Layman’s abstract for paper on exact sequential analysis for multiple weighted binomial end points

Every few days, we will be publishing layman’s abstracts of new articles from our prestigious portfolio of journals in statistics. The aim is to highlight the latest research to a broader audience in an accessible format.

The article featured today is from Statistics in Medicine, with the full article now available to read in early view here.

Silva, IR, Gagne, JJ, Najafzadeh, M, Kulldorff, M. Exact sequential analysis for multiple weighted binomial end points. Statistics in Medicine. 2019; 1– 12. doi: 10.1002/sim.8405

Sequential analysis can be used to prospectively monitor the safety and effectiveness of medications once they come onto the market. However, existing statistical methods for sequential analysis are based on a single outcome. Stakeholders may often be interested in how medications compare across a range of outcomes, including those indicating both benefits and harms. The authors developed a new flexible sequential statistical testing approach when multiple binomial outcomes are monitored. Each outcome is assigned a weight reflecting its severity. The sequential test indicate when there is sufficient evidence to conclude that the net benefit-harm profile of one treatment is superior to that of another treatment. The approach was demonstrated using two examples. In the first, the authors compared patients initiating rofecoxib (a drug that was withdrawn from the market for causing heart attack) to patients initiating other anti-inflammatory medications. The outcomes were heart attack and gastrointestinal bleeding, with the former having a larger weight. The second example compared a new medication for osteoporosis to an older medication, with a mix of effectiveness and safety endpoints – hip and pelvis fracture, forearm fracture, humerus fracture, serious infection, and pneumonia. The new sequential method has been implemented in the free R Sequential package and can be used in post-approval monitoring as well as in randomized clinical trial settings.