Outcomes and healthcare resource utilization associated with medically attended hypoglycemia in older patients with type 2 diabetes initiating basal insulin in a US managed care setting.

OBJECTIVE: To assess health outcomes and the economic burden of hypoglycemia in older patients with type 2 diabetes initiating basal insulin (BI). RESEARCH DESIGN AND METHODS: Medicare Advantage claims data were extracted for patients with type 2 diabetes initiating BI and patients were stratified into two groups: those with medically attended hypoglycemia during the first year of BI treatment (HG group) and those without (non-HG group). Main outcome measures were hospitalization, mortality, healthcare utilization and costs 1 year before and 1 year after BI initiation. RESULTS: Of 31,035 patients included (mean age 72 years [SD 9.2]), 3066 (9.9%; HG group) experienced hypoglycemia during 1 year post-BI initiation. After adjustment for demographic, comorbidity and medication history, hypoglycemia was associated with risk of hospitalization (HR 1.59; 95% CI: 1.53-1.65) and death (HR 1.50; 95% CI: 1.40-1.60). Healthcare utilization was higher pre-index and showed greater increases post-BI initiation in the HG vs. the non-HG group. Per-patient healthcare costs were substantially higher for the HG group than the non-HG group, both pre-index ($54,057 vs. $30,249, respectively) and post-BI initiation ($75,398 vs. $27,753, respectively). CONCLUSIONS: Based on available claims data, hypoglycemia during the first year of BI treatment is associated with risk of hospitalization or death in older people, increasing healthcare utilization and costs. Due to the observational nature of this study, causality cannot be attributed, and further prospective studies into the effect of hypoglycemia on health outcomes in this population are warranted.

The peculiar economics of life-extending therapies: a review of costing methods in health economic evaluations in oncology.

Published literature lacks consensus, and most guidelines lack definitive recommendations as to whether cost-effectiveness analyses (CEAs) should include all "future" costs or distinguish between related and unrelated medical costs. This systematic review of oncology CEAs evaluated cost methods used and the impact on the cost-effectiveness of incorporating different cost categories, including costs due to study intervention, related medical costs of the treated condition, and unrelated medical costs. Of the 59 studies reviewed, none included medical costs unrelated to the treated condition and 14 studies (32%) excluded direct medical costs related to the condition but not the evaluated intervention. Recomputing ICERs using different cost categories altered overall cost-effectiveness conclusions. The authors propose conventional CEA methods may implicitly penalize therapies that add "expensive" life years for chronically ill patients. Presenting ICERs computed with and without disease-attributable costs can help better convey how much the treatment itself contributes to overall costs.

Systemic sclerosis prevalence and comorbidities in the US, 2001-2002.

OBJECTIVE: Large, population-based assessments of systemic sclerosis (SSc) prevalence and comorbidity in the United States (US) are rare. We explored autoimmune disease and other comorbidity patterns among SSc patients in the US from 2001 to 2002 and compared these with controls. RESEARCH DESIGN AND METHODS: Two US datasets with patient-level medical and drug claims were used to assess SSc prevalence and comorbidity: IMS Health Integrated Administrative Claims Database (IMS Health) and the MarketScan Commercial Claims and Encounters Database (MarketScan). SSc patients and comorbidities were identified by International Classification of Diseases (ICD), 9th revision diagnostic codes appearing on medical claims. Patients without SSc diagnostic codes (controls) were selected and matched 4:1 to SSc patients based on sex, age, Census Bureau region, and previous insurance coverage. The prevalence relative risk (RR) statistic compared comorbidity occurrence between SSc patients and controls, with 95% confidence intervals estimated using the Mantel-Haenszel method. Several sensitivity analyses tested methods used for identifying SSc cases and the prevalence of comorbidities. RESULTS: In both databases, SSc prevalence was 0.05% using the standard population model, 0.03% under sensitivity analysis. Among SSc patients the risks for inflammatory bowel disease (IBD) and multiple sclerosis (MS) were notably higher across datasets than for those without SSc: RR 3.2-6.6 for MS, RR 2.1-2.2 for IBD, in MarketScan and IMS Health, respectively (p < 0.05 for all). The chronic disease burden of SSc patients was much higher than that of controls, as confirmed by two chronicity measures (Chronic Disease Score, Elixhauser Comorbidity Index). The risks for cardiovascular, renal, liver and several neuropsychiatric diseases were higher for SSc patients across both datasets. Sensitivity analyses supported these findings. CONCLUSIONS: These data provide a population-based estimate of US prevalence of SSc and document the higher risk for certain other autoimmune diseases among SSc patients when compared to controls. Patients with SSc also had a higher chronic disease burden than those without SSc. These findings are limited by the unknown validity of ICD-9 codes for SSc case identification, unbalanced regional representation, and a likely 'healthy worker' effect in these databases.

Presenting results of probabilistic sensitivity analysis: the incremental benefit curve.

Cost-effectiveness acceptability curves have become a common way of presenting the results of probabilistic sensitivity analysis. However, these curves do not provide information on what the loss of welfare or net benefit (NB) is for cases where a given intervention is not the optimal one. We describe an alternate approach to presenting the results of probabilistic sensitivity analysis called the incremental benefit curve that presents the entire distribution of incremental NB of each intervention for a given WTP value. The incremental benefit curve provides the decision maker information regarding the potential welfare loss with a given intervention for scenarios in which it is not the optimal intervention, and thus would be a useful complement to the acceptability curve.

Overview of the pharmacoeconomics of pharmacogenetics.

Pharmacoeconomics and pharmacogenetics are two fields converging together as it is increasingly recognized that genetic markers predicting efficacy and toxicity to drugs can cost-effectively improve patient care. While pharmacogenetics aims at identifying genetic markers underlying the response to drugs, pharmacoeconomics aims at delivering healthcare cost-effectively. Several studies have investigated the potential cost-effectiveness of pharmacogenetic-based approaches. Recent evidences include screening for thiopurine methyltransferase gene polymorphisms to prevent azathioprine-induced myelosuppression, or screening for human leukocyte antigen (HLA)B5701 to prevent hypersensitivity reactions to abacavir therapy. Furthermore, examples suggesting a cost-effectiveness of markers predicting drug efficacy include screening the angiotensin-converting enzyme gene polymorphisms for statins therapy, the alpha-adducin gene variant for diuretic therapy and the assessment of human epidermal growth factor receptor (HER2) expression for trastuzumab therapy. However, thus far, all these pharmacoeconomic analyses are exploratory and validations in prospective randomized clinical trials are warranted.

Co-occurrence and comorbidities in patients with immune-mediated inflammatory disorders: an exploration using US healthcare claims data, 2001-2002.

OBJECTIVE: Research in immune-mediated inflammatory disorders (IMIDs) suggests that several diseases share disruptions in key cytokines. A common pathogenesis may present as similar patterns of disease co-occurrence and comorbidity, which could be observed through the analysis of healthcare claims datasets. METHODS: Adult patients continuously enrolled from 2001-2002 were identified in two US healthcare datasets containing medical and drug claims from health plans and self-insured employers. Patients with treatment records indicating an IMID were selected (e.g., rheumatoid arthritis, psoriasis, Crohn's disease); controls for each disorder were matched 3:1 based on age, gender, region, and previous insurance coverage. IMID cohorts and comorbidities were identified using International Classification of Diseases, 9th revision codes. Prevalence relative risk was used to assess co-occurrence and comorbidity rates in IMID cohorts and controls. Medical and drug utilization patterns were also explored. RESULTS: Findings were similar across the two datasets. IMID patients represented about 4% of the population; specific IMID prevalence matched the epidemiology literature. Patients with at least one IMID had a higher risk for another IMID when compared to controls. The risk for infectious, renal, liver, and ulcerative comorbidities was also elevated. Selected drug utilization patterns confirmed comorbidity findings. IMID patients used more healthcare resources compared to controls; findings were robust under sensitivity analyses. CONCLUSIONS: IMID patients were generally more likely than controls to have another IMID, supporting the concept that the diseases are related. These patients also had higher comorbidity rates. Findings may be limited by the nature of claims datasets and the confounding effect of current treatments. Prospective studies are needed to determine whether IMIDs have a common pathogenesis.

Optimal assignment of treatments to health states using a Markov decision model: an introduction to basic concepts.

Assessing the cost effectiveness of a new health intervention often requires modelling to estimate the impact of the intervention on cost, survival and quality of life over the lifetime of a cohort of patients. Markov modelling is a methodology that is commonly employed to estimate these long-term costs and benefits. As commonly used, these models assume that the patients continue to get the treatments assigned regardless of the change in health states. In this paper, we describe an extension to the Markov modelling approach, called Markov decision modelling. Such a model starts with a set of health states and treatments and optimally assigns treatments to each of the health states. A Markov decision model can be used to identify the optimal treatment strategy not just for the initial disease state, but also as the disease state changes over time. We present a dynamic programming approach to identifying the optimal assignment of treatments, and illustrate this methodology using an example. The Markov decision modelling approach provides an efficient way of identifying optimal assignment of treatments to health states, but, like the standard Markov model, may be of limited use when probabilities of future events depend on past history in a complex fashion. Even with its limitations, Markov decision models offer an opportunity for health economists to inform healthcare decision-makers on how to modify current treatment pathways to incorporate new treatments as they become available.

Pharmacogenomics and the evolution of healthcare : is it time for cost-effectiveness analysis at the individual level?

The efficacy and toxicity of any given drug can vary substantially from one individual to another. The heterogeneity in individual genetics contributes, in part, to this variability. Pharmacogenomics uses each patient's individual genetic information to identify the drug with the best efficacy-safety profile for that patient. However, heterogeneity is also present in individuals' preferences for alternate efficacy-safety profiles. We argue that as healthcare evolves towards individualised drug therapy, preference elicitation and cost-effectiveness analysis should also be performed at the individual level to maximise societal welfare.

Functional status and radiographic joint damage are associated with health economic outcomes in patients with rheumatoid arthritis.

OBJECTIVE: This analysis examines the relationship between the functional and radiographic measures of disease activity and the employment status in patients with rheumatoid arthritis (RA). We also assessed the influence of improvement in physical function on employability, healthcare costs, and quality of life, utilizing data collected in the ATTRACT trial. METHODS: During the ATTRACT trial, the Health Assessment Questionnaire (HAQ) disability index, radiographic damage measured by the van der Heijde modified Sharp (vdH-Sharp) score, employment status, healthcare resource utilization, and quality of life measured by Medical Outcomes Survey Short Form-36 were assessed at baseline and again periodically through Week 54. Clinically important improvement was defined as an improvement in the HAQ of > or = 0.25 from baseline to Week 54. RESULTS: There was a significant association at baseline between functional status and the percentages of patients employed. Increased radiographic joint damage was associated with lower full-time employment rate, with patients in the 2 highest quartiles (vdH-Sharp score > 51.5) of radiographic damage having lower rates of full-time employment than those with less damage. During the ATTRACT trial, patients who achieved a clinically important improvement in HAQ scores had a significant improvement in their employability (21% vs 3%; p < 0.001), in their time lost from work (7 vs 30 days; p = 0.012), in their total/direct medical costs (7093/6791 US dollars vs 11,712/10,039 US dollars; p < 0.001), and in their quality of life (p < 0.001) compared with those who did not demonstrate this improvement. CONCLUSION: Functional disability and radiographic joint damage are correlated with employment in patients with RA. Clinically important improvement in HAQ scores is associated with substantial health economic and quality of life benefits for patients with RA.

Conditions for the near equivalence of cost-effectiveness and cost-benefit analyses.

The equivalence of cost-effectiveness analysis (CEA) and cost-benefit analysis (CBA) has been vigorously debated in the health economic literature. In this paper we review and refine the conditions for the equivalence of CEA and CBA. The previously stated conditions require that 1) each individual's willingness to pay (WTP) per quality-adjusted life year (QALY) is constant and does not vary with the magnitude of QALY gains, and 2) the WTP per QALY is identical for every individual in society. Based on mathematical programming formulations of CEA and CBA, we note that condition 2 can be replaced with two other conditions, which together are less restrictive than the requirement that every individual have the same WTP per QALY. Even with this less restrictive set of conditions, CEA and CBA are unlikely to be equivalent under real world conditions. When CEA and CBA do lead to different resource allocation decisions, the most appropriate framework for health economic analysis depends on the perspective regarding distribution issues. We also examine the equivalence of two different definitions of CEA provided in the literature and discuss the problems that could arise when there are multiple optima.

Economic value of thrombolysis with adjunctive abciximab in patients with subacute peripheral arterial occlusion.

BACKGROUND AND OBJECTIVE: Glycoprotein (GP) IIb/IIIa receptor inhibitors enhance thrombolysis in patients with acute coronary syndromes. This analysis evaluates the economic impact of abciximab, a GP IIb/IIIa inhibitor, as an adjunct to urokinase in peripheral artery occlusions of less than 6 weeks duration. STUDY DESIGN: A post-hoc economic analysis was performed using clinical data and inpatient resource utilisation derived from the prospective comparative phase II Platelet Receptor Antibodies in Order to Manage Peripheral Artery Thrombosis (PROMPT) pilot study. Study endpoints were amputation-free survival and survival without open surgery or major amputation after 90 days, and the rate of major complications at 30 days. PERSPECTIVE: Third-party payer and the societal perspective. PATIENTS AND METHODS: Seventy patients with lower extremity thrombi were randomised (2 : 5 ratio) to urokinase plus placebo or to urokinase plus abciximab. Economically relevant data were retrospectively derived from the clinical study database from a specific evaluation of patient records and from expert opinion. RESULTS: From the viewpoint of the society, average total per-patient direct and indirect costs accruing over 3 months were more favourable for treatment with abciximab plus urokinase than for urokinase alone [9723 euros (EUR) vs EUR10 322; 2000 values], despite higher initial hospitalisation costs of the combination therapy. Abciximab plus urokinase was the dominant strategy at 3 months due to a clinically higher rate of survival without amputation or bypass surgery coupled with a lower average per-patient cost. From the perspective of the third-party payer, treatment with abciximab plus urokinase was economically also superior to urokinase alone (EUR8773 vs EUR9663). CONCLUSIONS: Based on the preliminary findings of the PROMPT trial, the use of abciximab as an adjunct to urokinase in patients with subacute peripheral artery occlusions may be the favourable strategy compared with urokinase alone, in terms of clinical and economic outcomes. Further trials are needed to confirm these clinical and economic findings. The preliminary clinical benefits experienced by patients treated with abciximab plus urokinase in the PROMPT trial translated into cost savings in terms of reduced direct medical costs at 3 months. These cost savings more than offset the cost of abciximab. The use of abciximab as an adjunct to urokinase in patients with subacute peripheral artery occlusions may be the favourable strategy compared with urokinase-alone in terms of clinical and economic outcomes, but further trials are needed to confirm the these clinical and economic findings.

Application of cost-effectiveness analysis to multiple products: a practical guide.

The appropriate interpretation of cost-effectiveness results and its use in treatment decision making have been debated vigorously. In this report we have summarized the decision rules described in the health economics literature to determine which intervention should be chosen from among multiple treatment options for a given disease, using a graphic framework. The implications of different means of expressing budget constraints on treatment choice are examined.