Friday, August 26, 2016

John Dickerson defends his Ph.D. thesis at CMU, on kidney exchange

John Dickerson will defend today:
Computer Science Thesis Defense, Friday, August 26, 2016 - 2:00pm to 3:30pm
8102 Gates-Hillman Center (or, via Skype, for those of us who are far away).

Here's his summary of what he's preparing to defend:

"The exchange of indivisible goods without money addresses a variety of constrained economic settings where a medium of exchange — such as money — is considered inappropriate. Participants are either matched directly with another participant or, in more complex domains, in barter cycles and chains with many other participants before exchanging their endowed goods. This thesis addresses the design, analysis, and real-world fielding of dynamic matching markets and barter exchanges. We present new mathematical models for static and dynamic barter exchange that more accurately reflect reality, prove theoretical statements about the characteristics and behavior of these markets, and develop provably optimal market clearing algorithms for models of these markets that can be deployed in practice. We show that taking a holistic approach to balancing efficiency and fairness can often practically circumvent negative theoretical results. We support the theoretical claims made in this thesis with extensive experiments on data from the United Network for Organ Sharing (UNOS) Kidney Paired Donation Pilot Program, a large kidney exchange clearinghouse in the US with which we have been actively involved. Specifically, we study three competing dimensions found in both matching markets and barter exchange: uncertainty over the existence of possible trades (represented as edges in a graph constructed from participants in the market), balancing efficiency and fairness, and inherent dynamism. For each individual dimension, we provide new theoretical insights as to the effect on market efficiency and match composition of clearing markets under models that explicitly consider those dimensions. We support each theoretical construct with new optimization models and techniques, and validate them on simulated and real kidney exchange data. In the cases of edge failure and dynamic matching, where edges and vertices arrive and depart over time, our algorithms perform substantially better than the status quo deterministic myopic matching algorithms used in practice, and also scale to larger instance sizes than prior methods. In the fairness case, we empirically quantify the loss in system efficiency under a variety of equitable matching rules. Next, we combine all of the dimensions, along with high-level human-provided guidance, into a unified framework for learning to match in a general dynamic model. This framework, which we coin FutureMatch, takes as input a high-level objective (e.g., "maximize graft survival of transplants over time") decided on by experts, then automatically (i) learns based on data how to make this objective concrete and (ii) learns the "means" to accomplish this goal — a task that, in our experience, humans handle poorly. We validate FutureMatch on UNOS exchange data and make policy recommendations based on it. Finally, we present a model for liver exchange and a model for multi-organ exchange; for the latter, we show that it theoretically and empirically will result in greater social welfare than multiple individual exchanges. Thesis Committee: Tuomas Sandholm (Chair) Avrim Blum Zico Kolter Ariel Procaccia Craig Boutilier (Google/University of Toronto) Alvin Roth (Stanford University)"

He'll be teaching CS at Maryland in the Fall.

Thursday, August 25, 2016

Compensating bone marrow (blood stem cell) donors: still in legal limbo

Whether it will remain legal to compensate donors of bone marrow (blood stem cells) remains in limbo (see my various posts on the subject here).  The WSJ has an op-ed that summarizes the situation: 

Briefly, the 9th Circuit Court of Appeals lifted the ban on paying blood stem cell donors (if the technology was non-surgical), but the Department of Health and Human Services proposed a new regulation that would restore the ban. The regulation went out for public comment, and many comments were received, mostly against reinstating the ban.  The WSJ op-ed writes about that this way (in a way that makes me reflect on some of the oddities of news coverage):

"But a year after Ms. Flynn won her case, the Department of Health and Human Services announced that it might enact a regulation effectively nullifying the court’s ruling—and thus Ms. Flynn’s victory. In September 2013, HHS sought public comment. Hundreds of comments poured in favoring compensation for blood stem-cell donors who use apheresis, including support from Nobel Prize-winning economist Alvin Roth, who has long written on organ-donation policy. Only a handful of comments were opposed."

As you can imagine, I was one among many signers of the comment that I supported (which you can read here): the others, all economists, were
Theodore Bergstrom, University of California at S. Barbara, Stefano DellaVigna, University of
California at Berkeley, Julio J. Elias, Universidad del CEMA, Argentina,
Rodney Garratt, University of California at S. Barbara,
Michael Gibbs, University of Chicago, Judd Kessler, University of Pennsylvania, Nicola Lacetera,
University of Toronto, Stephen Leider, University of Michigan, John List, University of Chicago,
Mario Macis, Johns Hopkins University, Daniel McFadden, University of California at Berkeley, Matthew Rabin, University of California at Berkeley, Alvin Roth, Stanford University, Damien Sheehan-Connor, Wesleyan University, Robert Slonim, University of Sydney, Alex Tabarrok, George Mason University

If you have the time, you can read all 527 comments here.

Wednesday, August 24, 2016

Economists getting jobs as engineers

Bloomberg (Noah Smith) notices market design: All of a Sudden, Economists Are Getting Real Jobs

" Instead of holding forth on policy issues or the welfare of nations, many  are working with companies to create the kind of ideal markets that were previously confined to the pages of their academic papers. In other words, Keynes’ dream of economic dentistry -- or, more accurately, economic engineering -- might at last be coming true."

Tuesday, August 23, 2016

More on starting kidney exchange chains with deceased donor kidneys

Here's a forthcoming letter to the editor in the American Journal of Transplantation: We need to take the next step, by Marc L. Melcher, John P. Roberts, Alan B. Leichtman, Alvin E. Roth, and Michael A. Rees

It replies to another letter: A potential solution to make best use of living donor- deceased donor list exchange by VB Kute, HV Patel, PR Shah, PR Modi, VR Shah, HL Trivedi

which was prompted by our earlier article: Melcher, Marc L., John P. Roberts, Alan B. Leichtman, Alvin E. Roth, and Michael A. Rees, “Utilization of Deceased Donor Kidneys to Initiate Living Donor Chains,” American Journal of Transplantation, 16, 5, May 2016, 1367–1370.


Here's a post about that earlier article:

Using deceased donor kidneys to start living donor kidney exchange chains


and here's a post about the followup we hope to do:

Monday, August 22, 2016

Quality control of transplant centers, and the choice of who to transplant (and which organs to accept)

Transplant centers are regulated by measures such as their one-year graft-survival rate, so they feel pressure not to transplant patients, or organs, that have too high a risk to meet the required measure of success.

Here's a recent paper from the Journal of the American College of Surgeons that discusses some of the consequences:

Background

The central tenet of liver transplant organ allocation is to prioritize the sickest patients first. However, a 2007 Centers for Medicare and Medicaid Services regulatory policy, Conditions of Participation (COP), which mandates publically reported transplant center performance assessment and outcomes-based auditing, critically altered waitlist management and clinical decision making. We examine the extent to which COP implementation is associated with increased removal of the “sickest” patients from the liver transplant waitlist.

Study Design

This study included 90,765 adult (aged 18 years and older) deceased donor liver transplant candidates listed at 102 transplant centers from April 2002 through December 2012 (Scientific Registry of Transplant Recipients). We quantified the effect of COP implementation on trends in waitlist removal due to illness severity and 1-year post-transplant mortality using interrupted time series segmented Poisson regression analysis.

Results

We observed increasing trends in delisting due to illness severity in the setting of comparable demographic and clinical characteristics. Delisting abruptly increased by 16% at the time of COP implementation, and likelihood of being delisted continued to increase by 3% per quarter thereafter, without attenuation (p < 0.001). Results remained consistent after stratifying on key variables (ie, Model for End-Stage Liver Disease and age). The COP did not significantly impact 1-year post-transplant mortality (p = 0.38).

Conclusions

Although the 2007 Centers for Medicare and Medicaid Services COP policy was a quality initiative designed to improve patient outcomes, in reality, it failed to show beneficial effects in the liver transplant population. Patients who could potentially benefit from transplantation are increasingly being denied this lifesaving procedure while transplant mortality rates remain unaffected. Policy makers and clinicians should strive to balance candidate and recipient needs from a population-benefit perspective when designing performance metrics and during clinical decision making for patients on the waitlist.
It drew this headline in the news:
Hospitals are throwing out organs and denying transplants to meet federal standards

Sunday, August 21, 2016

Who Gets What and Why shortlisted for German Business Book Prize (to be announced in October)

Deutscher Wirtschaftsbuchpreis 2016: Die Shortlist

Google translates: Dusseldorf (ots) - The finalists of the German Business Book Prize have been announced: Ten books have made ​​it to the final round for 2016th A distinguished jury selected this year for the tenth time from the titles shortlisted the best business book of the year. The Executive Jury has Gabor Steingart, publisher of Handelsblatt. The prize will be awarded on 21 October at the Frankfurt Book Fair.
The German Business Book Award is themed "Understanding Business".Handelsblatt, the Frankfurt Book Fair and the investment bank Goldman Sachs award the prize to promote the economic literature. The three partners aim to emphasize the distinction the importance of economy section in mediating economic relationships. The selection criteria therefore include not only innovative agenda-setting or a new perspective and understanding and readability. The prize is endowed with 10,000 euros.
The ten books shortlisted provides Handelsblatt in the weeks prior to the literature page in the weekend edition. All other information on the award, the jury and of the initiators can be found at: www.deutscher-wirtschaftsbuchpreis.de

Die Shortlist 2016:
George Akerlof, Robert Shiller: Phishing for Fools. Manipulation und Täuschung in der freien Marktwirtschaft. Econ, Berlin 2016, 416 Seiten, 24 Euro
Adam Grant: Nonkonformisten. Warum Originalität die Welt bewegt. Droemer, München 2016, 384 Seiten, 22,99 Euro
Christoph Keese: Silicon Germany. Wie wir die digitale Transformation schaffen. Knaus, München 2016, 368 Seiten, 22,99 Euro
Paul Mason: Postkapitalismus. Grundrisse einer kommenden Ökonomie. Suhrkamp, Berlin 2016, 430 Seiten, 26,95 Euro
Alec Ross: Die Wirtschaftswelt der Zukunft. Plassen, Kulmbach 2016, 400 Seiten, 24,99 Euro
Alvin E. Roth: Wer kriegt was und warum? Bildung, Jobs und Partnerwahl: Wie Märkte funktionieren. Siedler, München 2016, 304 Seiten, 24,99 Euro
Wolfgang Schäuble (und Michel Sapin): Anders gemeinsam (im Gespräch mit Ulrich Wickert). Hoffmann und Campe, Hamburg 2016, 256 Seiten, 22 Euro
Mark C. Schneider: Volkswagen. Eine deutsche Geschichte. Berlin Verlag, 2016, 304 Seiten, 22 Euro
Hans-Werner Sinn: Der Euro. Von der Friedensidee zum Zankapfel. Hanser, München 2016, 560 Seiten, 24,90 Euro
Sahra Wagenknecht: Reichtum ohne Gier. Wie wir uns vor dem Kapitalismus retten. Campus, Frankfurt 2016, 292 Seiten, 19,95 Euro

Saturday, August 20, 2016

Differential privacy at Apple

The MIT Technology Review has an article about Apple's use of differential privacy, that caught my eye for several reasons: Apple’s New Privacy Technology May Pressure Competitors to Better Protect Our Data: The technology is almost a decade-old idea that’s finally coming to fruition.

"On a quarterly investor call last week, Apple CEO Tim Cook boasted that the technology would let his company “deliver the kinds of services we dream of without compromising on individual privacy.” Apple will initially use the technique to track trends in what people type and tap on their phones to improve its predictive keyboard and Spotlight search tool, without learning what exactly any individual typed or clicked.
...
“It’s exciting that things we knew how to do in principle are being embraced and widely deployed,” says Aaron Roth, an associate professor at University of Pennsylvania who has written a textbook on differential privacy. “Apple seems to be betting that by including privacy protections, and advertising that fact, they will make their product more attractive.”
In the version of differential privacy Apple is using, known as the local model, software on a person’s device adds noise to data before it is transmitted to Apple. The company never gets hold of the raw data. Its data scientists can still examine trends in how people use their phones by accounting for the noise, but are unable to tell anything about the specific activity of any one individual.
Apple is not the first technology giant to implement differential privacy. In 2014 Google released code for a system called RAPPOR that it uses to collect data from the Chrome Web browser using the local model of differential privacy. But Google has not promoted its use of the technology as aggressively as Apple, which has this year made a new effort to highlight its attention to privacy (see “Apple Rolls Out Privacy-Sensitive Artificial Intelligence”)."