Fixing the flu vaccine supply chain

On-time delivery of influenza vaccine to the American public is hardly a straight shot from federal health officials to vaccine manufacturers to physicians to patients. A recurring and vexing part of the process is a supply-chain hitch that can leave patients waiting for flu shots even when the supply of the medicine is abundant.

zhangGuest Blogger: Fuqiang Zhang, associate professor of Operations & Manufacturing Management



A recent study by Tinglong Dai (Johns Hopkins University), Soo-Haeng Cho (Carnegie Mellon University), and myself offers a potential remedy by proposing a new kind of supply chain contract between vaccine manufacturers and the retailers (such as physicians’ offices and pharmacies) that purchase and dispense the shots to patients. We recommend an approach that they say would encourage a prompt and steady flow of vaccine from producers to dispensers, primarily through a pledge by the vaccine makers to buy back all unused doses and pay rebates for late deliveries.

Examining more than 20 years of vaccine-purchasing practice at Johns Hopkins Hospital, Dai, Cho, and I have written what we believe is the first paper about the commercial side of the flu vaccine supply chain. Previous research has looked at the interplay between the government and manufacturers. This study breaks new ground by considering the role of retailers as well.

The vaccine process begins several months before the first patients get their arms swabbed with alcohol, explain the authors. Around January 1, manufacturers are already looking ahead to next autumn’s flu season and starting production of a vaccine to match the expected strains. Their aim is to make a large quantity that will be ready to sell to retailers by vaccination season, which usually runs from late September to mid-November.

At best, however, the manufacturers make only an educated guess when they try to get a jump on a production regimen that routinely lasts about six months. The federal Food and Drug Administration, with input from the World Health Organization, selects the targeted strains of the coming flu season. But the agency doesn’t finalize them until February or March. If manufacturers have guessed wrong on any of the three or four chosen strains, they must restart that part of production.

flu shot syringeRetailers, aware of the manufacturers’ risk, anticipate potential delay in shipments and may commit to smaller orders than are ideal. They want to avoid requesting a large supply that could arrive in late fall, when public demand will be down, and then being left with unused doses. The manufacturers, knowing that retailers will follow this cautious strategy, are thus reluctant to produce a large amount before the FDA’s final announcement.

We describe this dance between the manufacturers and retailers as a classic example of a negative feedback loop. The result can be a shortage during the period of peak demand, as happened in 2014.

“It’s a coordination problem,” Dai said in an interview. “The parties involved are not behaving in the optimal way from the supply chain’s perspective because they’re trying to anticipate the other party’s moves.”

As we show, manufacturers have experimented in recent years with various types of contracts to promote both production and purchases. For example, one company offered full credits to retailers on returns of up to 25 percent of unused doses shipped before November 15, then up to 50 percent of unused doses shipped after that date; a subsequent contract by the same manufacturer bumped that date up to October 15. Another company ruled out returns or rebates for late deliveries one year, then extended a 10 percent rebate for returns of post-September 30 shipments the next year, and then went back to no returns or rebates the following year. Small wonder, Dai says, that the process is riddled with uncertainty and frustration.

Because the Buyback and Late Rebate (BLR) solution we propose is based on familiar aspects of industry practice and supply-chain theory, we believe it could be comfortably adopted by manufacturers and retailers alike. As we explain in our paper, buybacks occur in businesses involving time-sensitive products (such as newspapers, holiday goods, and fashion apparel). And while late rebates happen less frequently, they have been used in Walmart’s dealings with its overseas suppliers as well as in the vaccine business. However, we know of no other industry that employs the combined BLR approach.

“BLR is similar to those other practices, but with a slight twist,” Dai said. “What we propose is that the retailer would not be limited in his ability to return any unused product. He can return 100 percent of it, not just a specified fraction of it, and instead of the usual full refund, he would get a partial one. In addition, he would receive a partial rebate on all late deliveries.”

We believe that the combination of 100-percent buyback and rebates on late shipments would motivate retailers to commit to larger orders, which in turn would lead manufacturers to commit to greater and more prompt production. The negative feedback loop would be broken, and a steady, reliable flow of vaccine would be available, to the benefit of manufacturers, retailers, and patients.

faculty researchOur research paper “Contracting for On-Time Delivery in the U.S. Influenza Vaccine Supply Chain,” won the 2012 William Pierskalla Award for the Best Paper in Healthcare (runner-up) from the Institute for Operations Research and the Management Sciences (INFORMS), and an Honorary Mention for the 2013 College of Supply Chain Management Best Student Paper Competition from Production and Operations Management Society (POMS). Link to the paper.

Images: Top:Flu vaccine production – vials – packing line, Copyright Sanofi Pasteur – March 2009, Photographer: Vincent Moncorgé. Rights granted by the author for a period of 10 years for all use worldwide excluding paid advertisement. Above left: Flickr Creative Commons, Daniel Paquet, Flu Vaccination Grippe


In Business & Research
Tag , , , , , ,

2 Responses to "Fixing the flu vaccine supply chain"

  1. avatar Flu Guy

    This is a very innovative supply contract… But I have a couple questions/observations.

    1) How do you account for revenue recognition in this model? With 100% buy-back guarantee, what can be recognized as revenue upon receipt of shipment?

    2) This also assumes that all flu vaccines are the same. We have seen the rise of specialized flu vaccine offerings with clear differentiators and clinical benefits. My point here is that, some flu vaccines may be worth the wait — but only up to a certain point. Thoughts?


    • avatar Fuqiang Zhang

      Thanks for the comments. Regarding the first question, we think a vaccine manufacturer can use an estimated proportion of returns as the basis for revenue recognition. A relevant webpage is here:

      Regarding the second question, yes a more effective flu vaccine may be worth the wait. However, it must be carefully analyzed whether demand sensitivity correlates with the effectiveness of flu vaccines. If the demand windows of different vaccines are largely the same, then on-time delivery would still be of crucial importance. In addition, our model more or less captures different waiting behavior by using the fraction of customers who will come back for vaccination later.

Leave a Reply

Your email address will not be published. Required fields are marked *

Solve the equation: *