ORIGINAL ARTICLE

Hidden Costs Associated With Venous Thromboembolism Impact of Lost Productivity on Employers and Employees Robert L. Page II, PharmD, MSPH, Vahram Ghushchyan, PhD, Brian Gifford, PhD, Richard Allen Read, MS, Monika Raut, PhD, Brahim K. Bookhart, MBA, MPH, Ahmad B. Naim, MD, C. V. Damaraju, PhD, and Kavita V. Nair, PhD Objective: To determine productivity loss and indirect costs with deep vein thrombosis (DVT) and pulmonary embolism (PE). Methods: Medical and pharmacy claims with short-term disability (STD) and long-term disability (LTD) claims from 2007 to 2010 were analyzed from the Integrated Benefits Institute’s Health and Productivity Benchmarking (IBI-HPB) database (STD and LTD claims) and IMS LifeLinkTM data (medical and pharmacy claims), which were indirectly linked using a weighting approach matching from IBIHPB patients’ demographic distribution. Results: A total of 5442 DVT and 6199 PE claims were identified. Employees with DVT lost 57 STD and 440 LTD days per disability incident. The average per claim productivity loss from STD and LTD was $7414 and $58181, respectively. Employees with PE lost 56 STD and 364 LTD days per disability incident. The average per claim productivity loss from STD and LTD was $7605 and $48,751, respectively. Conclusions: Deep vein thrombosis and PE impose substantial economic burdens.

V

enous thromboembolism (VTE), consisting of deep vein thrombosis (DVT) and pulmonary embolism (PE), is the third most common cardiovascular disease behind coronary artery disease and stroke, making the condition a leading public health concern in the United States.1–3 Although the current US-wide incidence of VTE varies within the literature, clinical administrative databases and hospital- and community-based studies have suggested the annual incidence to be between 1 and 2 per 1000 population, which equates to 300,000 to 600,000 new cases each year.4 Nevertheless, this rate may be an underestimate as recent models propose that more than 900,000 incident or recurrent cases of fatal and nonfatal VTE occur in the United States annually.5 Deitelzweig et al6 calculated that about 1 million Americans suffer from a VTE each year, a number that is projected to increase to 1.82 million by 2050. On the basis of 2007 to 2009 data from the National Hospital Discharge Survey, the Centers for Disease Control estimated that 547,596 adults 18 years of age and older are hospitalized annually for VTE, with an estimated annual average of 348,558 hospitalizations for DVT and 277,549 for PE.7 The risk for the development of VTE is multifactorial. Traditional risk factors consist of prolonged immobilization, infection, surgical procedures, trauma, cancer, and inherited coagulation From the Department of Clinical Pharmacy (Drs Page, Ghushchyan, and Nair), Skaggs School of Pharmacy and Pharmaceutical Sciences; Department of Physical Medicine (Dr Page), School of Medicine, University of Colorado, Aurora; College of Business and Economics (Dr Ghushchyan), American University of Armenia, Yerevan; Integrated Benefits Institute (Dr Gifford), San Francisco, Calif; Peakstat Statistical Services (Mr Read), Evergreen, Colo; and Janssen Scientific Affairs (Dr Raut, Mr Bookhart, and Dr Naim), LLC, Raritan, NJ. This study was supported by Janssen Pharmaceuticals, Raritan, New Jersey. The authors declare no conflicts of interest. Send correspondence to: Kavita V. Nair, PhD, Skaggs School of Pharmacy and Pharmaceutical Sciences, University of Colorado, Mail Stop C238, 12580 E Montview Blvd, Aurora, CO 80045 (Kavita.Nair@ucdenver). C 2014 by American College of Occupational and Environmental Copyright  Medicine DOI: 10.1097/JOM.0000000000000208

disorders.8–10 As the majority of these risk factors are closely linked to hospitalization, it is not surprising that between 59% and 75% of all VTE cases occur within the institutional setting.11 Nonetheless, VTE can develop in patients who are neither hospitalized nor recovering from a major illness. Pregnancy, use of hormone replacement or oral contraceptives, and advanced age (eg, >75 years) have all been associated with an increased risk for VTE.10 VTE also shares many classic, modifiable risk factors with that of coronary artery disease, consisting of obesity, cigarette smoking, hypertension, diabetes, and hypercholesterolemia.10 After the first VTE event, long-term complications can be major sources of both increased morbidity and mortality. In the United States, VTE is responsible for at least 100,000 and up to nearly 300,000 deaths annually in which 2% to 10% of all hospital deaths are attributable to PE.12,13 Within 1 to 2 years of the index diagnosis, 20% to 50% of patients with a DVT can develop postthrombotic syndrome and about 5% of patients with a PE develop chronic thromboembolic pulmonary hypertension.13–16 Although the majority of patients with postthrombotic syndrome suffer from chronic pain and lower extremity swelling, about 5% to 10% are diagnosed with severe symptoms such as debilitating pain, ulceration, and incapacitating limb swelling—all associated with diminished quality of life and loss in functional status.13–16 In addition, once diagnosed, the risk of VTE recurrence remains excessively high with 5% experiencing a subsequent VTE within 3 months, 6% to 8% within 1 to 2 years, and up to 25% within 5 years of the index event.13 Nonetheless, this disease is highly preventable in approximately 18% to 65% of cases.17 Within the hospital setting, appropriate intervention programs, which incorporate risk factor identification and stratification as well as administration of appropriate drug and nondrug therapies, can reduce VTE as much as 74% to 86%.18–20 Although VTE is more prevalent in an older adult population, a significant number of younger adults also suffer from this debilitating condition.10 Data from 2007 to 2009 from the National Hospital Discharge Survey suggested that 60 per 100,000 population aged 18 to 39 years, 143 for persons aged 40 to 49 years, 200 for persons aged 50 to 59 years, and 391 for persons aged 60 to 69 years were hospitalized for VTE during the study period.7 With this in mind, a diagnosis of VTE could have upsetting social and economic ramifications on both employees and employers because of the potential for disability and lost productivity. These types of indirect costs may be hidden and are not necessarily addressed in the literature. To better define the real-world economic burden that VTE has on the employer and the employee, we determined the productivity losses and calculated their subsequent costs associated with VTE.

METHODS Study Design This study was a descriptive, retrospective data analysis of medical, pharmacy, and disability insurance claims from January 1, 2007, through December 31, 2010, of patients with DVT and PE using two separate data sets: Integrated Benefits Institute’s Health and Productivity Benchmarking (IBI-HPB) and LifeLinkTM health

JOEM r Volume 56, Number 9, September 2014 Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

979

JOEM r Volume 56, Number 9, September 2014

Page et al

plan claims. The primary study objective was to estimate the lost productivity and its subsequent indirect costs limited to short-term disability (STD) and long-term disability (LTD) costs for employees with DVT and PE. The secondary objective was to estimate the direct costs associated for patients who had at least one hospital admission for a DVT or PE.

Data Sources Integrated Benefits Institute’s Health and Productivity Benchmarking The IBI-HPB database was used to obtain indirect cost estimates for STD and LTD incidents because of VTE. STD policies pay replacement wages (at some fraction of regular earnings) to employees with medical conditions that temporarily prevent them from working. Benefits typically begin 1 or 2 weeks after the date of a disabling event (such as a medical emergency or a scheduled medical procedure) and continue until an employee returns to work or until the duration of absences reaches a defined maximum (usually 6 months). LTD policies pay replacement wages when STD benefits have reached their maximum but employees are still unable to return to work. The duration of benefits is not fixed, but can last until an employee recovers sufficiently to return to work or until he or she reaches the normal social security retirement age. Integrated Benefits Institute (IBI) is a leading nonprofit supplier of health and productivity research, measurement, and benchmarking to employers and their supplier partners, with resources for demonstrating the business value of health. Each year, 14 major US disability insurers provide IBI with STD and LTD claims from more than 45,000 employer disability policies.21 The earliest available calendar year of claims was 2007, and the latest (available at the time of the study) was 2010. The data contained claims for which payments were ceased by the end of the calendar year (ie, closed claims) and claims for which payments continued to be paid at the end of the calendar year (ie, open claims). These claims included data on costs and durations of disability, as well as claim and claimant characteristics such as industry, only the primary diagnosis on the basis of the International Classification of Diseases, Ninth Revision (ICD-9), code, date of birth, and sex. Disability costs include wages paid by the policy to employees for days on which they were absent from work. Because STD and LTD benefits typically are administered separately from health care benefits, the claims do not include medical or pharmacy costs for treatments received for the duration of disability. Within this data set, the STD claim is considered closed when an employee returns to work or when the claim reaches its maximum duration (whether or not an employee returned to work), whichever comes sooner; LTD claims can remain open until an employee reaches social security retirement age, receives a lump-sum payout from a policy carrier, or returns to work. To ensure that a claim has time to reach its total duration and to avoid double counting claims that appear in the data twice (once in the calendar year it opened and once in the calendar year it closed), IBI retains all closed claims with a disability date in 2007 to 2010. Any remaining claims with 2007 or later disability data that were still open at the end of the year in 2010 are also retained.

LifeLinkTM Claims The LifeLinkTM claims data set managed by IMS Health was used to obtain medical and pharmacy claims for patients with DVT and PE. This data set includes adjudicated medical and pharmaceutical claims along with demographic information (eg, age, sex, and geographic region) and dates of enrollment for 40 million unique patients from more than 90 health plans across the United States and is representative of the national, commercially insured population on a variety of demographic measures. We obtained a 10% random sample of the entire LifeLinkTM data set (standard extract available for conducting research) for our analysis to identify our VTE cohort. 980

Sample Selection For the IBI-HPB database, the study sample consisted of patients who were 18 years of age and older with STD and LTD claims for DVT and PE during the study period (2007 to 2010). The following ICD-9 codes were used to capture patients with DVT: 451.1, 451.11, 451.18, 451.19, 451.2, 451.8x, 451.9, 453.1, 453.2, 453.3, 453.4x, 453.5x, 453.6, 453.7x, 453.8x, and 453.9. Patients with PE were identified using the following ICD-9 codes: 415.1, 415.11, 415.12, 415.19, and 416.2. For the LifeLinkTM claims data set, the study sample consisted of patients who were 18 years of age and older with at least one ICD-9 diagnosis for DVT and PE during the same study period. The first DVT or PE diagnosis was considered to be the index date. Within the LifeLinkTM claims data set, only patients with at least 12 months of continuous enrollment after this index date with at least 12 months of continuous enrollment with no DVT or PE diagnosis before the index data were included.

Outcome Variables Indirect Costs Using the IBI-HBP data, the following types of indirect costs were estimated: estimated wage loss per claim and productivity loss per disability claim; disability durations defined as the total number of STD and LTD days, and actual workdays compensated for an STD and LTD period. Demographic variables from the IBI-HBP data consisted of age, sex, and type of industry. From the entire IBI-HBP claims data set, STD and LTD disability days were determined for common cardiovascular and other health conditions during the study period. These conditions were identified by ICD-9 codes and consisted of the following: arthritis (710.xx, 711, 712.xx to 719xx), asthma (493.xx), acute coronary syndrome (410.xx, 411.xx, 413.1, 413.9), breast cancer (174.xx, 175.xx), depression (296.1, 296.2), diabetes (250.xx), anxiety/neurotic disorders (300.xx), heart failure (428.xx), hypercholesterolemia (272.xx), hypertension (401.xx to 405.xx), migraine/headache (346.xx), obesity (278.xx), rheumatoid arthritis (717.xx), and stroke (430, 431, 432.xx to 435.xx).

Direct Health Care Expenditures From the LifeLinkTM claims data set, DVT and PE-related health care costs consisted of hospitalizations, emergency department visits, outpatient visits, and prescription medications during the 12 months after the index date. Age, sex, and other comorbid conditions were also captured from the LifeLinkTM claims data. The burden of comorbid conditions was captured using the chronic condition index score.22 The chronic condition index score is calculated from the chronic condition indicator, which is a family of databases and software tools developed by the Health Cost and Utilization Project to categorize ICD-9 diagnosis into chronic and not chronic. A subset of the LifeLinkTM sample was used to calculate the direct costs of DVT and PE and consisted only of those patients with at least one DVT or PE hospital admission.

Statistical Analysis All demographic and outcome measures were analyzed descriptively. For calculating DVT and PE claims rates, each claim contained a code linking it back to a specific employer’s disability policy. For about 70% of employers, information on the number of covered lives in their policy was also included with each claim. An overall claims rate was created for an employer by dividing its total number of new claims in a calendar year by its covered lives. In principle, the DVT and PE claims rate could be calculated in the same manner; however, not all claims with covered lives information have valid ICD-9 information (and vice versa), which could distort the results. Instead, to estimate the DVT and PE-related disability incidence rate within the IBI-HPB claims data set, the DVT and PE

 C 2014 American College of Occupational and Environmental Medicine

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

JOEM r Volume 56, Number 9, September 2014

Hidden Costs of VTE

proportion of claims within the claims with valid ICD-9 information was multiplied by the overall claims rate to arrive at this estimate. Estimated wage loss per claim is an estimate of wages forgone by an employee for the duration of a disability claim. The measure calculates the difference between what an employee would have earned if he or she had worked for the duration of a claim (given the number of missed workdays and daily wages) and what he or she would have likely received in disability replacement wages. For each claim with a Standard Industrial Classification or North American Industrial Classification system code, we appended the Bureau of Labor Statistics’ estimates industry average disability wage replacement rates from their March 2011 Employee Benefits Survey and the industry average daily wage from the Bureau of Labor Statistics’ Occupational Employment Statistics program (inflated from 2010 averages to 2011 dollars using the consumer price index).23,24 To calculate daily wages on a full-time equivalent basis, Bureau of Labor Statistics’ mean annual wages were divided by 260 workdays. Rather than assume that employees on disability were paid at 100% of wages during any required waiting period, the full and replacement wages were calculated on the basis of the total (calendar) disability days on a 5-day workweek. The following equation describes the calculation of estimated lost wages in practice:

patient in the LifeLinkTM data were generated. For example, if there were 50 females in the 45- to 50-year age group in the IBI-HBP data and 100 females in the same age group in the LifeLinkTM data, then the analytic weight for each female in that age group in the LifeLinkTM data would be equal to 0.5. This method allowed comparison of direct and indirect cost estimates. Although these costs were obtained from different data sets, they were homogenous in terms of age and sex distribution. Using these analytic weights, the weighted age and sex distribution of the LifeLinkTM sample was statistically similar to the age and sex distribution of the IBI-HBP data. These analytic weights were used to derive DVT and PE-related health care costs. Direct costs were estimated using only patients with at least one DVT or PE hospital admission. This method was chosen as 59% to 75% of all VTE cases occur within the institutional setting.11 In addition, to be consistent with prior research, many studies estimating direct costs associated with VTE use the index hospitalization to identify their patient population.11,13,17,25–27 Cost estimates from different years were adjusted to a common 2011 dollar value using the medical care component of the consumer price index.28 All analyses R were conducted using STATA version 11 (College Station, TX).

RESULTS Estimated Wage Loss per Claim = (1 − Wage Replacement Rate) × Average Daily Wages   5 ×Total Disability Days × 7 Productivity loss per disability claim was estimated as the wage value of each missed workday (using the industry average daily wage). Missed workdays were calculated assuming a five-day workweek for the total disability duration (ie, total disability days × 5/7). Disability durations are typically enumerated in calendar days. STD days were calculated as the calendar days between the date of a disability for which a claim was filed and the closing date (for closed claims) or December 31, 2010 (for open claims). LTD days were calculated as the calendar days between the date when LTD disability payments began and the closing date (for closed claims) or December 31, 2010 (for open claims). For direct costs, on the basis of the age and sex distribution of both the IBI-HBP and LifeLinkTM data, analytic weights for each

For calculation of indirect costs, a total of 5442 DVT and 6199 PE claims were identified from the IBI-HPB data set. For calculation of direct costs, a total of 10,634 patients met inclusion criteria (9120 with a diagnosis of DVT and 1514 with a diagnosis of PE) (Fig. 1). From this cohort, 2000 patients had at least one hospital admission for a DVT and 651 for PE. When evaluating the number of disability claims, 5085 of DVT claims were for STD and 358 for LTD in which 5824 of PE claims were for STD and 374 for LTD (Tables 1 and 2). For both conditions the sex distribution was similar, 54% of DVT and 50% of PE claims were for male employees, 96% of employees with DVT and PE were 64 years of age or younger with a mean age of 48 ± 11 and 47 ± 11 years for DVT and PE respectively (Table 1). As the data from LifeLinkTM were weighted by analytic weights for sex and age, these findings were similar to the IBI-HPB cohort. The majority of STD claims originated from employees who were working in the manufacturing industry (29% and 26% for DVT and PE, respectively), followed by the services industry (19% and 21% for DVT and PE, respectively) (Fig. 2). Unlike STD claims, the majority

Total beneficiaries in the IMS LifeLink™ dataset between January 1, 2007 – December 31,2010 (N= 3,971,920 )

Beneficiaries over the age of 18 with a diagnosis of VTE (N=30,387)

Beneficiaries with 12 months of connuous enrollment before and aer index VTE date (N= 12,233)

Excluding beneficiaries under the age of 18 and no diagnosis of VTE (N= 3,941,533) Excluding beneficiaries without 12 months of connuous enrollment before and aer index VTE date (N=18,154)

Excluding beneficiaries with both PE and DVT (N=1,599)

FIGURE 1. Study population for the determination of direct costs of PE and VTE from the LifeLinkTM data set. DVT, deep vein thrombosis; PE, pulmonary embolism; VTE, venous thromboembolism.

Diagnosis of DVT only (N=9120) Diagnosis of PE only (N=1,514) At least one hospital admission for a DVT (N=2000) At least one hospital admission for a PE (N=651)

 C 2014 American College of Occupational and Environmental Medicine

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

981

JOEM r Volume 56, Number 9, September 2014

Page et al

TABLE 1. Demographics for Short-Term and Long-Term Disability Claims From the IBI-HPB Data Set and Weighted LifeLinkTM Claims Sample* DVT

Sex Male Female Age groups, yrs

Hidden costs associated with venous thromboembolism: impact of lost productivity on employers and employees.

To determine productivity loss and indirect costs with deep vein thrombosis (DVT) and pulmonary embolism (PE)...
256KB Sizes 0 Downloads 3 Views