RESEARCH
Patterns of use and expenses associated with mail-service pharmacy in adults with diabetes Jun Wu, Mary Lynn Davis–Ajami, and Virginia Noxon
Abstract Objectives: To identify socioeconomic factors associated with mail-service pharmacy use and compare the differences in disease-specific prescription medication and medical utilization expenses in a nationally representative sample of adults with diabetes. Design: A retrospective, longitudinal, cross-sectional study. Setting: United States in 2006–11. Participants: Medical Expenditure Panel Survey household component (MEPS-HC) participants aged 18 years or older diagnosed with diabetes and prescribed antidiabetic medications. Main outcome measures: Likelihood of mail-service pharmacy use, diabetes-related medical utilization, and medication expenses. Results: Among 4,430 eligible participants identified in the 2006–11 surveys, representing more than 83 million U.S. individuals, nearly 13% of the participants obtained two-thirds or more of their antidiabetic medications via mail service predominantly. Mail-service pharmacy users were older, had high school or college degrees, had higher incomes, and were more likely to be covered by private insurance. There were no significant differences in diabetes-related medical utilization and drug expenses between the two groups. Conclusion: Besides pharmacy benefit design, sociodemographic and economic factors influenced drug dispensing channel use (mail service versus community pharmacy). No significant differences in diabetes-related drug and medical expenses between mail-service and community pharmacy users were observed. J Am Pharm Assoc. 2015;55:41–51. doi: 10.1331/JAPhA.2015.14058
Journal of the American Pharmacists Association
j apha.org
Jun Wu, PhD, Assistant Professor, Department of Clinical Pharmacy and Outcomes Sciences, South Carolina College of Pharmacy, University of South Carolina, Greenville, SC Mary Lynn Davis–Ajami, PhD, MBA, MS, FNP-C, RN, Assistant Professor, Nursing Administration & Leadership, Department of Adult Health and Nursing Systems, School of Nursing, Virginia Commonwealth University, Richmond, VA Virginia Noxon, MS, Graduate Student, Department of Clinical Pharmacy and Outcomes Sciences, South Carolina College of Pharmacy, University of South Carolina, Columbia, SC Correspondence: Jun Wu, PhD, South Carolina College of Pharmacy, University of South Carolina, 701 Grove Road, Health Sciences Administration Bldg. (MIPH), Greenville, SC 29605; E-mail: wujun@ sccp.sc.edu Disclosure: The authors declare no relevant conflicts of interest or financial relationships. Previous presentation: International Society for Pharmacoeconomics and Outcomes Research 19th Annual International Meeting, Montreal, Quebec, Canada, June 1–4, 2014 Received April 3, 2014. Accepted for publication July 1, 2014. Published online in advance of print December 24, 2014.
JA N /FEB 2015 | 55:1 |
JAPhA 41
RESEARCH
MAIL-SERVICE PHARMACY IN DIABETES CARE
M
anaging a chronic disease such as diabetes may involve ongoing medication use, and medication adherence influences patient outcomes and medical costs in such patients.1–3 In diabetes, the pharmacy plays a significant role in the variation in medication adherence.4 For patients with chronic diseases, maintenance medication refills often provide a point of contact with the health care system and an indicator of medicationuse behavior.5 In the United States, most maintenance medications are dispensed through two types of outpatient dispensing channels, community-based pharmacies (chain or independently owned pharmacies) and mail-service pharmacies. Community pharmacies now offer 30- or 90-day supply refills and face-to-face consultations.6 Mail-service pharmacies offer direct mail delivery, telephonic consultations, and 90-day medication refills. Pharmacy benefit managers (PBMs) and health plan sponsors often suggest and may require clients to use mail-service pharmacies as a means of providing cost savings to members, medication management services, and a 90day medication supply.7–9
Key Points Background: ■■ Successful and timely prescription refills are necessary to ensure good medication adherence. Different services provided by mailservice and community pharmacies have the potential to affect patient medication-use behavior. ■■ The knowledge about how patient socioeconomic factors affect the choice of pharmacy dispending channels in the United States is limited. ■■ Whether mail-service pharmacy can produce cost savings for both patients and third party payers is equivocal because of sampling and methodology used in the studies. It is not clear how mail-service pharmacy use affects total health expenses in diabetes care. Findings: ■■ Nearly 13% of the eligible participants with diabetes filled at least two-thirds of their antidiabetic medications via mail service in the study. ■■ Age, family size, marital status, education and income levels, and insurance coverage were associated with the choice of pharmacy dispensing channels. ■■ No significant differences in health expenses were observed in mail-service and community pharmacy users.
42 JAPhA | 5 5 : 1 | JAN/F E B 2 0 1 5
ja p h a.org
Among those with diabetes, mail-service pharmacy use shows significantly greater medication adherence,6,10–14 trends toward lower diabetes-related medical costs,10 and lower emergency department use.11 Other studies report slightly greater medication adherence for maintenance medications dispensed through community pharmacies on a propensity-matched basis.15 Other factors besides 90-day supply, such as pharmacist–patient interaction, may influence whether the community or mail-service pharmacy significantly influences medication adherence.16 A few studies compare utilization and costs between mail-service and community pharmacies. One study reports mail service predominantly but not always has lower gross reimbursement rates per unit costs.17 Another study comparing utilization rates and costs for 90-day prescriptions among payers with equivalent benefits found that those insured by the payers chose community pharmacy over mail service by a ratio of 4:1 and had no significant differences in allowed charges per claim.18 However, pharmacy benefit design such as outof-pocket payments or copayments for brand medications versus generics can significantly influence choice of dispensing channel and drug utilization.19 In addition, many studies that examine differences between community and mail-service pharmacies use pharmacy claims data.8,10,12,14–19 To our knowledge, no studies use population-based cohorts to examine patient-level socioeconomic factors influencing the utilization of mail-service versus community pharmacies. There is a paucity of research assessing population-level information about maintenance-medication drug expenses in mail-service and community pharmacies.
Objectives The objectives of this study were to identify socioeconomic factors associated with mail-service pharmacy use in a nationally representative sample of adults with diabetes and to compare the differences in oral antiglycemic and insulin antidiabetic prescription medication expenses and diabetes-related medical expenses between mail-service and community pharmacies in this cohort.
Methods
Data source A retrospective, observational, longitudinal, cross-sectional study covering the years 2006–11 was conducted among Medical Expenditure Panel Survey household component (MEPS-HC) participants aged 18 years and older diagnosed with diabetes. Publicly available data from MEPS longitudinal data files were used, specifically panels 11 (years 2006–07), 12 (years 2007–08), 13 (years 2008–09), 14 (years 2009–10), and panel 15 (years 2010–11) as well as the medical conditions and event Journal of the American Pharmacists Association
MAIL-SERVICE PHARMACY IN DIABETES CARE RESEARCH
files. The longitudinal files provided information about demographic and socioeconomic characteristics and the medical conditions. Event files provided information about medical diagnosis, prescription medications, type of pharmacy, medical utilization, and medical expenses. The MEPS-HC samples are derived from a nationally representative subsample of household participants from the prior year’s national Health Interview Survey. MEPS gives nationally representative estimates of health care use, expenditures, sources of payment, and health insurance coverage of the U.S. civilian noninstitutionalized population.20 The MEPS-HC survey uses an overlapping panel design to collect detailed information about demographic and socioeconomic characteristics, health status, access to health care, and health care utilization and expenses. Participants complete interviews for only one panel (2-year time frame). Study population The study included eligible MEPS respondents age 18 years and older with a diagnosis for diabetes who were prescribed one or more prescription antidiabetic medications (oral or insulin). Eligible subjects were defined as participating the entire 2-year panel period and completing all five interview rounds, as evidenced by having longitudinal weights greater than zero (>0). A diabetes diagnosis was identified using the clinical classification software (CCS) variable that categorizes the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) into clinically meaningful categories (CCS code = 049, diabetes without complication; and CCS code = 050, diabetes with complications). CCS classifications were chosen over the MEPS ICD-9-CM codes variable because the MEPS public use files only provide three digit ICD-9-CM codes, which precludes making any distinctions in type of diabetes or disease severity. Antidiabetic medications were identified using the variables for the American Hospital Formulary Service (AHFS) Pharmacologic-Therapeutic Classification MULTIM therapeutic class and subclasses (antidiabetic agents, insulin, and oral antidiabetic drugs) from the MEPS prescribed medicines files. The study excluded individuals younger than 18 years and to exclude those with gestational diabetes, pregnant women. Outcome measures Type of medication dispensing channel was measured using data from the prescribed medicines files showing the type of pharmacy where medications were purchased including mail-service and community pharmacies (independent and traditional chain pharmacies, grocery stores, and pharmacies located in other retail settings). Previous studies reported the proportion of patients exclusively using mail service was small.6 This study defined mail-service pharmacy users as those Journal of the American Pharmacists Association
who obtained as least 66.7% of their antidiabetic medication from a mail-service pharmacy.6,12 Community pharmacy users were defined as those with at least 66.7% of their prescription antidiabetic medications from community pharmacies. Participants without at least 66.7% of their antidiabetic medication from either mail-service or community pharmacies were excluded.6 Economic outcomes included person-level mean diabetes-related health care expenses per participant per year for medical utilization and antidiabetic prescription drugs. The diabetes-related medical expenses and utilization were identified by CCS codes from medical event linked to medical condition files, including officebase visits, hospitalization, outpatient visits, and emergency department (ED) visits. The use of hospitalization and ED were standardized to the number of events per 1,000 patient–months. Each eligible participant in the study had 24 person–months. The antidiabetic drug expenses were identified from prescribed medicines files. These medication files provided validated drug expenses based on pharmacy claim records and patient self-reported costs. Out-of-pocket payments and the amounts paid by insurance were measured respectively. All the expenses were inflated to year 2013 using the Consumer Price Index.21 Potential variables associated with mail-service pharmacy use The Statistical Brief on population characteristics of persons purchasing prescription drugs from mail-service pharmacies issued by the Agency for Healthcare Research and Quality in 2008 informed our selection of potential variables.22 Socioeconomic and health characteristics between mail-service and community pharmacy users were compared, including demographic variables (age, gender, race, marital status, region, family size, education and income levels, and insurance coverage) and health-related variables (perceived health status, body mass index, type of antidiabetic medication use, and comorbidity). The Charlson Comorbidity Index (CCI) was constructed using the three-digit ICD-9-CM codes from the medical condition files and calculated as a measure of comorbidity.23 Data analysis T-tests for continuous variables and chi-square tests for categorical variables were employed to compare differences in mail-service and community pharmacy users. Variables with P
Patterns of use and expenses associated with mail-service pharmacy in adults with diabetes Jun Wu, Mary Lynn Davis–Ajami, and Virginia Noxon
Abstract Objectives: To identify socioeconomic factors associated with mail-service pharmacy use and compare the differences in disease-specific prescription medication and medical utilization expenses in a nationally representative sample of adults with diabetes. Design: A retrospective, longitudinal, cross-sectional study. Setting: United States in 2006–11. Participants: Medical Expenditure Panel Survey household component (MEPS-HC) participants aged 18 years or older diagnosed with diabetes and prescribed antidiabetic medications. Main outcome measures: Likelihood of mail-service pharmacy use, diabetes-related medical utilization, and medication expenses. Results: Among 4,430 eligible participants identified in the 2006–11 surveys, representing more than 83 million U.S. individuals, nearly 13% of the participants obtained two-thirds or more of their antidiabetic medications via mail service predominantly. Mail-service pharmacy users were older, had high school or college degrees, had higher incomes, and were more likely to be covered by private insurance. There were no significant differences in diabetes-related medical utilization and drug expenses between the two groups. Conclusion: Besides pharmacy benefit design, sociodemographic and economic factors influenced drug dispensing channel use (mail service versus community pharmacy). No significant differences in diabetes-related drug and medical expenses between mail-service and community pharmacy users were observed. J Am Pharm Assoc. 2015;55:41–51. doi: 10.1331/JAPhA.2015.14058
Journal of the American Pharmacists Association
j apha.org
Jun Wu, PhD, Assistant Professor, Department of Clinical Pharmacy and Outcomes Sciences, South Carolina College of Pharmacy, University of South Carolina, Greenville, SC Mary Lynn Davis–Ajami, PhD, MBA, MS, FNP-C, RN, Assistant Professor, Nursing Administration & Leadership, Department of Adult Health and Nursing Systems, School of Nursing, Virginia Commonwealth University, Richmond, VA Virginia Noxon, MS, Graduate Student, Department of Clinical Pharmacy and Outcomes Sciences, South Carolina College of Pharmacy, University of South Carolina, Columbia, SC Correspondence: Jun Wu, PhD, South Carolina College of Pharmacy, University of South Carolina, 701 Grove Road, Health Sciences Administration Bldg. (MIPH), Greenville, SC 29605; E-mail: wujun@ sccp.sc.edu Disclosure: The authors declare no relevant conflicts of interest or financial relationships. Previous presentation: International Society for Pharmacoeconomics and Outcomes Research 19th Annual International Meeting, Montreal, Quebec, Canada, June 1–4, 2014 Received April 3, 2014. Accepted for publication July 1, 2014. Published online in advance of print December 24, 2014.
JA N /FEB 2015 | 55:1 |
JAPhA 41
RESEARCH
MAIL-SERVICE PHARMACY IN DIABETES CARE
M
anaging a chronic disease such as diabetes may involve ongoing medication use, and medication adherence influences patient outcomes and medical costs in such patients.1–3 In diabetes, the pharmacy plays a significant role in the variation in medication adherence.4 For patients with chronic diseases, maintenance medication refills often provide a point of contact with the health care system and an indicator of medicationuse behavior.5 In the United States, most maintenance medications are dispensed through two types of outpatient dispensing channels, community-based pharmacies (chain or independently owned pharmacies) and mail-service pharmacies. Community pharmacies now offer 30- or 90-day supply refills and face-to-face consultations.6 Mail-service pharmacies offer direct mail delivery, telephonic consultations, and 90-day medication refills. Pharmacy benefit managers (PBMs) and health plan sponsors often suggest and may require clients to use mail-service pharmacies as a means of providing cost savings to members, medication management services, and a 90day medication supply.7–9
Key Points Background: ■■ Successful and timely prescription refills are necessary to ensure good medication adherence. Different services provided by mailservice and community pharmacies have the potential to affect patient medication-use behavior. ■■ The knowledge about how patient socioeconomic factors affect the choice of pharmacy dispending channels in the United States is limited. ■■ Whether mail-service pharmacy can produce cost savings for both patients and third party payers is equivocal because of sampling and methodology used in the studies. It is not clear how mail-service pharmacy use affects total health expenses in diabetes care. Findings: ■■ Nearly 13% of the eligible participants with diabetes filled at least two-thirds of their antidiabetic medications via mail service in the study. ■■ Age, family size, marital status, education and income levels, and insurance coverage were associated with the choice of pharmacy dispensing channels. ■■ No significant differences in health expenses were observed in mail-service and community pharmacy users.
42 JAPhA | 5 5 : 1 | JAN/F E B 2 0 1 5
ja p h a.org
Among those with diabetes, mail-service pharmacy use shows significantly greater medication adherence,6,10–14 trends toward lower diabetes-related medical costs,10 and lower emergency department use.11 Other studies report slightly greater medication adherence for maintenance medications dispensed through community pharmacies on a propensity-matched basis.15 Other factors besides 90-day supply, such as pharmacist–patient interaction, may influence whether the community or mail-service pharmacy significantly influences medication adherence.16 A few studies compare utilization and costs between mail-service and community pharmacies. One study reports mail service predominantly but not always has lower gross reimbursement rates per unit costs.17 Another study comparing utilization rates and costs for 90-day prescriptions among payers with equivalent benefits found that those insured by the payers chose community pharmacy over mail service by a ratio of 4:1 and had no significant differences in allowed charges per claim.18 However, pharmacy benefit design such as outof-pocket payments or copayments for brand medications versus generics can significantly influence choice of dispensing channel and drug utilization.19 In addition, many studies that examine differences between community and mail-service pharmacies use pharmacy claims data.8,10,12,14–19 To our knowledge, no studies use population-based cohorts to examine patient-level socioeconomic factors influencing the utilization of mail-service versus community pharmacies. There is a paucity of research assessing population-level information about maintenance-medication drug expenses in mail-service and community pharmacies.
Objectives The objectives of this study were to identify socioeconomic factors associated with mail-service pharmacy use in a nationally representative sample of adults with diabetes and to compare the differences in oral antiglycemic and insulin antidiabetic prescription medication expenses and diabetes-related medical expenses between mail-service and community pharmacies in this cohort.
Methods
Data source A retrospective, observational, longitudinal, cross-sectional study covering the years 2006–11 was conducted among Medical Expenditure Panel Survey household component (MEPS-HC) participants aged 18 years and older diagnosed with diabetes. Publicly available data from MEPS longitudinal data files were used, specifically panels 11 (years 2006–07), 12 (years 2007–08), 13 (years 2008–09), 14 (years 2009–10), and panel 15 (years 2010–11) as well as the medical conditions and event Journal of the American Pharmacists Association
MAIL-SERVICE PHARMACY IN DIABETES CARE RESEARCH
files. The longitudinal files provided information about demographic and socioeconomic characteristics and the medical conditions. Event files provided information about medical diagnosis, prescription medications, type of pharmacy, medical utilization, and medical expenses. The MEPS-HC samples are derived from a nationally representative subsample of household participants from the prior year’s national Health Interview Survey. MEPS gives nationally representative estimates of health care use, expenditures, sources of payment, and health insurance coverage of the U.S. civilian noninstitutionalized population.20 The MEPS-HC survey uses an overlapping panel design to collect detailed information about demographic and socioeconomic characteristics, health status, access to health care, and health care utilization and expenses. Participants complete interviews for only one panel (2-year time frame). Study population The study included eligible MEPS respondents age 18 years and older with a diagnosis for diabetes who were prescribed one or more prescription antidiabetic medications (oral or insulin). Eligible subjects were defined as participating the entire 2-year panel period and completing all five interview rounds, as evidenced by having longitudinal weights greater than zero (>0). A diabetes diagnosis was identified using the clinical classification software (CCS) variable that categorizes the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) into clinically meaningful categories (CCS code = 049, diabetes without complication; and CCS code = 050, diabetes with complications). CCS classifications were chosen over the MEPS ICD-9-CM codes variable because the MEPS public use files only provide three digit ICD-9-CM codes, which precludes making any distinctions in type of diabetes or disease severity. Antidiabetic medications were identified using the variables for the American Hospital Formulary Service (AHFS) Pharmacologic-Therapeutic Classification MULTIM therapeutic class and subclasses (antidiabetic agents, insulin, and oral antidiabetic drugs) from the MEPS prescribed medicines files. The study excluded individuals younger than 18 years and to exclude those with gestational diabetes, pregnant women. Outcome measures Type of medication dispensing channel was measured using data from the prescribed medicines files showing the type of pharmacy where medications were purchased including mail-service and community pharmacies (independent and traditional chain pharmacies, grocery stores, and pharmacies located in other retail settings). Previous studies reported the proportion of patients exclusively using mail service was small.6 This study defined mail-service pharmacy users as those Journal of the American Pharmacists Association
who obtained as least 66.7% of their antidiabetic medication from a mail-service pharmacy.6,12 Community pharmacy users were defined as those with at least 66.7% of their prescription antidiabetic medications from community pharmacies. Participants without at least 66.7% of their antidiabetic medication from either mail-service or community pharmacies were excluded.6 Economic outcomes included person-level mean diabetes-related health care expenses per participant per year for medical utilization and antidiabetic prescription drugs. The diabetes-related medical expenses and utilization were identified by CCS codes from medical event linked to medical condition files, including officebase visits, hospitalization, outpatient visits, and emergency department (ED) visits. The use of hospitalization and ED were standardized to the number of events per 1,000 patient–months. Each eligible participant in the study had 24 person–months. The antidiabetic drug expenses were identified from prescribed medicines files. These medication files provided validated drug expenses based on pharmacy claim records and patient self-reported costs. Out-of-pocket payments and the amounts paid by insurance were measured respectively. All the expenses were inflated to year 2013 using the Consumer Price Index.21 Potential variables associated with mail-service pharmacy use The Statistical Brief on population characteristics of persons purchasing prescription drugs from mail-service pharmacies issued by the Agency for Healthcare Research and Quality in 2008 informed our selection of potential variables.22 Socioeconomic and health characteristics between mail-service and community pharmacy users were compared, including demographic variables (age, gender, race, marital status, region, family size, education and income levels, and insurance coverage) and health-related variables (perceived health status, body mass index, type of antidiabetic medication use, and comorbidity). The Charlson Comorbidity Index (CCI) was constructed using the three-digit ICD-9-CM codes from the medical condition files and calculated as a measure of comorbidity.23 Data analysis T-tests for continuous variables and chi-square tests for categorical variables were employed to compare differences in mail-service and community pharmacy users. Variables with P
