To determine through a literature review how economic evaluations incorporate utility data into models when estimating long-term cost-effectiveness of lipid lowering therapies in primary and/or secondary cardiovascular disease (CVD) prevention. We used MEDLINE to identify published articles reporting cost-effectiveness models. Inclusion criteria included: English language articles published from 2000-2013, studies in adults ≥18 years old, studies assessing statins or ezetimibe for primary and/or secondary CVD prevention, and studies in the US, Canada, UK, Spain, Germany, Australia, Sweden, France, and Italy. Two researchers independently identified articles and disagreements were resolved by consensus. Sixty-one studies fulfilled the inclusion criteria. Forty-eight studies assessed the long-term cost-effectiveness (≥10 years) of lipid-lowering therapy— 20 studies were in primary CVD prevention, 19 were in secondary prevention, and 9 were in both primary and secondary prevention. 36/48 studies incorporated utility parameters into their models. 23/36 models differentiated short-term versus long-term utility impact of cardiovascular events. Eleven models were able to differentiate between short-term and long-term utilities through model structure, where separate health states were created for events versus post-events (e.g., stroke versus post-stroke), and utility values could be assigned to these health states accordingly. Twelve models were able to differentiate short-term versus long-term utilities by changing the model inputs, where different utility values were applied to a health state according to length of time after the event (e.g., event year versus subsequent years for a health state, or <6 months versus >6 months from time of event). Eleven models accounted for the occurrence of multiple events by applying disutilities or combining utilities multiplicatively for patients experiencing ≥1 event over time. Around two-thirds of the published long-term models differentiated short-term versus long-term utility impacts of cardiovascular events through model structure or utility inputs, which should be incorporated into future models on this topic.

Targeted therapies are hoped to deliver high-quality, effective treatments that control cost growth. Companion diagnostics (CDs) –biomarker tests to identify patients likely to benefit– are key to this potential. However, as the US's IOM has noted, reimbursement for CDs may not provide optimal incentives to develop critical CDs. To illustrate, we examined the cost:benefit of CDs based on current reimbursement levels and the clinical and economic benefit allowed by biomarker targeting. We identified 6 approved CD/therapeutic combinations, all in oncology. Several parameters were obtained: efficacy of therapeutic in indicated but otherwise un-targeted patients and in patients with the biomarker, therapeutic and diagnostic costs, and prevalence of the biomarker. CD clinical benefit was measured by the improvement in therapeutic efficacy in targeted versus untargeted patients. CD economic benefit was based on therapy cost avoided assuming that patients in a non-screened scenario undergo 1-month trial. To compare, we estimated a similar measure of the clinical cost:benefit for all oncology therapies approved since 2000. Estimated net economic benefit of CDs ranged from about $250 to $8,000. Estimated economic cost: benefit ranged from approximately $0.05 to $0.55 per USD saved. Estimates of the clinical cost:benefit of CDs ranged from approximately $1.50 to over $15 per one-percent improvement in clinical efficacy. Comparison oncology therapies are reimbursed at rates that imply an average clinical cost:benefit of about $750 per one-percent improvement in efficacy for non-OS benefits to $2400 per one-percent improvement in OS. Our calculations support the IOM statement that current reimbursement for CDs may not be optimal. Relative to the value placed on oncology therapeutics, the reimbursed value CDs is a small fraction of what would be expected under value-based pricing. This has implications for the structure of the CD industry as well as the potential for future innovations in diagnostics.

To investigate the relationship between persistence with drug therapy and treatment costs in patients with schizophrenia. Paid claims data from a commercial insurer was used to identify patients with schizophrenia. The unit of observation was the drug treatment episode. A total of 21,570 episodes were included in ordinary least squares (OLS) regression models of post-treatment cost as a function of alternative definitions of compliance adjusting for patient demographics, drug use history, prior medical care use, schizophrenia diagnosis and co-morbid medical conditions. Adding a second medication within one year is the key factor in this study. When compared to patients achieving one year of continuous therapy on the initial drug without switch or augment on a second drug; patients completing one year of therapy on their initial drug while adding an augmenting medication experienced increased drug costs by $1,565 (p<0.0001) and increased medical costs of $6,583(p<0.0001). Patients who did not achieve compliance on their initial medication, but did achieve compliance on their second medication exhibit higher total costs $11,381(p<0.0001). Patients using less than one year of psychotropic medication despite switching to or augmenting with a second drug experienced reduced drug costs, but also experienced increased medical costs. Similarly, patients who discontinued their initial therapy within one year but restart therapy using these medications also experienced reduced drug costs and increased medical costs. Finally, patients who quit their initial and any other therapy within one year save drug costs but also experience significantly higher medical costs. Compliance is associated with significant reductions in medical costs only if the initial drug regimen is maintained for 12 months. On the other hand, patients using less than one year of psychotropic medication experienced reduced drug costs relative to the most compliant patients(the comparison group), but also experienced increased medical costs.