Today Bernie Madoff pleaded guilty to the largest case of financial fraud in history. He was taken to jail in handcuffs and will likely spend the rest of his life in prison for pulling off a $65 billion Ponzi scheme. The tendrils of his deception have touched numerous people and organizations – wiping out personal fortunes, bankrupting charities, and probably leading to two suicides. Madoff is guilty of monumental sociopathic fraud.

But there is perhaps a bigger fraud story in the news this week also – not bigger in dollars but in the betrayal of trust. Anesthesiologist Scott S. Reuben, MD, of Baystate Medical Center in Springfield, MA is being accused of fabricating data on 21 published studies.

The Fraud

Reuben has been a leading researcher into what is called multimodal analgesia, specifically for post-operative pain management. Essentially, rather than simply sedating patients with narcotics, Reuben advocated using non-steroidal antiinflammatory analgesics (NSAIDS, or aspirin-like drugs) and neuropathic pain agents, like pregabalin. His published research showed the efficacy of this multimodal approach, which has been widely embraced.

The fraud came to light when Baystate could not find approval for human investigation to support two papers Reuben submitted last Spring. Essentially, he was presenting data that looked like he had performed research on human subjects, but there was no paperwork indicating that he had permission to do such research. It therefore came to light that this is because he never did the research – he just made up the data.

As is typical is such cases, once fraud comes to light and people take a closer look at the individual there is more fraud under the surface. When Reuben’s body of published research is viewed with a suspicious eye it appears that he has produced an unlikely number of soundly positive studies. Most researchers, when chance rather than fraud is operating, will produce a mix of positive, negative, and equivocal studies.

Further, Reuben has been accused of publication fraud. Evan Eckman, MD says his name was placed on two of Reuben’s papers even though he had no involvement with them.

In the end 21 of Reuben’s papers are considered likely fraudulent and have been or will soon be retracted by the journals who published them. None of his co-authors have been implicated in fraud, but I’m sure they are not happy about their affiliation with such fraud.

Two of the drugs that Reuben’s studies promoted are the COX2 inhibitor, Celebrex, and the neuropathic pain agent Lyrica – both produced by Pfizer, who partially funded Reuben’s research. Pfizer is not suspected of any wrongdoing, and they immediatly pulled any reference to Reuben’s studies from their promotional material when the fraud came to light.

The Implications

Like the Madoff scandal, the tendrils of this fraud go far and deep.  Reuben’s research was influential in the acceptance of multimodal analgesia. This acceptance now has to be reconsidered. The fraud does not mean that these treatments are ineffective or that multimodal analgesia is not a good idea. It just means that researchers have to go back to square one and think through the remaining data. Many studies will have to be replicated. A possibly legitimate approach to pain control has now been tainted by this fraud, and there is a big mess for other researchers to clean up.

Other anesthesiologists based their practice in part on Reuben’s data, which means patients were treated based upon fraudulent data.

Baystate has been hurt by this, as well as the publications that now have to retract published studies.

The Betrayal of Science

In addition to patients, doctors, journals, and his institution – Reuben has betrayed science and science-based medicine. And that is perhaps the deepest cut of all. The implications of that aspect of his betrayal is impossible to measure.

Science as a human endeavor depends upon honesty and transparency. Fraud like this cuts the heart out of the institution of science. It diminishes the very necessary trust that all scientific practitioners must place in the published research. It also diminishes the trust that the public places in the institutions of science. Further, the association with Big Pharma – while likely completely incidental in this case – will further fuel conspiracy theories and undermine faith in the medical profession as a whole.

Where do we go from here

When these stunning cases of scientific fraud come to light, such as the South Korean stem-cell fraud, it always raises the question of how to prevent such fraud from occurring in the future. I don’t think there is any system that can catch and eliminate all such fraud. There will always be a certain amount of fraud in science. But we want to minimize it.

Current peer-review practices are not really designed to catch fraud. Most cases of fraud are brought to light because someone blows the whistle, or because of some glaring and stupid mistake on the part of the individual committing the fraud – such as submitting data from patients you never had permission to study.

Another method for discovering fraud is that fraudulent work may not be in line with reality, and so it will not hold up to replication. A very famous fraud, that of Piltdown Man, was uncovered in this fashion. These fake pre-human fossils were out of step with later fossil discoveries, to the point that Piltdown Man was increasingly ignored as irrelevant to the real story of human evolution, and eventually just had to be reexamined and was indeed found to be fraudulent. By then the discovery of fraud was a relief, it meant paleontologists no longer had to account for the Piltdown anomaly.

In the end replication is probably the best hedge against fraud in science. The downside is that it may take a long time to work out. In medicine this may mean that treatment decisions are based for years on bad data, until the system has a chance to work out the real story.

This is exactly why I and others advocate for science-based medicine rather than just evidence-based medicine (although I see how this may seem counterintuitive at first). Science-based medicine considers prior plausibility, and recognizes the limitations of clinical evidence. Fraud is just one of many limitations of clinical evidence.

What this means is that when a finding does not make sense based upon prior plausibility the bar for evidence should be raised. This includes requiring independent replication as a hedge against fraud, or just incredibly sloppy research.

Another way to catch fraud is the to-good-to-be-true principle. If a researcher produces data that seems too clean, too good to be true and/or his research is always suspiciously positive, that should raise a red flag and prompt some review. The problem is there is no real mechanism for such review. Journals are typically only concerned with an individual study presented to them for publication – not the body of research by the author.

The community may detect the suspicious pattern, by which I mean the author’s colleagues or their institution. However, this does not seem to catch fraud in most cases. As with this case, suspicious patterns only come to light in retrospect, after fraud was detected by some other means.

Perhaps major journals, or professional organizations, need to add a layer of fraud detection to their procedures. This may seem like an inquisition (and if abused could become one) but if it is handled carefully and transparently, legitimate researchers should have nothing to fear. The minor inconvenience such a process would produce, if it really works as an early detection system for scientific fraud, could increase the reliability and reputation of the scientific process.

But it must also be recognized that it is impossible to completey eliminate all fraud. It is part of any human system. Systemic efforts are necessary to minimize fraud, which should also be punished severely.  But we need to expect that it will crop up every now and then. And this further means we need to keep our skepticism at an appropriate level – considering the plausibility of any claim, and witholding judgement until new ideas are adequately replicated.

__________________

Orac has also commented on this case.