Assessing the Cost-effectiveness of a Hypothetical Disease-modifying Therapy With Limited Duration for the Treatment of Early Symptomatic Alzheimer Disease.

PURPOSE: Clinical trials have produced promising results for disease-modifying therapies (DMTs) for Alzheimer's disease (AD); however, the evidence on their potential cost-effectiveness is limited. This study assesses the cost-effectiveness of a hypothetical DMT with a limited treatment duration in AD. METHODS: We developed a Markov state-transition model to estimate the cost-effectiveness of a hypothetical DMT plus best supportive care (BSC) versus BSC alone among Americans living with mild cognitive impairment (MCI) due to AD or mild AD. AD states included MCI due to AD, mild AD, moderate AD, severe AD, and death. A hypothetical DMT was assumed to confer a 30% reduction in progression from MCI and mild AD. The base case annual drug acquisition cost was assumed to be $56,000. Other medical and indirect costs were obtained from published literature or list prices. Utilities for patients and caregivers were obtained from the published literature and varied by AD state and care setting (community care or long-term care). We considered 3 DMT treatment strategies: (1) treatment administered until patients reached severe AD (continuous strategy), (2) treatment administered for a maximum duration of 18 months or when patients reached severe AD (fixed-duration strategy), and (3) 40% of patients discontinuing treatment at 6 months because of amyloid plaque clearance and the remaining patients continuing treatment until 18 months or until they reached severe AD (test-and-discontinue strategy). Incremental cost-effectiveness ratios (ICERs) were calculated as the incremental cost per quality-adjusted life-year (QALY) gained. FINDINGS: From the health care sector perspective, continuous treatment with a hypothetical DMT versus BSC resulted in an ICER of $612,354 per QALY gained. The ICER decreased to $157,288 per QALY gained in the fixed-duration strategy, driven by large reductions in treatment costs. With 40% of patients discontinuing treatment at 6 months (test-and-discontinue strategy), the ICER was $125,631 per QALY gained. In sensitivity and scenario analyses, the ICER was the most sensitive to changes in treatment efficacy, treatment cost, and the initial population AD state distribution. From the modified societal perspective, ICERs were 6.3%, 20.4%, and 25.1% lower than those from the health care sector perspective for the continuous, fixed-duration, and test-and-discontinue strategies, respectively. IMPLICATIONS: Under a set of assumptions for annual treatment costs and the magnitude and duration of treatment efficacy, DMTs used for a limited duration may deliver value consistent with accepted US cost-effectiveness thresholds.

A comparison of reweighting estimators of average treatment effects in real world populations.

Regulatory agencies typically evaluate the efficacy and safety of new interventions and grant commercial approval based on randomized controlled trials (RCTs). Other major healthcare stakeholders, such as insurance companies and health technology assessment agencies, while basing initial access and reimbursement decisions on RCT results, are also keenly interested in whether results observed in idealized trial settings will translate into comparable outcomes in real world settings-that is, into so-called "real world" effectiveness. Unfortunately, evidence of real world effectiveness for new interventions is not available at the time of initial approval. To bridge this gap, statistical methods are available to extend the estimated treatment effect observed in a RCT to a target population. The generalization is done by weighting the subjects who participated in a RCT so that the weighted trial population resembles a target population. We evaluate a variety of alternative estimation and weight construction procedures using both simulations and a real world data example using two clinical trials of an investigational intervention for Alzheimer's disease. Our results suggest an optimal approach to estimation depends on the characteristics of source and target populations, including degree of selection bias and treatment effect heterogeneity.

Reweighting Randomized Controlled Trial Evidence to Better Reflect Real Life - A Case Study of the Innovative Medicines Initiative.

Evidence from randomized controlled trials available for timely health technology assessments of new pharmacological treatments and regulatory decision making may not be generalizable to local patient populations, often resulting in decisions being made under uncertainty. In recent years, several reweighting approaches have been explored to address this important question of generalizability to a target population. We present a case study of the Innovative Medicines Initiative to illustrate the inverse propensity score reweighting methodology, which may allow us to estimate the expected treatment benefit if a clinical trial had been run in a broader real-world target population. We learned that identifying treatment effect modifiers, understanding and managing differences between patient characteristic data sets, and balancing the closeness of trial and target patient populations with effective sample size are key to successfully using this methodology and potentially mitigating some of this uncertainty around local decision making.

Preference for pharmaceutical formulation and treatment process attributes.

PURPOSE: Pharmaceutical formulation and treatment process attributes, such as dose frequency and route of administration, can have an impact on quality of life, treatment adherence, and disease outcomes. The aim of this literature review was to examine studies on preferences for pharmaceutical treatment process attributes, focusing on research in diabetes, oncology, osteoporosis, and autoimmune disorders. METHODS: The literature search focused on identifying studies reporting preferences for attributes of the pharmaceutical treatment process. Studies were required to use formal quantitative preference assessment methods, such as utility valuation, conjoint analysis, or contingent valuation. Searches were conducted using Medline, EMBASE, Cochrane Library, Health Economic Evaluation Database, and National Health Service Economic Evaluation Database (January 1993-October 2013). RESULTS: A total of 42 studies met inclusion criteria: 19 diabetes, nine oncology, five osteoporosis, and nine autoimmune. Across these conditions, treatments associated with shorter treatment duration, less frequent administration, greater flexibility, and less invasive routes of administration were preferred over more burdensome or complex treatments. While efficacy and safety often had greater relative importance than treatment process, treatment process also had a quantifiable impact on preference. In some instances, particularly in diabetes and autoimmune disorders, treatment process attributes had greater relative importance than some or all efficacy and safety attributes. Some studies suggested that relative importance of treatment process depends on disease (eg, acute vs chronic) and patient (eg, injection experience) characteristics. CONCLUSION: Despite heterogeneity in study methods and design, some general patterns of preference clearly emerged. Overall, the results of this review suggest that treatment process has a quantifiable impact on preference and willingness to pay for treatment, even in many situations where safety and efficacy were the primary concerns. Patient preferences for treatment process attributes can inform drug development decisions to better meet the needs of patients and deliver improved outcomes.

Comparative Effectiveness of Carotid Artery Stenting Versus Carotid Endarterectomy Among Medicare Beneficiaries.

BACKGROUND: Effectiveness of carotid artery stenting (CAS) relative to carotid endarterectomy (CEA) among Medicare patients has not been established. We compared effectiveness of CAS versus CEA among Medicare beneficiaries. METHODS AND RESULTS: We linked Medicare data (2000-2009) to the Society for Vascular Surgery's Vascular Registry (2005-2008) and the National Cardiovascular Data Registry's (NCDR) Carotid Artery Revascularization and Endarterectomy Registry (2006-2008/2009). Medicare patients were followed up from procedure date until death, stroke/transient ischemic attack, periprocedural myocardial infarction, or a composite end point for these outcomes. We derived high-dimensional propensity scores using registry and Medicare data to control for patient factors and adjusted for provider factors in a Cox regression model comparing CAS with CEA. Among 5254 Society for Vascular Surgery's Vascular Registry (1999 CAS; 3255 CEA) and 4055 Carotid Artery Revascularization and Endarterectomy Registry (2824 CAS; 1231 CEA) Medicare patients, CAS patients had a higher comorbidity burden and were more likely to be at high surgical risk (Society for Vascular Surgery's Vascular Registry: 96.7% versus 44.5%; Carotid Artery Revascularization and Endarterectomy Registry: 71.3% versus 44.7%). Unadjusted outcome risks were higher for CAS. Mortality risks remained elevated for CAS after adjusting for patient-level factors (hazard ratio, 1.24; 95% confidence interval, 1.06-1.46). After further adjustment for provider factors, differences between CAS and CEA were attenuated or no longer present (hazard ratio for mortality, 1.13; 95% confidence interval, 0.94-1.37). Performance was comparable across subgroups defined by sex and degree of carotid stenosis, but there was a nonsignificant trend suggesting a higher risk of adverse outcomes in older (>80) and symptomatic patients undergoing CAS. CONCLUSIONS: Outcomes after CAS and CEA among Medicare beneficiaries were comparable after adjusting for both patient- and provider-level factors.

The time horizon matters: results of an exploratory study varying the timeframe in time trade-off and standard gamble utility elicitation.

INTRODUCTION: The purpose of this study was to examine whether the time horizon of time trade-off (TTO) and standard gamble (SG) utility assessment influences utility scores and discrimination between health states. METHODS: In two phases, UK general population participants rated three osteoarthritis health states in TTO and SG procedures with two time horizons: (1) 10-year and (2) a time horizon derived from self-reported additional life expectancy (ALE). The two time horizons were compared in terms of mean utilities and discrimination among health states. RESULTS: In Phase 1, the 10-year tasks were completed by 80 participants, 35 of whom also completed utility assessment with the ALE. In Phase 2, all 101 participants completed procedures with both time horizons. Utility scores tended to be lower with the ALE than the 10-year, a difference that was statistically significant for two health states with SG in Phase 1 (P < 0.05), two health states with TTO in Phase 2 (P < 0.01), and one health state with SG in Phase 2 (P < 0.001). In Phase 1, rates of discrimination between mild and moderate osteoarthritis health states were significantly higher with the ALE than the 10-year (TTO: P = 0.03; SG: P = 0.001). This pattern of discrimination was similar in Phase 2. DISCUSSION: Results suggest that the time horizon could influence utility scores and discrimination among health states. When designing utility evaluations, researchers should carefully consider the time horizon so that the value of health states is accurately represented in cost-utility models.

Relationship Between Physician and Hospital Procedure Volume and Mortality After Carotid Artery Stenting Among Medicare Beneficiaries.

BACKGROUND: Clinical trials demonstrated the efficacy of carotid artery stenting (CAS) relative to carotid endarterectomy when performed by physicians with demonstrated proficiency. It is unclear how CAS performance may be influenced by the diversity in CAS and non-CAS provider volumes in routine clinical practice. METHODS AND RESULTS: We linked Medicare claims to the Centers for Medicare and Medicaid Services' CAS Database (2005-2009). We assessed the association between 30-day mortality and past-year physician (0, 1-4, 5-9, 10-19, ≥20) and hospital (<10, 10-19, 20-39, ≥40) CAS volumes and past-year hospital coronary and peripheral stenting volumes (<200, 200-399, 400-849, ≥850) among beneficiaries at least 66 years of age. Unadjusted 30-day mortality risk was 1.8% (95% confidence interval [CI], 1.6-2.0) for 19 724 patients undergoing CAS by 2045 physicians in 729 hospitals. Median past-year CAS volume was 9 (interquartile range, 4-19) for physicians and 23 (interquartile range, 12-41) for hospitals. Compared to physicians performing ≥20 CAS in the past year, lower CAS volumes were associated with higher adjusted risks of 30-day morality (P value for trend < 0.05): 1.4 (95% CI, 0.9-2.3) for 0 past-year CAS, 1.3 (95% CI, 0.9-1.8) for 1 to 4, 1.1 (95% CI, 0.8-1.6) for 5 to 9, and 0.9 (95% CI, 0.7-1.4) for 10 to 19. An inverse relationship between 30-day mortality and past-year CAS hospital volume as well as past-year hospital non-CAS volume, past-year hospital non-CAS volume, and 30-day mortality was also noted. CONCLUSIONS: Among Medicare patients, an inverse relationship exists between physician and hospital CAS volumes and hospital non-CAS stenting volume and 30-day mortality, even after adjusting for all pertinent patient- and hospital-level factors.

Outcomes after carotid artery stenting in Medicare beneficiaries, 2005 to 2009.

IMPORTANCE: Despite increased carotid artery stenting (CAS) dissemination following the 2005 National Coverage Determination, to our knowledge, periprocedural and long-term outcomes have not been described among Medicare beneficiaries. OBJECTIVE: To describe the incidence of outcomes during and after the periprocedural period among Medicare beneficiaries undergoing CAS. DESIGN, SETTING, AND PARTICIPANTS: Observational study with a mean follow-up time of approximately 2 years among 22,516 fee-for-service Medicare beneficiaries at least 66 years old undergoing CAS (2005-2009) who were linked to the Centers for Medicare & Medicaid Services' CAS database. Database procedure dates were required to fall during a Medicare hospitalization for CAS. MAIN OUTCOMES AND MEASURES: Periprocedural (30-day) and long-term risks of mortality and stroke or transient ischemic attack, as well as periprocedural myocardial infarction. Subgroups were based on sociodemographic, clinical, and center-level factors, as well as the Stenting and Angioplasty With Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial or Carotid Revascularization Endarterectomy vs Stenting Trial (CREST) enrollment criteria. RESULTS: The mean patient age was 76.3 years, 60.5% were male, 93.8% were of white race, 91.2% were at high surgical risk, 47.4% were symptomatic, and 97.4% had carotid stenosis of at least 70%. Crude 30-day mortality, stroke or transient ischemic attack, and myocardial infarction risks were 1.7% (95% CI, 1.5%-1.8%), 3.3% (95% CI, 3.0%-3.5%), and 2.5% (95% CI, 2.3%-2.7%), respectively. Mortality during a mean follow-up time of 2 years was 32.0% (95% CI, 31.0%-33.0%), with rates of 37.3% (95% CI, 35.8%-38.7%) among symptomatic patients and 27.7% (95% CI, 26.4%-28.9%) among asymptomatic patients. Older age, symptomatic carotid stenosis, and nonelective hospital admission were associated with increased adjusted hazards of mortality and stroke or transient ischemic attack during and after the periprocedural period. The presence of a stroke center, government ownership, and a hospital bed capacity of 500 or more were associated with increased adjusted hazards of periprocedural mortality and stroke or transient ischemic attack. Few patients met the SAPPHIRE trial or CREST enrollment criteria primarily because physicians did not meet proficiency requirements either due to exceeding periprocedural complication trial thresholds or not meeting minimum CAS volume requirements. CONCLUSIONS AND RELEVANCE: Competing risks may limit the benefits of CAS in certain Medicare beneficiaries, particularly among older and symptomatic patients who have higher periprocedural and long-term mortality risks. The generalizability of trials like the SAPPHIRE or CREST to the Medicare population may be limited, underscoring the need to evaluate real-world effectiveness of carotid stenosis treatments.

Exploration of claims-based utilization measures for detecting potential nonmedical use of prescription drugs.

BACKGROUND: Nonmedical use of prescription medication is a significant and growing public health concern in the United States. Drug utilization measures that can reliably quantify the extent of nonmedical use of a given medication or medication class would greatly facilitate efforts to identify, monitor, and constrain nonmedical prescription drug use. Measures making use of prescription claims data would be especially valuable given the ready availability of these data among health care plans and payers. OBJECTIVE: To explore the extent to which claims-based utilization measures can provide information that aids in identifying and quantifying nonmedical drug use. METHODS: Prescription claims from a large employer-based administrative claims database (MarketScan) were used to evaluate drug utilization during the first year after an index prescription for 6 classes of drugs with a known abuse potential and 3 classes without. Traditional population-level measures of adherence (i.e., medication possession ratio [MPR] and proportion of days covered [PDC]) and a novel measure of overlapping days supply (MPR/PDC ratio) were calculated for all medications. Measures of asymmetrical use within a population were evaluated with the Lorenz curve, representing the total drug supply used by the heaviest 1%, 5%, and 50% of all users. All of the measures across compounds were compared with Spearman nonparametric rank correlations, and the Friedman's test was performed to determine whether the rankings were consistent in pairwise analysis. The ability of each measure to discriminate between abusable and nonabusable compounds was evaluated using the c-statistic. RESULTS: The study cohort included 6,291,810 patients, mean age 52 years, 57.9% female. The mean MPR and mean PDC for drugs with known abuse potential were both lower than for drugs without known abuse potential. The MPR/PDC ratio (MPR:PDC) ranged from 1.02-1.09. Highest values for the Lorenz-1 curve were seen for acetaminophen with codeine, acetaminophen with oxycodone, oxycodone, and acetaminophen with hydrocodone. The individual MPR and PDC were strongly correlated to each other (r = 0.99; P less than 0.001 for both), moderately correlated with the MPR:PDC (r = 0.62 and 0.57, respectively) and moderately inversely correlated with Lorenz 1 (r = -0.83 and -0.86, respectively, P less than 0.001 for both). The MPR:PDC was inversely correlated with the Lorenz-1 (r = -0.39; P = 0.02). After rank ordering individual drugs by each measure from highest to lowest value, the MPR:PDC resulted in consistent distributions with the individual MPR (P = 0.0097) and PDC (P = 0.0018), but not the Lorenz-1 (P = 0.0835), nor Lorenz-50 (P = 0.1343). The MPR, PDC, Lorenz-1, and Lorenz-50 were able to discriminate between the drugs with known abuse potential and those without (c-statistic 0.979 to 1), while the MPR:PDC was not (c-statistic 0.592). CONCLUSIONS: When comparing classes of drugs with a known abuse potential with classes of drugs not prone to such use, using an array of drug utilization measures, significantly different patterns emerge, although the ability of these measures to serve as reliable indicators of the extent of nonmedical prescribing may be limited. There is a need for valid and reliable algorithms to detect the extent of nonmedical use at a population level to help target public health interventions aimed at constraining illicit use. Further work is warranted on the development of novel measures that make use of individual patient-level use patterns.

Challenges to time trade-off utility assessment methods: when should you consider alternative approaches?

In recent years, the time trade-off (TTO) method, most commonly with a 10-year time horizon, has been the most frequently used approach for direct health state utility assessment, likely due to National Institute for Health and Care Excellence (NICE) preference for comparability with the EQ-5D, which has a utility scoring algorithm derived via this method. Although comparability to previous utility studies is important, there are situations when the TTO method may not be appropriate. The purpose of the current review is to highlight challenges to the TTO method. Five challenges to the TTO method are discussed: mild health states, small differences among health states, temporary health states, pediatric health states, and assessment of samples with particular characteristics. Some of these challenges are associated with the 10-year time horizon, while other situations may raise issues for TTO methods regardless of the time horizon. Alternative approaches for valuing health states are suggested.

Impact of caregiver and parenting status on time trade-off and standard gamble utility scores for health state descriptions.

BACKGROUND: The purpose of this study was to examine the effect of caregiver status on time trade-off (TTO) and standard gamble (SG) health state utility scores. Respondents were categorized as caregivers if they reported that either children or adults depended on them for care. METHODS: This study was a secondary analysis of data from three studies in which general population samples rated health state descriptions. Study 1: UK; four osteoarthritis health states. Study 2: UK; three adult ADHD health states. Study 3: US; 16 schizophrenia health states. All three studies included time trade-off assessment. Study 1 also included standard gamble. Descriptive statistics were calculated to examine willingness to trade in TTO or gamble in SG. Utilities for caregivers and non-caregivers were compared using t-tests and ANCOVA models. RESULTS: There were 364 respondents including 106 caregivers (n = 30, 47, and 29 in Studies 1, 2, and 3) and 258 non-caregivers. Most caregivers were parents of dependent children (78.3%). Compared to non-caregivers, caregivers had more responses at the ceiling (i.e., utility = 0.95), indicating less willingness to trade time or gamble. All utilities were higher for caregivers than non-caregivers (mean utility difference between groups: 0.07 to 0.16 in Study 1 TTO; 0.03 to 0.17 in Study 1 SG; 0.06 to 0.10 in Study 2 TTO; 0.11 to 0.22 in Study 3 TTO). These differences were statistically significant for at least two health states in each study (p < 0.05). Results of sensitivity analyses with two caregiver subgroups (parents of dependent children and parents of any child regardless of whether the child was still dependent) followed the same pattern as results of the primary analysis. The parent subgroups were generally less willing to trade time or gamble (i.e., resulting in higher utility scores) than comparison groups of non-parents. CONCLUSIONS: Results indicate that caregiver status, including being a parent, influences responses in time trade-off health state valuation. Caregivers (i.e., predominantly parents) were less willing than non-caregivers to trade time, resulting in higher utility scores. This pattern was consistent across multiple health states in three studies. Standard gamble results followed similar patterns, but with less consistent differences between groups. It may be useful to consider parenting/caregiving status when collecting, interpreting, or using utility data because this demographic variable could influence results.

Analysis of cognitive, functional, health service use, and cost trajectories prior to and following memory loss.

OBJECTIVES: This brief report examines whether significant changes in cognition, functional dependence, health service use, and out-of-pocket medical expenditures (OOPMD) occur in the years prior to a physician-identified memory problem in a nationally representative sample of older adults. METHOD: Longitudinal data from the RAND-Health and Retirement Survey were utilized. Those who reported a physician-identified memory problem (n = 387) were compared with a randomly selected control group of similar age, race, and gender who did not indicate a memory problem (n = 387). Multilevel linear models were used to construct trajectories for various measures of cognition, function, health service use, and OOPMD in the years prior to and following memory problem identification. RESULTS: Several trajectories demonstrated significant rates of change in the years leading up to a physician-identified memory problem, including symptoms (mental status, fine motor skills, and instrumental activities of daily living) and utilization (OOPMD and overnight stays in hospital). DISCUSSION: Preclinical declines in mental status and function and increases in hospital use and OOPMD are apparent prior to the formal identification of memory problems. Earlier identification of these changes might provide a basis for interventions that could alter the clinical course of dementia.

Biologic disease-modifying drug treatment patterns and associated costs for patients with rheumatoid Arthritis.

OBJECTIVE: To assess the influence of biologic treatment patterns on healthcare costs for patients with rheumatoid arthritis (RA) initiating tumor necrosis factor-α (TNF-α) antagonist therapy. METHODS: Patients with 2 RA diagnoses (International Classification of Diseases, 9th ed, 714.xx), and without psoriasis or Crohn's disease, were identified in a US employer-based insurance claims database. A sample of 2545 was constructed based on an index event of initiating TNF-α antagonist therapy and 30 months of continuous enrollment. Baseline characteristics were assessed in the 6-month pre-index period and treatment patterns were determined during the 12-month post-index period. Medical service and prescription drug costs were analyzed for Months 13-24 using multivariate regression analysis to control for baseline characteristics and time-varying confounding associated with treatment and disease severity. RESULTS: In the first year after TNF-α initiation, 89% used a single TNF-α antagonist; only 9% and 2% had switched TNF-α antagonists or received non-TNF biologic disease-modifying antirheumatic drugs, respectively. Descriptive analyses revealed pairwise differences between groups (p < 0.05) in baseline characteristics (comorbidities, RA-related procedure use, and prescription drug use). Controlling for observed baseline characteristics, costs were greater for those treated with multiple vs single TNF-α antagonists: annual RA-related prescription drug costs ($8,340 vs $7,058; p = 0.012), RA-related healthcare costs ($15,048 vs $13,312; p = 0.008), and total healthcare costs ($26,697 vs $21,381; p < 0.001). CONCLUSION: In this sample, the majority of patients with RA were treated with a single TNF-α antagonist over the first year on therapy. For those who switched therapy, Year 2 RA-related and total direct healthcare costs were higher, adjusting for claims-based measures of RA disease severity.

Does caregiver burden mediate the effects of behavioral disturbances on nursing home admission?

OBJECTIVES: The primary objective of this study was to determine whether caregiving burden mediated the relationship between specific behavior disturbances and time to nursing home admission (NHA) for persons with dementia (i.e., Alzheimer disease or a related disorder). DESIGN: The study used secondary longitudinal data from the Medicare Alzheimer's Disease Demonstration, a Medicare-covered home care benefit and case management program for family caregivers of persons with dementia. Primary caregivers of persons with dementia were assessed via in-person and telephone interviews every 6 months over a 3-year period. SETTING: Dementia caregivers were recruited from eight catchment areas throughout the United States. PARTICIPANTS: The baseline sample included 5,831 dementia caregivers. Just more than 40% (43.9%; N = 2,556) of persons with dementia permanently entered a nursing home during the 3-year study period. MEASUREMENTS: Individual behavior problems were measured with the Memory and Behavior Problem Checklist. Caregiving burden was assessed with a short version of the Zarit Burden Inventory. Key covariates, including sociodemographic background, functional status, and service utilization, were also considered. RESULTS: Event history analyses revealed that time-varying measures of caregiver burden fully mediated the relationship between four behavioral disturbances (episodes of combativeness, property destruction, repetitive questions, and reliving the past) and NHA. CONCLUSIONS: The findings highlight the multifaceted, complex pathway to NHA for persons with dementia and their family caregivers. The results emphasize the need for comprehensive treatment approaches that incorporate the burden of caregivers and the behavioral/psychiatric symptoms of persons with dementia simultaneously.

Illness-associated productivity costs among women with employer-sponsored insurance and newly diagnosed breast cancer.

OBJECTIVE: Determine lost work time and job attrition for incident breast cancer (BC). METHODS: The cases were employed women, aged 18 to 64, with BC identified by a validated algorithm between 1999 and 2005, from claims (MarketScan) and attendance databases. Controls without cancer were matched 3:1 on age, comorbidity, and index year. RESULTS: First-year mean disability days were 60 (cases, N = 880) versus 5 (controls, N = 2640) (P < 0.001). The first-year disability costs were $4900 for cases versus $385 for controls (P < 0.001). In years 2 through 4, the disability days and associated costs were similar for the cases versus controls. After 4 years, 56.4% of cases were still enrolled in the employer-sponsored insurance programs compared to 6.5% of controls (P < 0.001). CONCLUSIONS: The lost work associated with BC is substantial in the first year after diagnosis. Employee retention is much higher for BC cases versus controls.

A retrospective claims analysis of combination therapy in the treatment of adult attention-deficit/hyperactivity disorder (ADHD).

BACKGROUND: Combination therapy in managing psychiatric disorders is not uncommon. While combination therapy has been documented for depression and schizophrenia, little is known about combination therapy practices in managing attention-deficit/hyperactivity disorder (ADHD). This study seeks to quantify the combination use of ADHD medications and to understand predictors of combination therapy. METHODS: Prescription dispensing events were drawn from a U.S. national claims database including over 80 managed-care plans. Patients studied were age 18 or over with at least 1 medical claim with a diagnosis of ADHD (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] code 314.0), a pharmacy claim for ADHD medication during the study period July 2003 to June 2004, and continuous enrollment 6 months prior to and throughout the study period. Dispensing events were grouped into 6 categories: atomoxetine (ATX), long-acting stimulants (LAS), intermediate-acting stimulants (IAS), short-acting stimulants (SAS), bupropion (BUP), and Alpha-2 Adrenergic Agonists (A2A). Events were assigned to calendar months, and months with combined use from multiple categories within patient were identified. Predictors of combination therapy for LAS and for ATX were modeled for patients covered by commercial plans using logistic regression in a generalized estimating equations framework to adjust for within-patient correlation between months of observation. Factors included age, gender, presence of the hyperactive component of ADHD, prior diagnoses for psychiatric disorders, claims history of recent psychiatric visit, insurance plan type, and geographic region. RESULTS: There were 18,609 patients identified representing a total of 11,886 months of therapy with ATX; 40,949 months with LAS; 13,622 months with IAS; 38,141 months with SAS; 22,087 months with BUP; and 1,916 months with A2A. Combination therapy was present in 19.7% of continuing months (months after the first month of therapy) for ATX, 21.0% for LAS, 27.4% for IAS, 23.1% for SAS, 36.9% for BUP, and 53.0% for A2A.For patients receiving LAS, being age 25-44 or age 45 and older versus being 18-24 years old, seeing a psychiatrist, having comorbid depression, or having point-of-service coverage versus a Health Maintenance Organization (HMO) resulted in odds ratios significantly greater than 1, representing increased likelihood for combination therapy in managing adult ADHD.For patients receiving ATX, being age 25-44 or age 45 and older versus being 18-24 years old, seeing a psychiatrist, having a hyperactive component to ADHD, or having comorbid depression resulted in odds ratios significantly greater than 1, representing increased likelihood for combination therapy in managing adult ADHD. CONCLUSION: ATX and LAS are the most likely drugs to be used as monotherapy. Factors predicting combination use were similar for months in which ATX was used and for months in which LAS was used except that a hyperactive component to ADHD predicted increased combination use for ATX but not for LAS.

Patient-specific decision modeling to guide the use of drotrecogin alpha (activated) in patients with severe sepsis.

PURPOSE: The expected benefit of treating severe sepsis with drotrecogin alpha (activated) for an individual patient may depend upon several clinical factors including disease severity. Our objective was to create a decision support tool incorporating patient-specific inputs to estimate the balance between treatment risks and benefits for individual patients with severe sepsis. MATERIALS AND METHODS: Logistic regression models were developed to calculate patient-specific mortality risk with and without treatment, which were then used as inputs into a 75-state Markov model. Patient-specific inputs included patient age, sex, and 12 readily available clinical characteristics. RESULTS: The expected benefit from drotrecogin alpha (activated) treatment was most dependent upon the underlying disease severity. For example, for a 56-year-old white man with severe sepsis and a 28-day mortality risk of 29%, the model predicted a treatment-related gain of 1.2 quality-adjusted life years (17.3 vs 16.1). Probabilistic sensitivity analyses demonstrated that this patient would benefit from therapy 85% of the time. CONCLUSIONS: A customizable decision model using patient-specific inputs can be used to inform the treatment decision when considering the use of drotrecogin alpha (activated) therapy by weighing the risks vs the benefits of therapy in the treatment of severe sepsis.

Characteristics of patients initiating raloxifene compared to those initiating bisphosphonates.

BACKGROUND: Both raloxifene and bisphosphonates are indicated for the prevention and treatment of postmenopausal osteoporosis, however these medications have different efficacy and safety profiles. It is plausible that physicians would prescribe these agents to optimize the benefit/risk profile for individual patients. The objective of this study was to compare demographic and clinical characteristics of patients initiating raloxifene with those of patients initiating bisphosphonates for the prevention and treatment of osteoporosis. METHODS: This study was conducted using a retrospective cohort design. Female beneficiaries (45 years and older) with at least one claim for raloxifene or a bisphosphonate in 2003 through 2005 and continuous enrollment in the previous 12 months and subsequent 6 months were identified using a collection of large national commercial, Medicare supplemental, and Medicaid administrative claims databases (MarketScan). Patients were divided into two cohorts, a combined commercial/Medicare cohort and a Medicaid cohort. Within each cohort, characteristics (demographic, clinical, and resource utilization) of patients initiating raloxifene were compared to those of patients initiating bisphosphonate therapy. Group comparisons were made using chi-square tests for proportions of categorical measures and Wilcoxon rank-sum tests for continuous variables. Logistic regression was used to simultaneously examine factors independently associated with initiation of raloxifene versus a bisphosphonate. RESULTS: Within both the commercial/Medicare and Medicaid cohorts, raloxifene patients were younger, had fewer comorbid conditions, and fewer pre-existing fractures than bisphosphonate patients. Raloxifene patients in both cohorts were less likely to have had a bone mineral density (BMD) screening in the previous year than were bisphosphonate patients, and were also more likely to have used estrogen or estrogen/progestin therapy in the previous 12 months. These differences remained statistically significant in the multivariate model. CONCLUSION: In this sample of patients enrolled in commercial, Medicare, and Medicaid plans, patients who initiated raloxifene treatment differed from those initiating bisphosphonates. Raloxifene patients were younger, had better overall health status and appeared to be less likely to have risk factors for new osteoporotic fractures than bisphosphonate patients. Differences in the clinical profiles of these agents may impact prescribing decisions. Investigators using observational data to make comparisons of treatment outcomes associated with these medications should take these important differences in patient characteristics into consideration.

Application of a decision support tool for anticoagulation in patients with non-valvular atrial fibrillation.

BACKGROUND: Atrial fibrillation affects more than two million Americans and results in a fivefold increased rate of embolic strokes. The efficacy of adjusted dose warfarin is well documented, yet many patients are not receiving treatment consistent with guidelines. The use of a patient-specific computerized decision support tool may aid in closing the knowledge gap regarding the best treatment for a patient. METHODS: This retrospective, observational cohort analysis of 6,123 Ohio Medicaid patients used a patient-specific computerized decision support tool that automated the complex risk-benefit analysis for anticoagulation. Adverse outcomes included acute stroke, major gastrointestinal bleeding, and intracranial hemorrhage. Cox proportional hazards models were developed to compare the group of patients who received warfarin treatment with those who did not receive warfarin treatment, stratified by the decision support tool's recommendation. RESULTS: Our decision support tool recommended warfarin for 3,008 patients (49%); however, only 9.9% received warfarin. In patients for whom anticoagulation was recommended by the decision support tool, there was a trend towards a decreased hazard for stroke with actual warfarin treatment (hazard ratio 0.90) without significant increase in gastrointestinal hemorrhage (0.87). In contrast, in patients for whom the tool recommended no anticoagulation, receipt of warfarin was associated with statistically significant increased hazard of gastrointestinal bleeding (1.54, p = 0.03). CONCLUSIONS: We have shown that our atrial fibrillation decision support tool is a useful predictor of those at risk of major bleeding for whom anticoagulation may not necessarily be beneficial. It may aid in weighing the benefits versus risks of anticoagulation treatment.

Actual versus predicted first-year utilization patterns of teriparatide in patients with employer-sponsored health insurance.

OBJECTIVE: To characterize first-year utilization patterns of teriparatide derived from a claims database analysis versus predictions from an economic model. RESEARCH DESIGN AND METHODS: Claims data for actual teriparatide utilization were obtained from an integrated administrative database of approximately 3.4 million beneficiaries. A control group included patients with osteoporosis but without the use of teriparatide. An economic model, which relied on first-year market share projections, predicted the utilization of teriparatide from the demographic characteristics of the plan. Predictions were compared to actual utilization for eight health plans within the database. MAIN OUTCOME MEASURES: Demographic and clinical characteristics, number of teriparatide patients, and days of teriparatide therapy. RESULTS: Less than 1% of patients diagnosed with osteoporosis received teriparatide. Teriparatide-treated patients, compared to other patients with osteoporosis, were older and more likely to have experienced a previous fracture or to have received previous osteoporosis pharmacotherapy. For the combined 505 300 lives in the eight plans used for the comparative analysis, there were 134 teriparatide patients; the model predicted 131. For individual plans, the predictions varied in their accuracy. The greatest under-prediction for one plan was 17 patients (40 predicted vs. 57 actual), while the greatest over-prediction was 18 patients (34 predicted vs. 16 actual). For the other 6 plans, the predictions were within four patients of the actual number of teriparatide users. A similar pattern of differences was observed by comparing actual versus predicted days of teriparatide therapy across the eight plans. LIMITATIONS: Some clinical details of the actual patient cohorts, such as bone mineral density results, were not available in the database. The comparisons made between the teriparatide model predictions and actual utilization were based on analyses of a single model and do not speak to the broader issue of the accuracy of predictive economic models in general. CONCLUSIONS: Overall, first-year teriparatide utilization was relatively limited, consistent with model predictions. Predictions for individual plans varied in their accuracy.