non-specialty (i.e., traditional) medications. Moreover, there has been an increased focus on research and development of specialty medications over the past decade. Consequently, a large number of specialty medications have received approval in recent years. Although only 26% of new drug approvals were specialty medications from 2005 to 2009, this percent of specialty new drug approvals increased to 54% from 2010 to 2014.7 Notably, in 2012 and 2013, specialty medications accounted for about 60 percent or more of new drug approvals.8,9 Despite the growing popularity of specialty pharmaceuticals, little is
Corresponding author. 4588 Parkview Place, St. Louis, MO, 63110, USA. E-mail address: [email protected] (T. Park). Research was conducted while the author (Chung, PD) was at the Washington University School of Medicine.
http://dx.doi.org/10.1016/j.sapharm.2017.10.007 Received 7 June 2017; Received in revised form 15 September 2017; Accepted 12 October 2017 1551-7411/ © 2017 Elsevier Inc. All rights reserved.
Please cite this article as: Park, T., Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.10.007
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Component files as poor/near poor/low income (< 200% FPL), middle income (200% to less than 400% FPL), and high income (≥400% FPL). Type of insurance included any private insurance, public insurance, and uninsured, as specified in the Household Component file. To measure prescription medication insurance status, a binary variable was created using 12 expenditure variables in the Prescribed Medicines file which showed the payment sources of prescription medications. This new variable was able to separate respondents without prescription medication insurance who paid all prescriptions out-of-pocket from those with prescription medication insurance who had payment source(s) other than out-of-pocket payments. To identify a usual source of care, a binary variable was created using the Household Component file. This variable indicated whether respondents visited a doctor's office, clinic, health center, or other place they usually go when sick or needing advice about their health. A binary variable indicating any use of mailorder pharmacy for TMUs, SMUs, and BMUs was also created using the Prescribed Medicines file. Finally, medical need factors included self-reported health status and comorbidity conditions. Self-reported health states were categorized as poor, fair, good, very good, and excellent based on the general health variable in the Household Component file. Comorbidities were estimated by computing the Charlson Comorbidity Index, as suggested by D'Hoore.24 Seventeen comorbidity conditions included in the Charlson Index were identified from the Medical Conditions file using the ICD-9 codes.
known about the characteristics associated with specialty medication use. Previous studies have focused largely on the increase in costs and utilization of specialty medications,10–14 clinical and economic outcomes related to a specialty pharmacy program,15–18 and people's willingness to pay for these medications.19,20 Recently, Hosseini Jebeli et al. investigated socioeconomic factors affecting demand for specialty medications using a survey instrument.21 They found that gender, income, education, and job categories significantly influenced user demand for these medications. This study selected four medical conditions (multiple sclerosis, hemophilia, thalassemia, and chronic kidney disease) in which specialty medications are frequently used in Iran. Because this study focused on socioeconomic factors as predictors for demand for specialty medications, it did not include other key variables such as clinical factors (e.g., individuals' health status and comorbidities) that could potentially affect specialty medication use.21 The objective of the current study was to identify characteristics associated with specialty medication use by employing a large, nationally representative sample of patients using specialty medications for their chronic conditions in the U.S. Comprehensive theory-based variables were considered as potential factors related to specialty medication use. 2. Methods 2.1. Data source Data were obtained from the Medical Expenditure Panel Survey (MEPS), a nationally representative survey sponsored by the Agency for Healthcare Research and Quality and the National Center for Health Statistics. In particular, the MEPS data for the years 2000 through 2013 were employed in this study to include as many individuals as possible who had used specialty medications. The MEPS collects comprehensive data on health care utilization, expenditures, insurance coverage, and sources of payment for the civilian, non-institutionalized U.S. population.22 Such data are collected using an overlapping panel design in which each panel is interviewed for a series of five rounds over 30 months.
2.4. Dependent variables Specialty pharmaceutical users were identified using the medication name variable in the MEPS Prescribed Medicines file. Among our study sample, those who used any specialty pharmaceuticals listed in the Zalesak et al. study3 were considered as specialty pharmaceutical users. Consequently, adults 18 or older who used only pharmaceuticals other than specialty medications were considered traditional medication users (TMUs). Because numerous individuals used not only specialty medications but also traditional medications concurrently, they were labeled as both medication users (BMUs) separately from those who used specialty medications only (SMUs). To verify that each specialty medication was associated with its designated condition, the conditions were manually matched with their indicated specialty medications using the International Classification of Diseases, Ninth Revision (ICD9) code in the MEPS Medical Conditions file for a random sample of 100 individuals. As the study objective was to identify characteristics of patients using specialty medications chronically when traditional medications were also available as a treatment option, specialty pharmaceuticals used for acute disease or supportive care were not included in this current study. Accordingly, traditional pharmaceuticals were not included if they were used for conditions where there were no viable specialty medication substitutes. For the same reason, specialty pharmaceuticals (e.g., orphan drugs) were also not included if there were no viable traditional medication options. Because the primary outcome of interest was use of specialty medications, a binary variable indicating a SMU was created to identify characteristics of SMUs compared with TMUs (i.e., those who used traditional medication only). Similarly, a binary variable indicating a BMU was generated to compare BMUs with TMUs.
2.2. Study sample Our study sample was the MEPS respondents aged 18 or older who received at least one medication during the years 2000 through 2013. 2.3. Independent variables Andersen's Health Services Utilization Model guided the selection of independent variables.23 According to this multilevel model, people's use of health services is a function of three major components – their predisposition to use health services (predisposing factors), factors which enable or impede use (enabling factors), and their need for care (need factors).23 As such, these three factors were considered in this study. Predisposing factors included age, gender, race/ethnicity, marital status, employment status, education, and census region of residence. Age was a continuous variable while race/ethnicity was categorized as white, black, Hispanic, and other. Marital status, employment status, and education were measured by creating binary variables, each of which indicated whether an individual was married, employed, and had college or higher post-graduate education, respectively. Census region of residence was categorized into four regional groups: Northeast, Midwest, South, and West. All the predisposing variables were identified from the MEPS Household Component files. Enabling factors included family income level, type of insurance coverage, prescription medication insurance status, usual source of health care, and use of mail-order service. Family income level was categorized based on the Federal Poverty Level (FPL) in the Household
2.5. Statistical analysis For the selected variables, descriptive statistics were examined by computing means, standard error, and proportions. To identify the characteristics of specialty pharmaceutical users compared with traditional pharmaceutical users, each variable between SMUs and TMUs as well as between BMUs and TMUs was compared using bivariate logistic regression. Multivariate logistic regression in a hierarchical fashion was 2
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Table 1 List of specialty medications. Medical condition
Specialty medications used in SMUs
Frequency in SMUs (%)
Specialty medications used in BMUs
Frequency in BMUs (%)
Total
Cancer
Anastrozole, Fluorouracil, Letrozole Ibandronic acid
66 (28.0%)
2225 (37.9%)
2291 (37.5%)
1567 (26.7%)
1638 (26.8%)
Glatiramer acetate, Interferon beta-1a Adalimumab, Certolizumab pegol, Etanercept Boceprevir, Entecavir, Ribavirin – – –
49 (20.8%)
Anastrozole, Capecitabine, Dasatinib, Exemestane, Fluorouracil, Goserelin, Letrozole, Nilutamide Ibandronic acid, Denosumab, Teriparatide, Zoledronic acid Dalfampridine, Glatiramer acetate, Interferon beta1a, Interferon beta-1b Adalimumab, Certolizumab pegol, Etanercept
768 (13.1%)
817 (13.4%)
568 (9.7%)
604 (9.9%)
211 (3.6%)
225 (3.7%)
318 (5.4%) 115 (2.0%) 78 (1.3%)
318 (5.2%) 115 (1.9%) 78 (1.3%)
–
– 236 (100.0%)
27 (0.5%) 5877 (100.0%)
27 (0.4%) 6113 (100.0%)
Osteoporosis Multiple sclerosis Inflammatory conditions Hepatitis HIV infection Pulmonary heart disease Infertility Asthma Total
71 (30.1%)
36 (15.3%) 14 (5.9%) – – –
Adefovir dipivoxil, Boceprevir, Entecavir, Peginterferon alfa-2a, Ribavirin, Telaprevir Enfuvirtide, Lamivudine Sildenafil Chorionic gonadotropin, Follitropin, Ganirelix acetate, Menotropins Omalizumab
SMUs, Specialty Medication Users; BMU, Both Medication Users.
service use with specialty medication use, as well as negative associations of high family income level and having a usual source of care with specialty medication use. The fully adjusted model suggested that the older adults were more likely to use specialty medications than young adults (OR = 1.05, 95% CI: 1.01–1.10). Married people, the employed, individuals with drug insurance, and those using mail-order service were also more likely to use specialty medications compared with unmarried people, the unemployed, individuals with no drug insurance, and those not using mail-order services, respectively (OR = 3.68, 95% CI: 1.56–8.68; OR = 5.82, 95% CI: 1.40–24.22; OR = 4.27, 95% CI: 1.12–16.33; and OR = 3.41, 95% CI: 1.01–11.60, respectively). However, individuals with middle or high income levels were about 70% less likely to use specialty medications than those with a low income level (OR = 0.30, 95% CI: 0.10–0.89 for the middle income level group and OR = 0.33, 95% CI: 0.13–0.88 for the high income level group, respectively). Individuals with a usual source of care were 65% less likely to use specialty medications compared with those without a usual source of care (OR = 0.35, 95% CI: 0.13–0.93). Table 4 shows the characteristics of individuals using specialty medications in addition to traditional medications (i.e., BMUs) compared with those using only traditional medications (i.e., TMUs). Results from bivariate regression showed that specialty medication use by BMU was associated with numerous variables including older age, female gender, black, Hispanic, and other race/ethnicity, high education level, middle income level, public or no health insurance, prescription drug insurance, usual source of care, mail-order service use, worse health status, and more comorbidities. When predisposing variables were added to the regression model, specialty medication use by BMUs was found to be positively associated with older age, female gender, high education level, and residence in the South, but negatively associated with black and other race/ethnicity. After adding enabling variables, high education, race/ethnicity, and residence in the South were no longer significantly associated with specialty medication use. Enabling variables such as prescription drug insurance and mail-order service use were positively associated with specialty medication use. In the fully adjusted model, older adults were more likely to use specialty medications in addition to traditional medications compared with younger adults (OR = 1.03, 95% CI: 1.02–1.04). Females were about 6 times more likely to use specialty medications than males (OR = 5.84, 95% CI: 3.48–9.80). Individuals having drug insurance and using mailorder services were about three times as likely to use specialty medications as those without drug insurance and not using mail-order services, respectively (OR = 3.02, 95% CI: 1.93–4.72 and OR = 2.92, 95% CI = 1.94–4.38). Finally, individuals with worse health states and more comorbidities were more likely to use specialty medications relative to
also conducted to identify associations between predisposing, enabling, and need variables and specialty medication use behavior. To account for the impact of time, the year variable was added as a covariate in multivariate logistic regressions. All analyses were conducted using Stata 14 (StataCorp, College Station, TX) by accounting for the clustered and stratified survey design and sampling weights in MEPS. This study was reviewed and approved by the institution's Institutional Review Board. 3. Results Table 1 shows specialty pharmaceuticals used by SMUs and BMUs. Overall, these pharmaceuticals were used for the treatment of cancer, osteoporosis, multiple sclerosis, inflammatory disorders, and hepatitis among SMUs and BMUs. BMUs also used their specialty pharmaceuticals for treating human immunodeficiency virus (HIV) infection, pulmonary heart disease, infertility, and asthma. The characteristics of the study sample are displayed in Table 2. A total of 204,283 individuals (weighted sample of 150,513,847) were identified as using at least one prescription medication from the year 2000 through 2013; of the study sample identified, the majority used traditional medications only (n = 202,961, weighted sample of 149,539,994). A total of 1267 individuals (weighted sample of 933,214) used specialty medication in addition to traditional medication, and 55 individuals (weighted sample of 40,640) used specialty medication only. Overall, the mean age was 49.9 (standard error, 0.13) and about 58% were female. More than half of the sample were white (64%), married (57%), fully or partially employed (66%), and college or higher post-graduate education (53%). About 73% had private health insurance and 19% had public health insurance, with 59% having prescription drug insurance. The percentage having a usual source of care (87%) was high; however, the percentage using mailorder service (13%) was low. Approximately 18% of the sample reported their health status as fair or poor. The mean Charlson comorbidity score was 0.7 ± 0.01 (range: 0–32). Table 3 presents the unadjusted and adjusted odds ratios (ORs) obtained from bivariate and multivariate logistic regressions when SMUs were compared with TMUs. Results from the bivariate logistic regressions indicated that active employment status and utilization of mail-order services were positively associated with specialty medication use. When predisposing variables were controlled for in the regression model, older age and active employment status were positively associated with specialty medication use. When enabling factors were also added to the model, results revealed additional positive associations of “married” marital status, having drug insurance, and mail-order 3
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Table 2 Description of the study sample (Unweighted n = 204,283; Weighted N = 150,513,847).a Variables
Weighted mean or weighted percent
Predisposing factors Age (year), mean ± standard error (range) Male, % Race/ethnicity, % White, non-Hispanic Black, non-Hispanic Hispanic Others Married, % Employment: full or part-time, % Education: college or post-graduate, % Region of residence, % Northeast Midwest South West Enabling factors Family income level, % Low Middle High Type of health insurance, % Any private Public Uninsured Prescription drug insurance, % Usual source of care, % Use of mail order service, % Medical Need factors Self-reported health status, % Excellent Very good Good Fair Poor Charlson Comorbidity Index, mean ± standard error (range)
49.9 ± 0.13 (18–90) 41.8 64.2 9.0 8.1 18.7 57.0 66.0 52.5 18.9 23.1 36.8 21.2
27.7 29.9 42.4 73.1 18.7 8.2 58.5 87.3 12.7
13.6 34.9 33.4 14.3 3.9 0.69 ± 0.01 (0–32)
a Includes traditional medication users (unweighted n = 202,961; weighted N = 149,539,994), specialty medication users (n = 55; N = 40,640), and both medication users (n = 1267; N = 933,214). Total N is slightly over the sum of these three Ns because of rounding.
medications for their chronic diseases including inflammatory conditions, cancer, and pulmonary arterial hypertension than adults less than 65 years old. Possibly, older adults may have needed a specialty medication with strong clinical efficacy since their health states were not as good as young adults. Alternatively, older adults were more likely to have prescription drug insurance typically through Medicare Part B or Part D, which may have led their specialty medication use. In addition, having prescription drug insurance was found to be positively correlated with using specialty medications in addition to traditional medications. Economic analysis shows that individuals consume more health care services as insurance covers a greater portion of the costs.25 It is expected that patients with prescription drug insurance were less likely to be sensitive to the price of medications, thereby having relatively low resistance to using high-cost specialty medications. In contrast, patients without prescription drug insurance might have faced considerable cost sharing that limited their access to these costly medications. Not surprisingly, prior studies have shown that high cost sharing was associated with reductions in specialty medication use or initiation.26–28 Notably, this finding implies that high cost sharing may restrict using preferred efficacious therapies, which can result in worsening conditions, more comorbidities, and higher overall health care costs. It raises an affordability and accessibility issue for high-cost medications with superior efficacy, suggesting the challenge of managing patients who need these medications but cannot afford them due to a lack of drug insurance. Another finding of more prevalent specialty medication use among individuals using mail-order services can be explained by the reduced
those with a better health state and less comorbidities, with ORs ranging from 1.40 (95% CI: 1.29–1.53) to 2.85 (95% CI: 1.32–6.15).
4. Discussion This study examined the characteristics of patients using specialty pharmaceuticals compared with those using traditional pharmaceuticals. Results from bivariate analysis suggested that active working status and using mail-order services were found to be positively associated with using specialty medications among SMUs. Factors related to the use of specialty medications in addition to traditional medications among BMUs included: older age, female gender, black, Hispanic, or other race/ethnicity, high education, middle income level, having public or no health insurance, having prescription drug insurance, having a usual source of care, using mail-order services, worse health state, and additional comorbidities. After controlling for predisposing, enabling, and need factors in the multivariate analyses, results revealed that three variables (i.e., older age, prescription drug insurance, and mail-order service use) were positively related to using specialty medications in both SMUs and BMUs. In other words, older adults, individuals with prescription drug insurance, or those using mail order pharmacy services were more likely to use specialty medications regardless of whether they used traditional medications concurrently or not. The first finding of more frequent use of specialty medications among the older adults compared with younger adults was also shown by a previous study.11 In this prior study, older adults (aged 65 and older) consumed more specialty 4
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Table 3 Patient attributes associated with specialty medication use: SMUs (unweighted n = 55; weighted N = 40,640) compared with TMUs (n = 202,961; N = 149,539,994). Unadjusted odds ratio [95% CI]
Predisposing factors Age (year) Gender Male Female Race/ethnicity White, non-Hispanic Black, non-Hispanic Hispanic Other Marital status Unmarried Married Employment Unemployed Full or part-time employed Education High school or less College or post-graduate Region of residence Northeast Midwest South West Enabling factors Family income level Low Middle High Type of health insurance Any private Public Uninsured Prescription drug insurance No Yes Usual source of care No Yes Use of mail order service No Yes Medical Need factors Self-reported health status Excellent Very good Good Fair Poor Charlson Comorbidity Index
Adjusted odds ratio [95% CI] Predisposing variables
Predisposing + Enabling variables
Predisposing + Enabling + Need variables
1.01 [0.99–1.03]
1.04 [1.01–1.07]**
1.05 [1.01–1.10]*
1.05 [1.01–1.10]*
Reference 1.34 [0.58–3.09]
Reference 2.22 [0.84–5.86]
Reference 2.23 [0.79–6.35]
Reference 2.78 [0.88–8.76]
Reference 0.38 [0.10–1.42] 0.95 [0.29–3.09] 0.47 [0.20–1.12]
Reference 0.58 [0.13–2.51] 1.24 [0.32–4.83] 0.80 [0.31–2.08]
Reference 0.78 [0.17–3.60] 1.48 [0.44–4.97] 0.89 [0.31–2.50]
Reference 0.33 [0.04–2.78] 1.49 [0.44–5.05] 0.57 [0.20–1.65]
Reference 2.11 [0.95–4.66]
Reference 2.09 [0.88–4.96]
Reference 3.08 [1.35–7.09]**
Reference 3.68 [1.56–8.68]**
Reference 2.40 [1.02–5.65]*
Reference 5.83 [1.63–20.87]**
Reference 5.23 [1.41–19.39]*
Reference 5.82 [1.40–24.22]*
Reference 2.12 [0.96–4.68]
Reference 1.34 [0.60–3.01]
Reference 1.34 [0.52–3.43]
Reference 1.29 [0.49–3.42]
Reference 1.04 [0.27–3.96] 0.73 [0.20–2.64] 2.01 [0.59–6.80]
Reference 0.80 [0.19–3.31] 0.73 [0.17–3.08] 2.27 [0.59–8.72]
Reference 0.77 [0.18–3.20] 0.68 [0.17–2.73] 1.81 [0.48–6.81]
Reference 0.81 [0.18–3.75] 0.64 [0.14–2.90] 1.78 [0.44–7.19]
Reference 0.96 [0.38–2.40] 0.97 [0.41–2.29]
Reference 0.36 [0.13–1.06] 0.36 [0.14–0.88]*
Reference 0.30 [0.10–0.89]* 0.33 [0.13–0.88]*
Reference 0.76 [0.27–2.16] 0.36 [0.11–1.22]
Reference 0.29 [0.07–1.26] 0.34 [0.06–1.83]
Reference 0.28 [0.07–1.11] 0.38 [0.07–2.12]
Reference 5.24 [1.83–15.02]
Reference 3.48 [1.09–11.10]*
Reference 4.27 [1.12–16.33]*
Reference 0.61 [0.27–1.35]
Reference 0.37 [0.15–0.93]*
Reference 0.35 [0.13–0.93]*
Reference 3.27 [1.45–7.38]**
Reference 4.14 [1.37–12.54]*
Reference 3.41 [1.01–11.60]*
Reference 0.66 [0.24–1.83] 0.98 [0.34–2.84] 0.38 [0.11–1.32] 0.12 [0.01–1.03] 1.01 [0.83–1.24]
Reference 1.47 [0.46–4.73] 2.93 [0.71–12.18] 1.26 [0.22–7.13] 1.65 [0.14–18.85] 1.05 [0.72–1.53]
*p < 0.05, **p < 0.01, ***p < 0.001.
threaten the safety of self-administering such medications as was pointed out by Schwartz et al.31 When comparing SMUs with TMUs, specialty medication use was found to be higher among married and actively working people. Given the scope of this study, the reasons for increased specialty medication use among married and actively working individuals are not fully understood. Future research may further elucidate such reasons. Individuals with higher incomes tended to use specialty medications less, which is consistent with the finding in the Hosseini Jebeli et al. study showing that an increase in income decreased the demand for specialty medications.21 Hosseini Jebeli et al. argued that high income might serve as a proxy for excellent health by reducing stress, concerns, or fear of being unable to afford high-cost medications. Their rationale could likewise apply to the current study. Post-hoc analysis showed that individuals with higher income were in better health states: the proportions in a good, very good, or excellent state (versus a fair or poor
copayment of patients using mail-order services. Mail-order pharmacies have been associated with lower copayments compared to community pharmacies.29 Another possible explanation is that the pharmacy benefit managers (PBMs) regularly negotiate contracts with larger financial discounts for the plan sponsors/employers to channel their business to the PBM-managed mail-order rather than retail pharmacies. Accordingly, the PBMs design their specialty prescription benefit plans with strong incentive (or requirement) for plan participants (i.e., patients) to utilize their mail-order facilities.30 Some specialty pharmaceuticals could have been mail-ordered if patients were able to take their oral medications with no difficulty or had low resistance to self-administer injectables due to their past experience of self-administration. Of note, this finding implies that particular attention should be paid to patients using mail-order services to ensure safety of specialty pharmaceuticals given their limited stability and short shelf life. For example, lack of temperature control of specialty medications during shipping can 5
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
Table 4 Patient attributes associated with specialty medication use: BMUs (unweighted n = 1267; weighted N = 933,214) compared with TMUs (n = 202,961; N = 149,539,994). Unadjusted odds ratio [95% CI]
Predisposing factors Age (year) Gender Male Female Race/ethnicity White, non-Hispanic Black, non-Hispanic Hispanic Other Marital status Unmarried Married Employment Unemployed Full or part-time employed Education High school or less College or post-graduate Region of residence Northeast Midwest South West Enabling factors Family income level Low Middle High Type of health insurance Any private Public Uninsured Prescription drug insurance No Yes Usual source of care No Yes Use of mail order service No Yes Medical Need factors Self-reported health status Excellent Very good Good Fair Poor Charlson Comorbidity Index
Adjusted odds ratio [95% CI] Predisposing variables
Predisposing + Enabling variables
Predisposing + Enabling + Need variables
1.04 [1.04–1.04]***
1.03 [1.03–1.04]***
1.04 [1.03–1.05]***
1.03 [1.02–1.04]***
Reference 4.05 [3.25–5.03]***
Reference 4.44 [3.49–5.64]***
Reference 4.51 [2.82–7.22]***
Reference 5.84 [3.48–9.80]***
Reference 0.57 [0.46–0.71]*** 0.47 [0.35–0.63]*** 0.29 [0.21–0.39]***
Reference 0.62 [0.49–0.79]*** 0.78 [0.57–1.05] 0.50 [0.37–0.69]***
Reference 1.02 [0.62–1.68] 1.49 [0.89–2.49] 0.58 [0.30–1.13]
Reference 0.92 [0.55–1.55] 1.43 [0.82–2.52] 0.50 [0.24–1.03]
Reference 0.99 [0.84–1.17]
Reference 1.17 [0.98–1.40]
Reference 1.00 [0.70–1.43]
Reference 1.02 [0.70–1.48]
Reference 0.40 [0.37–0.47]
Reference 0.86 [0.68–1.08]
Reference 1.10 [0.70–1.73]
Reference 1.48 [0.88–2.50]
Reference 1.24 [1.04–1.47]*
Reference 1.34 [1.11–1.61]**
Reference 1.08 [0.75–1.56]
Reference 1.23 [0.84–1.78]
Reference 0.93 [0.71–1.22] 1.23 [0.98–1.53] 1.11 [0.88–1.40]
Reference 0.95 [0.74–1.24] 1.25 [1.00–1.56]* 1.17 [0.93–1.48]
Reference 1.18 [0.70–1.98] 1.06 [0.68–1.68] 0.90 [0.53–1.52]
Reference 1.16 [0.68–1.97] 1.00 [0.63–1.58] 0.82 [0.48–1.39]
Reference 0.81 [0.67–0.98]* 1.03 [0.87–1.24]
Reference 1.28 [0.87–1.89] 1.40 [0.89–2.21]
Reference 1.32 [0.86–2.03] 1.60 [0.96–2.65]
Reference 1.29 [1.08–1.53]** 0.30 [0.19–0.48]***
Reference 1.39 [0.89–2.16] 1.48 [0.67–3.23]
Reference 1.20 [0.74–1.95] 1.29 [0.54–3.11]
Reference 3.87 [2.56–5.86]***
Reference 3.27 [2.14–5.01]***
Reference 3.02 [1.93–4.72]***
Reference 3.04 [2.24–4.14]***
Reference 1.60 [0.88–2.91]
Reference 1.43 [0.74–2.74]
Reference 3.57 [3.02–4.23]***
Reference 2.88 [1.91–4.33]***
Reference 2.92 [1.94–4.38]***
Reference 1.36 [0.99–1.87] 1.95 [1.43–2.66]*** 2.53 [1.83–3.49]*** 3.98 [2.79–5.66]*** 1.33 [1.30–1.37]***
Reference 1.37 [0.75–2.51] 2.44 [1.20–4.98]* 2.85 [1.32–6.15]** 2.77 [1.05–7.33]* 1.40 [1.29–1.53]***
*p < 0.05, **p < 0.01, ***p < 0.001.
specialty medications.21 A possible interpretation for this finding is that many specialty pharmaceuticals identified in this study were used for conditions related to women. Females have conditions such as breast cancer and osteoporosis which require chronic use of specialty medications for these conditions. An inclusion of these conditions in this study could have resulted in more prevalent specialty medication use among females. Patients with worse health states or more comorbidities (regardless of their ages) were highly likely to use more specialty medications in addition to traditional medications. These patients may have used specialty medications with superior therapeutic efficacy to achieve treatment goals that had not been achieved using traditional medications alone. This study has several strengths. To the authors' knowledge, this is the first study to employ a large, nationally-representative dataset to identify patient characteristics associated with specialty medication
state) were 68.0%, 83%, and 91% in the low income, middle income, and high income groups, respectively (p < 0.0001). Given their excellent health state, high income patients might not need more expensive specialty medications as much as those with a worse health state. Patients having a usual source of care were less likely to use specialty medications compared with those without a usual source of care. Although having a usual source of care has been related to the receipt of preventive care or screening services in earlier studies,32,33 its association with prescription drug use has not been explored. Future studies are needed to determine a provider's roles in the use of specialty medications in their patients. When comparing BMUs with TMUs, female gender was found to be a significant predictor for using specialty medications in addition to traditional medications. Hosseini Jebeli et al. have also shown a positive association between female gender and the increased use of 6
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
References
use. Findings from this study underscore the need for appropriate management strategies in using specialty medications with a high incidence of adverse events and compliance issues. For example, older adults or individuals with worse health states and more comorbidities who are more likely to use specialty medications should be closely monitored to ensure that they take these medications as directed. Risks of potential drug-drug interactions need to be avoided or minimized. Pharmacists/clinicians can provide education regarding drug dosing, missed-dose management, frequency of dosing, administration (e.g., self-injection technique), storage and disposal, expected therapeutic effect, adverse events, and strategies for adverse events in using specialty medications.34 Empirical evidence has shown that better management of specialty medication use improves patient health-related outcomes such as increased adherence and reduced health care costs.15–18 In addition, this study employed Andersen's health services utilization model as a theoretical framework. This model has been widely used elsewhere in health service research.35−38 As in these previous studies, the model guided the selection of comprehensive factors potentially associated with health services use – i.e., specialty medication use in this study. Several limitations of this study should be noted. First, because the MEPS data file did not include all specialty pharmaceuticals used chronically, the study sample might not be generalizable to all patients using specialty medications. However, this study included most therapeutic classes, including oncology, osteoporosis, multiple sclerosis, inflammatory conditions, hepatitis, and immune deficiency disorders encompassing the majority of specialty pharmaceutical use. Therefore, there would likely be little impact on the study findings from including additional new samples to the current data. Second, this study had a small sample of SMUs, making the study underpowered to detect significant differences in some characteristics between individuals using specialty medications alone and those using traditional medications. A small sample size of SMUs along with a much larger sample size of BMUs indicates that most patients with chronic diseases were likely to use a combination of specialty and traditional medications in their treatment regimen. To make the findings related to SMUs more robust, more patients using only specialty medications would be needed in a future study. Finally, although covariates were adjusted for in the regression models, residual confounding may nonetheless exist by unmeasured factors. This is always a concern from using secondary data inherent in observational studies. To the extent that any unmeasured characteristics affect outcome measures, the results will be biased. However, this bias is likely to be small, if any, since all the variables addressed in Andersen's model were comprehensively controlled for in this study.
1. Navarro RP, Johnson KA. Opportunities and challenges of specialty pharmaceuticals. J Manag Care Pharm. 2013;19:70–71. 2. Hirsch BR, Balu S, Schulman KA. The impact of specialty pharmaceuticals as drivers of health care costs. Health Aff Millwood. 2014;33:1714–1720. 3. Zalesak M, Greenbaum JS, Cohen JT, et al. The value of specialty pharmaceuticals - a systematic review. Am J Manag Care. 2014;20:461–472. 4. Schumock GT, Li EC, Suda KJ, et al. National trends in prescription drug expenditures and projections for 2014. Am J Health Syst Pharm. 2014;71:482–499. 5. Express Scripts Inc. 2014 Drug Trend Report. St. Louis: Express Scripts; 2015. 6. Express Scripts Inc. 2011 Drug Trend Report. St. Louis: Express Scripts; 2012. 7. PriceWaterhouseCoopers Health Research Institute. Medical Cost Trend: Behind the Number 2016. Available at: https://www.pwc.com/us/en/health-industries/behindthe-numbers/assets/pwc-hri-medical-cost-trend-chart-pack-2016.pdf. Accessed April, 2017. 8. Tharaldson A. Specialty Drug Approvals in 2013. Available at: http://lab.expressscripts.com/lab/insights/drug-options/specialty-drug-approvals-in-2013. Accessed May, 2017. 9. American Journal of Managed Care. The Growing Cost of Specialty Pharmacy - Is it Sustainable? Available at: http://www.ajmc.com/payer-perspectives/0213/thegrowing-cost-of-specialty-pharmacyis-it-sustainable. Accessed May, 2017. 10. Managed Care. Biggest Increases in Specialty Drug Costs Seen in Rheumatologic, Cancer, and Hemophilia Agents. Available at: https://www.managedcaremag.com/ linkout/2005/2/68. Accessed May, 2017. 11. Mullins CD, DeVries AR, Hsu VD, Meng F, Palumbo FB. Variability and growth in spending for outpatient specialty pharmaceuticals. Health Aff Millwood. 2005;24:1117–1127. 12. Jacobs MS, Johnson KA. Curbing the costly trend: exploring the need for a progressive approach to the management of specialty pharmaceuticals under the medical benefit. Am Health Drug Benefits. 2012;5:280–289. 13. Trish E, Joyce G, Goldman DP. Specialty drug spending trends among medicare and medicare advantage enrollees, 2007-11. Health Aff Millwood. 2014;33:2018–2024. 14. Gleason PP, Alexander GC, Starner CI, et al. Health plan utilization and costs of specialty drugs within 4 chronic conditions. J Manag Care Pharm. 2013;19:542–548. 15. Hlubocky JM, Stuckey LJ, Schuman AD, Stevenson JG. Evaluation of a transplantation specialty pharmacy program. Am J Health Syst Pharm. 2012;69:340–347. 16. Murphy P, Cocohoba J, Tang A, Pietrandoni G, Hou J, Guglielmo BJ. Impact of HIVspecialized pharmacies on adherence and persistence with antiretroviral therapy. AIDS Patient Care STDS. 2012;26:526–531. 17. Tschida S, Aslam S, Khan TT, Sahli B, Shrank WH, Lal LS. Managing specialty medication services through a specialty pharmacy program: the case of oral renal transplant immunosuppressant medications. J Manag Care Pharm. 2013;19:26–41. 18. Henderson RR, Visaria J, Bridges GG, Dorholt M, Levin RJ, Frazee SG. Impact of specialty pharmacy on telaprevir-containing 3-drug hepatitis C regimen persistence. J Manag Care Spec Pharm. 2014;20:1227–1234. 19. Goldman DP, Jena AB, Lakdawalla DN, Malin JL, Malkin JD, Sun E. The value of specialty oncology drugs. Health Serv Res. 2010;45:115–132. 20. Romley JA, Sanchez Y, Penrod JR, Goldman DP. Survey results show that adults are willing to pay higher insurance premiums for generous coverage of specialty drugs. Health Aff Millwood. 2012;31:683–690. 21. Hosseini Jebeli SS, Barouni M, Orojloo PH, Mehraban S. Estimating the marginal effect of socioeconomic factors on the demand of specialty drugs. Glob J Health Sci. 2014;7:28–37. 22. Cohen JW, Cohen SB, Banthin JS. The medical expenditure panel survey: a national information resource to support healthcare cost research and inform policy and practice. Med Care. 2009;47(7 Suppl 1):S44–S50. 23. Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav. 1995;36:1–10. 24. D'Hoore W, Bouckaert A, Tilquin C. Practical considerations on the use of the charlson comorbidity index with administrative data bases. J Clin Epidemiol. 1996;49:1429–1433. 25. Pauly MV. The economics of moral hazard: Comment. Am Econ Rev. 1968;58:531–537. 26. Gleason PP, Starner CI, Gunderson BW, Schafer JA, Sarran HS. Association of prescription abandonment with cost share for high-cost specialty pharmacy medications. J Manag Care Pharm. 2009;15:648–658. 27. Doshi JA, Li P, Ladage VP, Pettit AR, Taylor EA. Impact of cost sharing on specialty drug utilization and outcomes: a review of the evidence and future directions. Am J Manag Care. 2016;22:188–197. 28. Kim YA, Rascati KL, Prasla K, Godley P, Goel N, Dunlop D. Retrospective evaluation of the impact of copayment increases for specialty medications on adherence and persistence in an integrated health maintenance organization system. Clin Ther. 2011;33:598–607. 29. Valluri S, Seoane-Vazquez E, Rodriguez-Monguio R, Szeinbach SL. Drug utilization and cost in a medicaid population: a simulation study of community vs. mail order pharmacy. BMC Health Serv Res. 2007;7:122. 30. Pharmaceutical Care Management Association. The Management of Specialty Drugs. Available at: https://www.spcma.org/wp-content/uploads/2016/06/sPCMA_The_ Management_of_Specialty_Drugs.pdf. Accessed April 2017. 31. Schwartz RN, Eng KJ, Frieze DA, et al. NCCN task force report: specialty pharmacy. J Natl Compr Canc Netw. 2010;8(Suppl 4):S1–S12. 32. Ettner SL. The timing of preventive services for women and children: the effect of having a usual source of care. Am J Public Health. 1996;86:1748–1754. 33. DeVoe JE, Fryer GE, Phillips R, Green L. Receipt of preventive care among adults:
5. Conclusions Given the growing popularity of using specialty pharmaceuticals, understanding the characteristics of patients using these medications is essential. This study identified sociodemographic, economic, and clinical factors associated with specialty medication use among U.S. adults. As a next step, appropriate management strategies are needed in using specialty medications with a high incidence of side effects and compliance problems. Attention should be paid to ensure that patients requiring specialty medications actually have access to these medications and use them in an appropriate and safe way. Once identified, individuals receiving specialty medication can receive education and interventions specifically tailored for them by clinicians, thereby improving health outcomes.
Conflicts of interest All authors have nothing to declare. 7
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
T. Park et al.
2006;17:395–404. 37. Deshpande M, Chewning B, Mott D, Thorpe JM, Young HN. Asthma medication use among U.S. adults 18 and older. Res Soc Adm Pharm. 2014;10:e113–e123. 38. Look KA. Patient characteristics associated with multiple pharmacy use in the U.S. population: findings from the medical expenditure panel survey. Res Soc Adm Pharm. 2015;11:507–516.
insurance status and usual source of care. Am J Public Health. 2003;93:786–791. 34. Mullican KA, Francart SJ. The role of specialty pharmacy drugs in the management of inflammatory diseases. Am J Health Syst Pharm. 2016;73:821–830. 35. Aparasu RR, Mort JR, Brandt H. Polypharmacy trends in office visits by the elderly in the United States, 1990 and 2000. Res Soc Adm Pharm. 2005;1:446–459. 36. Farley JF, Cline RR, Gupta K. Racial variations in antiresorptive medication use: results from the 2000 medical expenditure panel survey (MEPS). Osteoporos Int.
8