Economic analysis of BRAF gene mutation testing in real world practice using claims data: costs of single gene versus panel tests in patients with lung cancer

Aims: To assess the time to BRAF testing, compare the characteristics of tested vs not-tested patients, and describe the costs for sequential vs next-generation sequencing (NGS) BRAF testing.

Methods: Patients diagnosed with lung cancer after December 1, 2013 were identified from two US claims databases; their characteristics were assessed during the 12 months before diagnosis (index date). Testing modalities were analyzed from the index date to end of continuous health plan enrollment or data availability (December 2015), based on combinations of Current Procedural Terminology (CPT) procedure codes. Time to BRAF testing was assessed using Kaplan-Meier analysis. Costs were analyzed from a payer’s perspective.

Results: A total of 28,011 patients newly-diagnosed with lung cancer were identified. Of them, 1,260 (4.5%) were tested for BRAF: 3.2% and 4.2% were tested at 6 and 12 months, respectively, after the index date. Compared to non-tested patients, tested patients were younger (58.3 vs 65.3 years; p?<?.001), had a lower Charlson Comorbidity Index (2.8 vs 2.9; p?=?.005), and a higher proportion had metastases (70.9% vs 43.4%; p?<?.001). In 76.0% of cases, BRAF was tested along with KRAS. BRAF was tested using NGS in 6.6% of cases. The average reimbursed amounts for the 10 most common CPT code combinations were $207–$2,074. Using the average costs of individual mutation tests, the total cost of sequential testing comprising KRAS, EGFR, ALK, ROS1, and BRAF tests was $3,763 ($464, $696, $1,070, $1,127, and $406, respectively), that of NGS was $2,860.

Limitations: Claims data did not include BRAF test results.

Conclusions: Among patients newly-diagnosed with lung cancer, 4.5% were tested for BRAF. Tested patients were younger and had a lower comorbidity burden, but more advanced disease. While reimbursed amounts varied greatly based on combinations of testing procedures, NGS testing was associated with cost savings compared to sequential testing of individual mutations.     

The Production Possibility Frontier as a Maximum Value Function: Concavity and Non-increasing Returns to Scale

This paper makes both a methodological as well as a substantive contribution to the literature on the concavity of the production possibility frontier (PPF). Rather than using the standard, calculus‐based techniques, the method here relies on the fact that the PPF is a maximum value function. The consequent simplification in analysis makes it possible to demonstrate that the conditions which are sufficient to guarantee global concavity of the PPF are considerably less stringent than those stated in the literature. Existing analyses have assumed that production functions are (a) concave and homothetic, or (b) display non‐increasing returns to scale (NIRS) and homogeneity. This paper shows that concavity without homotheticity, or NIRS and quasiconcavity without homogeneity (or even homotheticity) are sufficient, thus greatly increasing the generality of existing results. The analysis can be extended to include situations in which the input set includes industry‐specific factors.     

A note on Mill's theory of international values

This note provides an algebraic formulation of the analysis contained in section 7 of John Stuart Mill's chapter on international values. It is shown that Mill's results can be obtained as the solution to a pair of nonlinear programming problems. In addition, an error in Mill's analysis regarding the post-trade terms of exchange is pointed out.     

Decision rules for an investor in forward exchange markets

This paper obtains decision rules for a risk-averse, expected-utility-maximizing investor in forward exchange markets who may simultaneously combine covered arbitrage, spot and forward speculation. The introduction of margin requirements makes rates of return on all alternatives directly comparable, and causes initial wealth to constrain investment in all alternative. It is shown that combinations involving more than two alternatives need not be considered, and that a comparison of (expected) rates of return alone is sufficient to choose between the relevant combinations. The inadequacy of the ‘functional’ approach for analyzing individual decision-making is demonstrated.     

Estimation of innovation diffusion models with application to a consumer durable

A class of birth processes having a variety of practical applications in penetration of new services and products is considered. Typically, statistical inferences on these models are performed by means of simple error structures placed on the deterministic analogs of the underlying stochastic processes. Motivated by the poor performance of conventional estimation methods, the problem of estimating the parameters of these models is readdressed. We develop necessary formulae for performing the maximum likelihood estimation and weighted least squares estimation methods, and demonstrate their superiority through analyses of some real data and simulation studies.