Pharma Faces More Comparisons of Drug Costs & Effectiveness

Will emerging drug cost assessment initiatives provide useful, valid information on total drug value? Or will new “scores” on drug cost-effectiveness merely justify more no-coverage decisions by insurers and health plans? These are troubling prospects for biopharma companies, as well as for patients and prescribers, as medical authorities and health care analysts look to more broadly incorporate drug costs into treatment evaluations and recommendations.

The latest to join the cost-effectiveness stampede is the National Comprehensive Cancer Network (NCCN), which recently announced that it would add a new “tool” to its widely used guidelines that advise oncologists on cancer treatment strategies. Up to now the NCCN program has based recommendations on cancer type, patient characteristics, and drug safety and effectiveness; now the scores will also include drug price, starting in October with therapies for multiple myeloma and chronic myeloid leukemia.

The NCCN initiative comes soon after an announcement by the American  Society of Clinical Oncology (ASCO) of its plan to establish a new framework to calculate “net health benefit” scores on cancer treatments based on a range of factors, including drug benefit, toxicity and cost.  

Drug makers fear that such calculations may ignore the larger treatment value issues and could limit patient access to effective medicines if mis-interpreted by patients, prescribers and payers. For one thing, these assessments may omit key factors, such as superior effectiveness of a new therapy for certain patient populations; how outcomes may improve patient quality of life even without extending longevity; and critical toxicity and side effects.  

In commenting on the ASCO plan, the National Pharmaceutical Council (NPC) offers a number of recommendations for evaluating cancer treatments. Benefit calculations should include overall and progression-free survival, particularly in situations where a clinical trial may be stopped early due to overwhelming benefits, advises NPC chief science officer Robert Dubois. Increased toxicity in both new and comparator treatments should be fully assessed, along with whether a therapy can enhance comfort for patients with advanced disease. And treatment value should include patient reported outcomes for quality of life; potential for long-term (vs. median) survival; and factors contributing to treatment adherence, such as route of administration, regimen complexity, and impact of side effects. So many factors affect the effectiveness of a test drug and a comparator in a clinical trial, says Dubois, as to make such assessments “meaningless” when taken out of context.

NPC also takes issue with ASCO’s plan to measure costs based on drug acquisition prices, noting that actual outlays for treatment can vary due to site of care, provider mark-ups, manufacturer discounts, the need for more or less medical services, and the potential for reduced side effects and other benefits. And instead of relying on data from randomized clinical trials, this kind of analysis should look more to real-world evidence from observational studies, which may be more relevant for patients and better reflect differences in individual response to treatment.

A number of NPC research projects aim to address these and other issues related to drug value. A lead initiative is to develop a more valid framework for payer assessment of the value of medical services, which will identify domains of value, options for measuring these domains, and models for integrating them into an overall assessment. Other projects are examining the ethics and effects of high co-pays in tiered pharmacy benefit plans; how payer cost-effectiveness policies may differ for medical procedures vs. biopharmaceuticals; and the role drugs play in managing high-cost patients  in commercial, Medicare and Medicaid plans. A related study will explore possible inequities from certain payers shouldering a greater share of long-term costs for treating chronic conditions.