The seventh in our series of Friday 'Desert Island Reviews', featuring tutors from the MSc in EBHC (Systematic Reviews) programme, is from Dr Jeffrey Aronson .

Can you introduce yourself and share your role in the MSc EBHC (Systematic Reviews) program?

I’m Jeff Aronson. I qualified as a doctor in 1970 and went on to become a General Physician, working in hospital; my specialty is Clinical Pharmacology. When Kamal Mahtani asked me to write a blog for this series, I looked up all the systematic reviews in which I have taken part. To my surprise, I found that my resumé contains about 70 publications concerned in some way with systematic reviews. About half of those are published systematic reviews and most of the rest are the associated published protocols. Then there are about half a dozen papers that deal with aspects of the methods used in systematic reviews.

In most cases I have been involved in giving advice about different aspects of the use of medications. The methods papers have arisen mostly through discussions with colleagues about technical questions that arose while we were engaged in doing the systematic reviews.

Generally, others have done all the hard work in preparing the reviews, but in a few cases I have actually done the work myself: searching for the relevant publications, reading the titles and abstracts, selecting the papers to review, extracting the data, analysing them, and writing the paper. And those few reviews are my favourites.

 If you were stranded on a "Desert Island", which systematic review that you have led or been involved in might you take with you to read?

I have three favourite systematic reviews to chose from.

I published the first in 1975, and it was my first published paper. People had been doing systematic reviews for many years, but the idea of systematic reviews that included all published and unpublished data had not yet been proposed. Furthermore, Gene Glass had not yet invented the idea of meta-analysis and the word that went with it. So, I didn’t appreciate that what I was doing was a systematic review and meta-analysis. Indeed, today it wouldn’t be regarded as a proper systematic review, because I did it all on my own, without any help. And it wasn’t a proper meta-analysis, because I didn’t use modern statistical methods and I didn’t include forest plots—they weren’t invented until the 1980s and weren’t given the name until the 1990s. What I did was to review the use of plasma digoxin concentrations in monitoring digoxin therapy and specifically in the diagnosis of digoxin toxicity. I collected all the papers I could find on the subject and summarized the data. I did all the searching using reference lists in published papers and also whatever I could find listed in the Index Medicus printed catalogues, a total of 19 studies.

My second favourite systematic review was one that I did with another clinical pharmacologist, Robin Ferner, on the benefits and harms of laughter, which we published in the BMJ’s Xmas issue in December 2013. We called the review “MIRTH”, which stood for “Methodical Investigation of Risibility, Therapeutic and Harmful”, and it was reported in social media and newspapers worldwide. Although it was a whimsical account, including a lot of jokes, it also included a list of differential diagnoses of pathological laughter, which clinicians faced with that problem might find useful. Since 2018 it has had nearly 20,000 reads and over 1600 downloads; its Altmetrics score is over 800. One commentator reported "Never expected a systematic review to have me crying with laughter". There is also an article about it in the New York Times.

But my top favourite review is one that I began in 2008 and Igho Onakpoya continued, resulting in a paper that has been cited nearly 600 times. As the Editor from 1991 of 21 volumes of an annual book called Side Effects of Drugs Annual, each of which covered all the current published literature on adverse drug reactions, including 50 chapters and about 3500 references each year, I also included in each volume an annual essay on a topic of their choice by an invited author. One year I decided to write the essay myself and chose to review medicines that had been withdrawn from the market because of adverse drug reactions. When Igho took up the reins from me, he increased the number of cases we found from 280 or so to 462. We published the review in 2016 and followed it with six other papers on the subject, each covering a different aspect. I was also invited to write an editorial in the French journal Thérapie. Together those papers have been cited over 900 times.

What did your review show?

We identified 462 medicinal products that were withdrawn from the market between 1953 and 2013, the most common reason being hepatotoxicity, liver damage. The supporting evidence in 72% of cases consisted only of anecdotal reports, showing the importance of such reports, which people generally regard as being low down on the evidence scale. Only 43 of the drugs (9.3%) were withdrawn worldwide and 179 (39%) were withdrawn in one country only. Withdrawal was significantly less likely in Africa than in other continents. The delay between the launch of a drug and the first report of an adverse reaction that later occasioned withdrawal of the drug fell linearly with time. However, the delay between the first reported adverse reaction and the year of first withdrawal did not consistently shorten over time. I later described this, in my editorial, as pharmacovigilance success and regulatory failure.

We concluded that there were discrepancies in the patterns of withdrawal of medicinal products from the market when adverse reactions were suspected, and that withdrawals were inconsistent across countries.

We recommended that there should be greater co-ordination among drug regulatory authorities and increased transparency in reporting suspected adverse drug reactions.

What did you particularly enjoy about the review?

What I particularly enjoyed about doing this review was that it covered a wide range of medical topics, including analgesics and anti-obesity medicines, and a wide range of adverse drug reactions, including those affecting many different organs and in some cases deaths. It also had a simple message for the drug regulators. Although I don’t know whether they have taken that on board.

Reflecting on your reviews, what one lesson would you offer individuals embarking on a systematic review for the first time?

Whatever the question you want to answer in your review, don’t embark on the review on your own without help and advice from someone experienced in doing systematic reviews. There are courses on systematic reviews where you can learn more about how to do them and meet people you can work with.

Finally, If you were stranded on a "Desert Island" and about to read your review, what one food or drink treat would you bring with you?

I would bring a duck liver pâté with lots of brioche to spread it on and a half bottle of Château d'Yquem to wash it down. The English writer Sydney Smith once said that his idea of heaven was to eat pâté to the sound of trumpets; I wouldn’t need the trumpets.

 I would also take a book with me, to complement the review: How Science Takes Stock. The Story of Meta-Analysis by Morton Hunt (Russell Sage Foundation, 1997).

You can learn more about the MSc EBHC (Systematic Reviews) programme through the dedicated webpage or by contacting cpdhealth@conted.ox.ac.uk.

The views expressed in this blog are those of the author and do not necessarily represent the views of CEBM as a group.