Large cohort studies of people with a first-episode psychosis provide a unique opportunity for finding out why so many young people with schizophrenia spectrum disorders die at a young age. However, it seems that those psychiatrists who have access to the mortality data generally do not want the facts to come out. In published cohort studies, there is virtually always too little information or no information at all about the causes of death.
In 2012, Hegelstad et al. published 10-year follow-up data for 281 patients with a first-episode psychosis (the TIPS study). 1 Although their average age at entry into the study was only 29 years, 49 patients (17%) had died in less than 10 years. The authors’ detailed article was about recovery and symptom scores and they took no interest in all these deaths, which appeared in a flowchart of patients lost to follow-up and were not commented upon anywhere in their paper. In the text, however, they mentioned only 28 deaths (11%), so it is difficult to know how many died.
On March 5th 2017, I wrote to the first author: “In your interesting study from 2012, ‘Long-Term Follow-Up of the TIPS Early Detection in Psychosis Study: Effects on 10-Year Outcome’, 28 patients died of 281. Do you have the causes of death, which would be very interesting to know about? I noticed that most patients were still on antipsychotics 10 years after they started (table 2). I consider this very frightening, e.g. around half of them will have tardive dyskinesia (which is often masked by ongoing treatment) and many, if not all, will have developed permanent brain damage at this point, to mention just two of the important harms of long-term treatment. In contrast, only 17% of patients in Lappland were still on antipsychotics 5 years after their first episode (compared to 75% in Stockholm).”
I sent a reminder ten days later and was told I would get an answer shortly. On May 10th, I wrote again: “It is now another two months. You did an important study and it is important for the world to know what 10% of your very young patients died from in just 10 years. It is a frightening death rate. Do we need to file a Freedom of Information request to get this information?”
Hegelstad replied that, “We are preparing a manuscript detailing the information you are asking for. It will be submitted to a peer-reviewed journal. When published the information will be readily available to all.”
The preparation of the manuscript and its publication were very fast. Already in June, the paper was out, in World Psychiatry.2 With this record speed, I wondered what the peer review had been like, if there was any. It looked more like a letter to the editor, with no subheadings, and taking up only 1.5 pages. The number of deaths was now neither 49, nor 28, but 31.
As the information I had requested was not to be found in their article, Bob Whitaker and I wrote a letter to the editor of the journal, Professor Mario Maj, on August 16th with this message: “Melle et al. report in a 10-year prospective study of 281 patients with schizophrenia spectrum disorders that 11% died, although their average age at entry into the study was only 29 years. Their study might give us a unique insight into why so many patients with such disorders die so young, but there was too little detail in their paper to provide this much needed knowledge. We very much hope that you will help getting the knowledge Melle et al. have in their files published by publishing our short letter and asking them to respond to the issues we raise. That would be a great service to psychiatry, the patients, and everyone else with an interest in this vitally important issue.”
Our submitted letter was this one:
Why did so many young people with schizophrenia spectrum disorders die so early?
Melle et al. report that, in a 10-year prospective study of 281 patients with schizophrenia spectrum disorders, 11% died,1 although their average age at entry into the study was only 29 years.2 Their study might give us a unique insight into why so many patients with such disorders die so young, but there was too little detail in their paper to provide this much needed knowledge.
Melle et al. write that 16 died by suicide, 7 by accidental overdoses or other accidents, and 8 from physical illnesses, including 3 from cardiovascular illness. In order to attempt to separate iatrogenic causes of death from deaths caused by the disorder, we need to know:
When did the suicides occur? Suicides often occur early, after the patients have left hospital,3 and are sometimes iatrogenic. A Danish register study of 2,429 suicides showed that, compared to people who had not received any psychiatric treatment in the preceding year, the adjusted rate ratio for suicide was 44 for people who had been admitted to a psychiatric hospital.4 Such patients would of course be expected to be at greatest risk of suicide because they were more ill than the others (confounding by indication), but the findings were robust and most of the potential biases in the study were actually conservative, i.e. favoured the null hypothesis of there being no relationship. An accompanying editorial noted that there is little doubt that suicide is related to both stigma and trauma and that it is entirely plausible that the stigma and trauma inherent in psychiatric treatment — particularly if involuntary — might cause suicide.5
What does accidental overdoses and other accidents mean? Did the doctors overdose or did the patients overdose themselves by mistake, and which types of accidents were involved? Psychotropic drugs can lead to falls, which can be fatal, and suicides are sometimes miscoded as accidents.6
It is surprising that 8 young people died from physical illness. What were these illnesses exactly and what were the cardiovascular illnesses? If some of these people suddenly dropped dead, it could be because antipsychotics can cause QT prolongation.
Peter C. Gøtzsche1 and Robert Whitaker2
1 Nordic Cochrane Centre, Copenhagen, Denmark
2 Boston, United States
Conflicts of interest: none.
1. Melle I, Olav Johannesen J, Haahr UH et al. Causes and predictors of premature death in first-episode schizophrenia spectrum disorders. World Psychiatry 2017;16:217-8.
2. Hegelstad WT, Larsen TK, Auestad B et al. Long-term follow-up of the TIPS early detection in psychosis study: effects on 10-year outcome. Am J Psychiatry 2012;169:374-80.
3. Chung DT, Ryan CJ, Hadzi-Pavlovic D et al. Suicide rates after discharge from psychiatric facilities: a systematic review and meta-analysis. JAMA Psychiatry 2017;74:694-702.
4. Hjorthøj CR, Madsen T, Agerbo E et al. Risk of suicide according to level of psychiatric treatment: a nationwide nested case-control study. Soc Psychiatry Psychiatr Epidemiol 2014;49:1357–65.
5. Large MM, Ryan CJ. Disturbing findings about the risk of suicide and psychiatric hospitals. Soc Psychiatry Psychiatr Epidemiol 2014;49:1353–5.
6. Brown S. Excess mortality of schizophrenia. A meta-analysis. Br J Psychiatry 1997;171:502-8.
Eight days later we were told by Maj that, “Unfortunately, although it is an interesting piece, it does not compete successfully for one of the slots we have available in the journal for letters.”
Five days later, I replied and told Maj that we very much hoped he would reconsider his rejection of our letter:
“Allow me to add that people I have talked to in several countries about deaths in young people with schizophrenia — psychiatrists, forensic experts and patients — have all agreed that we desperately need the kind of information we asked you to ensure we get from the very valuable cohort of patients Melle et al. reported on in your journal.
There is widespread and well-substantiated suspicion that the reason we have not seen a detailed account of causes of death in cohorts like the one in the TIPS study by Melle et al. published in your journal is that the psychiatrists prioritise protecting their guild interests rather than protecting the patients. By declining to publish our letter and get the data out that Melle et al. have in their files, you contribute to that suspicion. We previously asked one of the investigators, Wenche ten Velden Hegelstad, to provide us with these data but were told on 10 May this year that they would be published … They have not been published, as what Melle et al. have published in your journal is not an adequate account of why these young people died.
Therefore, we call on you to ensure these data get out in the open, for the benefit of the patients. We believe it is your professional and ethical duty — both as a journal editor and as a doctor — to make this happen. This is not a matter about the slots you have available in the journal for letters. It is a matter of prioritization.”
We did not hear from Maj again. But since TIPS was supported by grants from the Research Council of Norway, I will ask the Council for the raw anonymised data on the deaths, as this is clearly in the public interest. TIPS was also supported by Lundbeck Pharma, Eli Lilly, and Janssen-Cilag Pharmaceuticals, which will hardly be interested in getting these data out in the open, so it will be interesting to see what happens.
In contrast to the authors of the TIPS study, professor of psychiatry Merete Nordentoft, Copenhagen, was forthcoming when I asked her about the causes of death for 33 patients after 10 years of follow-up in the OPUS study, also of patients with a first-episode psychosis.3 I specifically mentioned that suicides, accidents and sudden unexplained death could be drug related. I received a list of the deaths and Nordentoft explained that no one had a heart related cause of death registered but that this was probably because they were so young. In the certificates she had seen some patients had simply dropped dead, one of them while sitting in a chair.
This is how it should be. Openness is the way forward if we wish to reduce the many deaths that occur in young mental health patients. It shouldn’t be necessary to say this, but unfortunately there are few psychiatrists like Nordentoft. I have asked Hegelstad about the conflicting numbers of deaths and also asked to get the causes of death in more detail. I did not hear from Hegelstad again.
Conflicts of interest: none.
