Richard Smith: Science and journalism threatened in the high court

richard_smith_2014I wrote this piece some six weeks ago after giving evidence in a libel case reported by The BMJ and published on 30 July 2015 . I’ve had to wait until the case was over to post the blog.

I’ve just finished giving evidence for a day and a half in the high court in Toronto. I enjoyed the experience, despite being cross examined for five hours. Fulfilling his duty, the barrister tried to discredit me and my evidence. I don’t think he succeeded, but it’s for the jury not me to decide. It’s a 15 year story that led to me being in court, and I want to share that story, describe my experience in court, and reflect on important implications for science and medicine.

The background to the case

The first part of the story has been told many times, so I’ll be brief.  My part in the story begins with submission in 2000 to The BMJ, when I was the editor, of a randomised trial by a single author that showed that a vitamin supplement improved the memory of healthy patients aged 65 and above. The paper came from R K Chandra, a Canadian researcher with a global reputation for his work in nutrition and immunology.

Alarm bells rang at once with the editor who read the paper. The paper was a follow up of one published in the Lancet eight years before. Why the delay? Why not a longer follow up? How had an expert on nutrition and immunology carried out such complicated psychometric tests? And how credible was it that a single author could have conducted the trial alone?

We asked Chandra’s university, the Memorial University of Newfoundland, to investigate. It came back quickly, saying there was no problem. We were unconvinced and asked it what form the investigation had taken; in particular, had investigators examined the raw data. After a long delay, the university eventually informed us that Chandra had resigned his position and left the country. It couldn’t do anything more.

The BMJ did not publish the study but in September 2001 it appeared in the journal Nutrition. We debated with The BMJ ethics committee whether we should notify Nutrition about our doubts and were advised that we should. The committee also advised us to contact the Lancet over our doubts about the study it had published. In 2005 Nutrition retracted the 2001 study. The Lancet has not retracted the 1992 study.

In an article in The BMJ in 2005 I discussed who had responsibility for investigating the previous studies of the author of a retracted study after the Memorial University of Newfoundland argued that it was The BMJ’s, not its, responsibility. I argued that it lay with the university, the employer, as it had the moral responsibility, the legal legitimacy, and could ensure due process.

Three television programmes

In 2006 CBC (Canadian Broadcasting Corporation) broadcast three programmes about Chandra. The programmes showed that there had long been doubts about the integrity of Chandra and that the Memorial University of Newfoundland had conducted an inquiry into Chandra before The BMJ ever contacted it, and found that Chandra had committed scientific misconduct. Chandra threatened to sue the university, as he threatened many, including The BMJ and me, and the university caved into the threats. In the programme a university spokesperson defended this action, saying: “Universities like ours are publicly funded. I think if Dr Chandra had been working in a different area of science where … let’s say, life and death were involved, if he were claiming a cure for cancer, let’s say, that was fraudulent … it’s possible that the university might have acted differently.” So much for universities being about the pursuit of truth and valuing integrity.

The programmes also found from his divorce case that Chandra had 120 bank accounts around the world that contained some $2m. He had had at least 12 companies and organisations in his name, including a foundation incorporated in Liechtenstein, a tax haven. Where Chandra’s money came from has never been made clear, but he has patented his multivitamin pills. He claimed losses of $125m in his action against CBC.

The third programme asked who should be acting on cases of misconduct and exposed the complete inadequacy of science to do so, a point of great public importance. I appeared in the programmes and thought them important and powerful. I also thought that there was something shameful about the mass media doing what science was failing to do.

The case against CBC

In 2011, five years after the programmes were broadcast, Chandra sued CBC and the journalists for libel and invasion of privacy. His total claim against CBC was $132m. He also took an action against the Memorial University of Newfoundland, but the university settled the case before it reached court.

My part in the case against CBC

The build up to the case was long. The lawyers acting for CBC contacted me to ask if in principle I would give evidence. I said yes without thinking about it much: the same principle that dictated that we act on possible misconduct led me to agree to give evidence. I saw it as an extension of the process. I had a duty to follow it through. I was also intrigued. Would I travel to Toronto? I know that a trial is a complicated game, and the lawyers were clear that an appearance in person would be much better than appearing on video. I didn’t think too much about exactly what an appearance would involve or how long I would be on the witness stand. “Typical,” my wife would say.

The next step was for me to tell my story to the lawyers. I also contacted The BMJ about releasing material from the files. The BMJ lawyer was understandably cautious: it was CBC, not The BMJ, that was being sued, and her duty was to minimise risk to The BMJ, which might well mean staying away from the fight. The lawyers got most of The BMJ material via Nutrition anyway.

I first spoke to the CBC lawyers perhaps a year before I appeared in court. They were building the case, telling their story as carefully, compellingly, and clearly as they could. The case would be heard before a jury, a factor that must have dictated the content, form, and style of the case. Evidence mattered more than anything else, but the lawyers must have had to think hard about whom they wanted to give evidence. Some people–for example, the CBC journalists–would have to give evidence, but others were optional. It must be like casting a play. Would the witness convince, appeal to the jury, or be more of a liability than asset?

I had some benefits. I’d been the editor of The BMJ, surely somebody who could be trusted, even an authority of some sort. I have a CBE, which is mostly an embarrassment to me, but perhaps that might give me greater credibility in a court where they bow to the Queen at the beginning of the day. Maybe my English accent (even with its Cockney twang, a problem in Britain but indiscernible to Canadians) would be an advantage: in the US people tend to overestimate your smartness if you speak with an English accent, and the lawyers thought that might also be the case in Canada. But I had risks too: I’m a notorious, loose cannon. I wondered if the opposing lawyers might bring up my “disgraceful” blog posted on 31 December 2014 that argued that dying of cancer was the best way to die. (They didn’t.)

The next problem was finding a date. When would the trial begin? How long would the prosecuting lawyers take to make their case? Might it even be that Chandra or his lawyers would “throw in the towel” before I had to give evidence? I kept giving dates to the lawyers, but inevitably my diary began to fill.

At last a date

Eventually it became clear when I might give evidence just a few days before I was due to do so. I would fly over on a Sunday, give evidence on the Monday, and hope to fly back that night. Inevitably it was uncertain: the defence lawyer thought she could get through my “evidence in chief” in two hours, but she didn’t know how long cross examination would take or what they would ask.

I’d been sent a file of papers in advance, including a transcript of the programmes, and I’d read through them (on a train to Bristol). The transcript especially made fascinating reading; how, I wondered, could Chandra have the chutzpah to sue, when the case against him was so broad, deep, and damning? I pondered on this and reached an answer: he had the money, and he had nothing to lose. Even if there was a small chance of him winning it was worth taking. Plus he could see his tormentors, like me, squirm when cross examined, although I don’t think I did squirm.

While in London I had a call with the lawyer to “prep” me, outline the way she would present my evidence and the questions she would ask me. She emphasised that I must say only what I thought, and she was scrupulous in telling me nothing of the evidence that had been presented in the eight weeks the trial had already been underway.

Meeting the lawyer

Eventually I was on a plane to Canada. It was delayed. I was in bed and asleep by 8.15 pm and awake at 4.15 am. I had to be at the lawyer’s office by 7 am, an office way up in the clouds in one of the highest buildings in Toronto. She was starting at 6 am every morning during the trial, talking to witnesses and preparing her case. It was, I realised, an exhausting business, some 12 hours every day, with most of those hours being “on stage” or necessitating close attention to all that was happening.

We met in her office, and she took me through how she would present my evidence. Some of the questions would be open, but like a good lawyer she knew what I would say. We mustn’t go too fast or we would lose the jury; she thought it would take about two hours. Then the prosecuting lawyers would cross examine me. They would be out to destroy my evidence and credibility. That was their job. We speculated on what they might ask, but mostly it was guesswork.

She explained to me that everybody stood up when the judge and then the jury arrived. I had to decide whether to swear on the Bible or affirm (what atheists do). Once I was “in cross,” legal jargon for being cross examined, I could not talk to her, or anybody involved in the case, at all. She hoped that I’d be finished today and could fly home tonight, but it was uncertain.

At the court

We walked up to the court building, her assistant joining us. The lead lawyer wheeled a substantial pile of boxes. I offered to help: “litigators,” she told me, “transport their own evidence.” At the court the lawyers changed into their uniform of pinstripe trousers, white bow, and black gown. The thinking, the lawyer explained, is that all lawyers look the same in court, none grander than others. “But we don’t go as far as wigs.”

I waited while they changed, and then went up to court. I saw where I must perform. The witness box was right beside the judge’s bench, but he was higher. The court reporter would be immediately in front of me, and the clerk of the court beside her, immediately in front of the judge. The jury’s seats were beside me to my left, presumably close to the witness so that they could see if he or she was lying. The lawyers’ benches were in front of the judge, well below him, and the lawyer asking the questions stood at a podium behind the lawyers’ benches. The public benches were behind all that.

The court assembled, waiting for the judge and the jury. I sat on the public benches with only one or two others. The judge appeared just after 10 am. We all stood. He bowed to us, and we to him. The judge was British in origin and smart and reasonable, according to the CBC lawyer.

We sat again and waited for the jury. We stood again, and they, seven ordinary looking people, as they should be, trooped in. I was told that it took a long time to get a jury, partly because the case was expected to take around 11 weeks. A jury in a civil case has six members, and the court accepts a judgment of five to one. Originally there were eight members of the jury, two spares. One had been lost. Losing three would mean starting again.

The CBC lawyer had made an attempt to have the jury dismissed, arguing the case was too complex, but had failed. I imagine that she thought the judgement less predictable with a jury. Chandra’s lawyers had opted for a jury, perhaps for the reverse reason. The jury would also presumably increase the cost of the trial, so increasing the stakes.

In the witness box

The registrar made a grand statement about the court of Ontario being in session. The oldest of Chandra’s lawyers then announced that they had completed the prosecution case. If they had had more evidence to present, then I’d have had to wait longer. But they didn’t, and I was called to the witness box. I affirmed, and we were off. I essentially had to simply tell the story of my and The BMJ’s involvement in the story–after a few preliminaries about my background. My story was told largely through documents, letters I’d written years ago, that were flashed up on a screen. It’s perhaps worth remembering when you send an email that one day it might be shown in court. None of mine were that embarrassing.

I was calm, and the lawyer was skilful. Pace was important: we mustn’t waste time, but we mustn’t go too fast for the jury. She had said it would take about two hours, and it did. How, I wondered, did I sound to the jury: an honest editor worried about research integrity? Or a fanatic pursuing a witch hunt? Chandra’s lawyer interrupted a few times over bits of evidence and the nature of the questions. It was routine tussling, although the jury and I had to leave for some 10 minutes while the lawyers quibbled over a point.
One crucial point was to make clear that I was not paid. I was not an expert witness. I was part of the case, but at the same time as an ex-editor involved in many cases of scientific misconduct and with a longstanding interest in the subject I was expert. But, crucially, I was not a hired hand: I was speaking for myself, saying only what I believed.

Cross examination

After two hours, at 12.30 pm, my cross examination began. I wasn’t examined by the oldest lawyer, who, it was hinted, had suffered a setback, but by a thin, bald headed lawyer of about 40. He had something of a sneering attitude. I remembered being told that good lawyers could make anybody look an idiot if they wanted. It was his job to discredit me and my evidence, but he mustn’t be too aggressive for fear of making the jury sympathetic to me.

He started by asking me questions about myself that didn’t seem to lead anywhere. It was as if he was trying to find his line of attack. We didn’t seem to get anywhere before 1pm, when it was lunch. It was agreed before lunch that we would come back early, at 2 pm rather than 2.15 pm, with the aim of me being able to fly home that night.

We went through the rigmarole of standing for the jury leaving, sitting, and then standing again as the judge left. “In cross,” in the lawyers’ jargon, I had to lunch alone.
We began again at 2 pm, and Chandra’s lawyer launched his attack. One line was that I was no psychiatrist, psychologist, immunologist, or nutritionist, so what did I know relevant to the study? I repeated endlessly that I hadn’t acted alone, but that we were a team and had expert advice. Another line was that I had no right to challenge the Memorial University of Newfoundland on the nature of its investigation, but this didn’t seem a profitable line as the study was eventually retracted, and it must be apparent to everybody that if a simple supplement of vitamins and minerals could really reverse dementia (as the study seemed to show once the mental state of the “healthy” people was assessed) then it would be a major breakthrough and we’d know by now.

Chandra’s lawyer made much of the fact that Nutrition had never used the words fraud or research misconduct, but I suggested that that was because of fear of lawyers and that the retraction included such major doubts that no other reason apart from fraud could be imagined.

There was much discussion around the significance of missing data: “Could missing data not be found, Dr Smith?”

“I suppose that anything that is lost can be found,” I answered, adding unwisely “except perhaps a soul.” The judge sniggered. (Luckily I didn’t say “virginity,” which occurs to me now.) My silly, throwaway comment showed that I was relaxed. I had to concentrate hard, but I didn’t ever become anxious.

The revelation was that the data had been found, but nobody seemed to have examined them. I pointed out that they could be invented after the paper was published.
During the afternoon the CBC lawyer reminded Chandra’s lawyer and the judge that I hoped to be able to fly home, and Chandra’s lawyer said he hoped to finish. But when we came close to 4 pm, the finishing time, Chandra’s lawyer said that he needed another hour.

The cost of the trial

As it was, I left the witness box and had to leave the court without speaking to anybody. CBC’s lawyer said that my ticket would have to be changed and that that might require some administrative contact. As it happened, it didn’t. I wandered through the streets of Toronto, reflecting on the cost of the trial. Seven lawyers, a judge, seven jury members, and three or four court officials in court every day for around 12 weeks must cost millions. Then there were years of preparation, preliminary skirmishes in court, vast amounts of paperwork, expert witnesses, and people being flown across the world to give evidence.

I spent a pleasant evening eating beside Lake Ontario and watching boats move in and out of the mist, and by 9.45 am the next day I was back in court ready to continue my evidence.

A second day in the witness box

After the customary standing and bowing I was back being cross examined. Chandra’s lawyer seemed on better form than the previous day.

He questioned me on The BMJ informing the Lancet about our doubts about the earlier paper that Chandra had published in the Lancet in 1992. The Lancet had not found any problem with the paper and had not retracted it. Did this not mean that the paper was not fraudulent and that The BMJ had been wrong to argue that the later paper was fraudulent? I said that it was hard to prove a paper fraudulent simply by examining what was published and that the Lancet had presumably encountered the same problem as The BMJ in having the study investigated by the university. In short, just because the paper was not retracted did not mean it was not fraudulent.

Then the lawyer moved to the rapid responses to The BMJ’s article on Chandra. There were many, and most were supportive of Chandra. The lawyer read quotes from the responses. I pointed out that we had a very low threshold for posting rapid responses and that we couldn’t even be sure that the respondents existed. Chandra might have prompted people to write and might have sent responses himself under an assumed name. The responses, I suggested, were a very low form of evidence and should not be taken seriously.

Some of the rapid responses were from R B Singh, an Indian researcher, the quality of whose research and and The BMJ had also called into question. Was it coincidence that the two scientists I’d pursued were both Indian? Was I not a racist, pursuing these people because of my racism rather than any concern for the integrity of science? I knew this was coming because R B Singh had made the same accusation, including on the front page of the Wall Street Journal. I didn’t attempt to prove that I wasn’t a racist but said that I had complained about many other researchers, and simply two people being Indian was not much of a coincidence.

Chandra’s lawyer asked me if I was accusing Chandra of being a serial killer. In the programme I said: “I think what happens with these people who have been producing fraudulent studies for a very long time is that just—they get away with it so often, they get more and more blasé about it, a little bit like, in my mind, serial killers. It’s almost as if they eventually want to be found out. They get bolder and bolder about it.” I answered that I wasn’t accusing Chandra of being a serial killer, and the lawyer asked me if I’d like to apologise to him if I’d given that impression. I did.

But wasn’t the comparison, the lawyer continued, too extreme? I said that it might be, but I thought that the comparison held. I explained that I probably had in my mind Harold Shipman, a British general practitioner who had murdered hundreds of his patients and become ever more careless about doing so.

Chandra’s lawyer took two and a half hours not the one hour that had been mentioned on the previous day, and so I spent another lonely break.

As we came to the end of my cross examination Chandra’s lawyer produced a document that showed that the CBC had commissioned a statistician to examine 14 papers by Chandra and that the statistician had not found evidence of fraud? Did I not think that that information should have been included in the CBC’s programmes? The CBC lawyer leapt up and said that she didn’t think this question admissible. I hadn’t seen this document in advance, and there seemed to be some debate about the legitimacy of the document. After some debate Chandra’s lawyer asked me the question in some form, and I said that I couldn’t answer the question because I didn’t know the totality of the evidence assembled by CBC and so couldn’t comment on what was included and what was not.

After five and a half hours and a whole legal day my cross examination was over. CBC’s lawyer had the chance to ask me questions to reverse any “accidents” (as she had put it earlier) and she asked me again the very last question about what should and should not have been included in the programme–and I gave the same answer about not being able to answer it as I didn’t know the totality of the evidence. It felt as if that bit of evidence might be crucial.


I left the courtroom, and the CBC reporter and producer came and thanked me for giving evidence. The lawyer came out too and said I’d done well. I said that television people always said that even if you were useless and asked if the lawyers were the same. She said not. I felt that I had done well, but I was far from sure.

The importance of the case for journalism and science

It’s more than 10 years since CBC journalists started working on the three programmes about Chandra. They had exposed some outrageous behaviour and raised a crucial point about guarding the integrity of science. As a consequence they have had to live with a serious threat for 10 years, a threat that could destroy their reputation, jobs, and income. I was sure of the rightness and bravery of what they had done, but there is always uncertainty in a legal case. The case was also threatening to CBC, a corporation that like the BBC is under attack. Losing the case and having to pay millions in costs and damages would harm the corporation.

But the case mattered as well for science. The evidence presented in the case showed how the university and journals had been intimidated by legal threats. Chandra had used the law to sustain his reputation long after he had begun to behave fraudulently. If the case was lost then universities, publishers, and editors might become even more cautious about confronting fraud—when they are too cautious already.

I flew home hoping very much that CBC would win the case but recognising the inevitable uncertainty.

On Friday 23 July the jury found for the CBC, that what was said in the programme was true, and that the journalists had not invaded Chandra’s privacy.

Competing interest: Richard Smith had his expenses paid to attend the trial but was not paid a fee.