Monthly Archives: August 2012

Polls About Elections: What Are They Worth?

Two weeks away from the general elections in the Netherlands, and a little over three months away from the US elections, we are bombarded with polls. At least one everyday, sometimes more. What do the findings of these polls tell us exactly? The perennial answer that you will hear from a scientist would be: it depends. Some polls are better than others. Jelke Bethlehem from Statistics Netherlands, the agency that collects official statistics in the country, has developed a very handy checklist that allows journalists to assess the value of polls. Unfortunately for you guys across the pond the checklist is in Dutch. So here’s a flow chart that I created with yED based on the checklist.

The flow chart helps you interpret the results of polls. You can flow through the chart in this Prezi.

My advice: use it every time you find a report about polls based on survey data. You will rarely – if ever – end up at the green fields after the FINISH. The quality of polls depends on so many factors, that the answer to the question “What do the findings of these polls tell us exactly?” for a journalist writing for the general public is: not very much.

1 Comment

Filed under data, methodology

Will a presumed consent system increase the number of organ transplantations?

Update, February 28, 2018: The Netherlands Senate (Eerste Kamer) has voted in favor of the opt-out system proposed in parliament. The vote was close: 38 senators were in favor, 36 voted against. The ministry of Health, Wellbeing and Sports announced that the law will be implemented in the summer of 2020. The law creates another natural experiment with changes in defaults, which will be interesting to monitor. The evidence reported from other countries in the post below suggests that the change will increase the number of organs available for transplantation.

Update, September 16, 2016: The Netherlands parliament (Tweede Kamer) has recently voted in favor of a law changing the opt-in system to an opt-out system. In December 2015 Wales shifted to an opt-out system. The first results are promising. A study that was published in 2014 also suggests that the change will increase the number of transplanted kidneys. Whether and when the change will take effect in the Netherlands depends on a vote in the House of Lords (Eerste Kamer).

In the Netherlands, post mortem organ donation still requires active registration of consent. The number of post mortem organ transplantations is low. Is there a connection between these two circumstances? Would changing the system to presumed consent in case of non-registration increase the number of organ transplantations?

These questions are at the heart of a new debate about organ donation, spurred by a proposal of the liberal democratic party (D66) to change the system, presuming consent and introducing an ‘opt out’ rule. It is believed that such a change will save lives by increasing the number of organ donors.

Is this true? Will changing to an opt-out system save lives? What do we know about the effect of presumed consent? You will not be surprised that researchers from various scientific disciplines have examined the effects of organ donation policies. But before I give some of the details of these studies, we need to carefully formulate the research question. In this case the question is: what should be the dependent variable in the debate?

In my view, what counts is the number of ‘lives saved’. The number of patients that successfully received an organ of a post mortem organ donor is the relevant outcome variable, and not so much the number of organ donors available. Obviously a presumed consent system increases the number of organ donors available for transplantation. Here’s a graph showing the ‘effective consent rate’ – i.e., the proportion of the population not declining consent – from a 2003 article by Johnson & Goldstein in Science

The problem is that presuming consent does not automatically increase the number of successful transplantations. You can see that the difference in the number of transplantations between countries with an opt in (voluntary registration) or opt out (presumed consent) system is much less dramatic.


Now about the independent variable. Many studies – including the Johnson & Goldstein article from which the graphs above are taken – have looked at differences between countries. These differences, however, cannot tell us whether a change in the system will save lives.  The only feasible policy recommendation we can make from an analysis of differences between countries is about migration: to which country should people move in order to have the best chance of receiving an organ from a post mortem organ donor? If we want to know the effect of a change in the system on the number of lives saved, we should look at changes over time in the number of successful transplantations in countries that have changed the system to an opt out / presumed consent system.

The benefits of an opt-in system seem less positive when we look at the changes in the numbers of lives saved in several countries that have recently changed to an opt-in system, as described in a 2008 article by Coppen and colleagues in BMC Health Services Research. Italy changed the system in 1999 and shows an increase in the organ donation rate. The increase cannot be attributed fully to the change to an opt-in system, however. The increase already started in 1995. Sweden changed to an opt-in system in 1996, but shows no increase in the organ donation rate in the years thereafter. The study covered 10 countries in the period 1995-2005.

A more recent study by Bilgel in the European Journal of Health Economics, covering a larger set of 28 countries and a longer time period (1993-2006), did find positive effects of system changes.  The study also shows that it is important whether family or next of kin is asked for consent and can veto donation, regardless of donor consent. The study shows that changing into an opt-out system works best if family/next of kin are routinely asked  for consent.

Previous debates on organ donation policy in Dutch parliament did not result in a system change. If the system remains unchanged, there is still room for improvement, as shown by a recent study from the UK by Wellesley in the British Medical Journal. Citizens may be ‘nudged’ into donation simply by more actively soliciting registration choices, for instance when applying for a new driver’s licence. Similar experiments have been conducted in the Netherlands, and they seem to have been successful.

The Netherlands have also been successful in increasing the number of living organ donors, as shown by a graph from a recent article in the Economist (below). These are mainly kidney transplantations by kin.

The problem, however, is that these nudges do not move the masses. The nudge is in the right direction, Wellesley acknowledged, immediately asking the follow up question: “but is it enough?” If only new applicants for driver’s licences are nudged into donation, it will take ages before the current waiting lists for organs are cleared. “It is time to move to presumed consent for organ donation”, Bird and Harris, also from the UK, concluded in a response to the Wellesley study. In the British Medical Journal, Rieu, an Oxford Ethicist, provides another good summary of the debate and comes to a similar conclusion.

The change to a presumed consent system may change the norm, also for family and kin who have the power to veto the deceased’s consent with donation. Citizens in countries with an opt-out system are more willing to consent with donation of another person’s organ, as shown in a 2008 study using survey data by Mossialos, Costa-Font and Rudisill.

Leave a comment

Filed under altruism, incentives, law, methodology, organ donation, policy evaluation