Stephen O’Neill became anxious in 2016 and was given Sertraline. This caused an immediate toxic reaction that killed him, The doctors treating him invented a fictitious mental illness and both they and the coroner found it convenient to blame this fiction for Stephen’s death. See The Death of Stephen O’Neill and related posts.
The coroner and doctors all but laughed at the idea that the Doxycyline, a serotonin reuptake inhibiting drug, Dr Brannigan had given Stephen a few weeks beforehand might have produced the anxiety that led Dr Brannigan to prescribe Sertraline – not a treatment approved by MHRA or NICE Guidelines for Doxycycline toxicity. To recognise the link would have entailed recognising that they missed something sitting right in front of them.
A month after Stephen’s inquest, an horrific death of a young woman grabbed the attention of many people, not just in the UK. The drug she was on – doxycycline – was immediately in the frame. I was asked to provide a report and another coroner had to come to a conclusion. Could he find a Sanity Clause?
HM Senior Coroner
City of Milton Keynes
March 16th 2020
Dear Mr Osborne
You have asked me to prepare a report on the possible effects of doxycycline on AC taken by her for malaria prophylaxis when undertaking a field trip in 2019 as part of her university studies.
I am …
Background to AC’s Case
By way of background materials, you have provided me with
AC’s Medical History & Death
It appears from her medical records and the statements from her university supervisors that AC was a well-balanced and healthy young woman. She appeared well suited to the project she was undertaking and, barring an accident, there was little reason to think anything would go wrong and no reason to suspect the eventual outcome.
Suicides in women of this age are rare. Suicides in women of this age with no history of mental illness are vanishingly rare. There appear to be no financial, legal, relationship or other factors in this case pointing to any increase in risk and a lot of factors protective against suicide.
The statements of – several people – point to a change in AC’s mental state in the days immediately prior to the event. She appears to have had an increase in anxiety amounting to paranoia at several points, lability of mood, and what XY described as features of catatonia.
A’s mental state on both the day before and the day of the event appears to have been confused and distracted. The changes were such that they seem likely to have contributed to the event and it is difficult to see any reason for the event other than these changes.
When it comes to determining what has happened in an event like this, the medical process is very similar to the judicial process. The details of an event and those leading up to it are examined for possible explanations and the person in question, if present, or witnesses, are examined and cross-examined to establish the most likely explanation. This counts for more than any supposed science.
In this case, based on the risk profile, and the descriptions offered by those closest to AC, the most likely explanation is that AC entered a toxic state and became delirious. This seems to be the explanation that those looking at the events at close quarters appear to have reached.
Toxic states arise primarily from physical causes such as infections, physical illnesses, drug treatments or other medical interventions. Extreme psychological shocks can also give rise to a “functional” delirium.
In this case, there is no prior evidence for an infection or other physical illness, and it is not possible from the autopsy to establish whether there was an unrecognised physical disorder.
There are no reports of a psychological shock or at present reason to suspect one.
There were two changes of drug treatment in the weeks prior to the event, one involving X and the other her use of doxycycline for malaria prophylaxis.
Seems extraordinarily unlikely for these reasons
AC was also prescribed doxycycline, a tetracycline antibiotic commonly used for malaria prophylaxis, for the treatment of acne and for other infections.
She appears to have been aware that doxycycline can sometimes cause difficulties, and therefore took some test doses for approximately a week prior to leaving. She appears to have had some increase in anxiety and sleep difficulties but after reporting these to her parents she opted to stick with this treatment for her trip.
It appears that she took 11 doxycycline tablets, which is consistent with taking her medicines as she had been initially supposed to take them. Toward the end, she had apparently been advised to stop taking doxycycline, but it appears she did not stop, perhaps because she didn’t register the advice and was not registering much.
She may have been offered Nytol, diphenhydramine, to take as a sleep aid in the last few days but does not appear to have taken these tablets.
During the 11 days taking doxycycline, AC moved from being calm and confident to nervous and lacking in self-confidence. These changes are consistent with a drug induced behavioural toxicity. Her changes in mental state were marked in both nature and severity and sufficient to contribute substantially, even entirely, to the events that led to her death.
My impression is that a majority of observers looking at this scenario, with X labelled Factor A, and doxycycline labelled Factor B and other possibilities labelled Factor C would come to the conclusion that Factor B had likely played a part in AC’s death.
There is an extra element in this case that points to doxycycline. It appears that AC had tried it for a few days before leaving the UK and it caused some changes consistent with the changes she later displayed. This appears to offer a case of Challenge, De-challenge and Re-challenge (CDC).
When assessing a drug effect, CDC offers the strongest cause and effect evidence. This is a Christmas tree lightbulb approach to causality. Almost everyone will have had the experience of attempting to start old-style (pre-LED) Christmas tree lights after they had been put away for a year. When the lights failed to work, the standard procedure was to unscrew each bulb in turn and if the lights came on when a particular bulb was unscrewed and went off again when that bulb was screwed in again then this was taken as conclusive evidence that this bulb was the cause of the problem. The answer was to unscrew that bulb completely and leave the light set one bulb short or else substitute a new bulb.
CDC operates in the same way and AC’s case appears to offer an instance of this. Her prior test of doxycycline may in fact have inadvertently set her up to have a more rapid onset of an adverse response to doxycycline that she appears to have been predisposed to than she might otherwise have had.
There is one further element that may have played a part – Nytol. This drug is an antihistamine with serotonin reuptake inhibiting properties. While it is often used for sedation, it can cause the kind of agitation that selective serotonin reuptake inhibiting (SSRI) antidepressants can cause. Doxycycline also has serotonin reuptake inhibiting properties and people who become agitated on an SSRI may also become agitated on doxycycline and vice versa.
Doxycycline and Behavioural Events: Data
Making a diagnosis involves deciding which pattern (Gestalt) is most appropriate and the doxycycline pattern fits best and is likely to be the explanation most people would reach for other than those who find it difficult to believe that an “antibiotic” could cause a problem like this. There are two key elements in the pattern to note – one is the background difficulties consistent with the eventual outcome that is more evident with doxycycline and the other is, all things being equal, that a drastic outcome like this is more likely to occur in a shorter time-frame on a drug than a longer one.
When adverse events happen on drugs, doctors and increasingly other healthcare professionals and more recently those affected themselves can report to regulators such as the Food and Drug Administration (FDA) in the USA or the Medicines and Healthcare products Regulatory Agency (MHRA) in the UK. The FDA has the largest body of easily accessible reports
As of late 2019, the data for suicidal events and suicide-related events on doxycycline in OpenFDA data is shown below. Close to 1/3rd of reports on doxycycline are for behavioural events. These reports primarily come from doctors and are sent through companies to FDA. Roughly 1/6th of MHRA reports on this drug are for similar behavioural events – again primarily from doctors. These reports concern patients taking this drug for acne, malaria prophylaxis and other infections.
FDA reports – Doxycycline
Terms included in each group
This profile of effects on doxycycline likely arises in part because many of those reporting have been prompted to report by their surprise at a suicidal or agitated reaction on an “antibiotic”. The proportion of such reports on doxycycline is higher than the proportion for many drugs carrying warnings about suicide risk.
Another point to note though is that the designation of a drug as antibiotic or antidepressant is somewhat arbitrary. Prozac (fluoxetine), for instance, among the best-selling antidepressants, has antibiotic properties and doxycycline inhibits serotonin reuptake like Prozac does and can cause emotional numbing as Prozac and other SSRIs can and dependence and withdrawal reactions as this group of drugs can.
There is another dataset stemming from the use of isotretinoin (Accutane) for acne:
Sundstroem A, Alfredsson L, Sjoelin-Forsberg L, Gerden B, Bergman U, Jokinen J. Association of suicide attempts with acne and treatment with isotretinoin: retrospective Swedish cohort study. BMJ 2010; 342; c5812.
This paper carried the image below. All details of the data that went into this graph can be found in the published paper. The authors argue these figures indicate that acne leads people to commit suicide and therefore the suicide risk from isotretinoin can be discounted. This interpretation hinges on a view they offer that there is no signal that tetracycline antibiotics cause suicide – most people with acne will be given doxycycline or other tetracycline before being given isotretinoin. In the light of the FDA data above, this figure appears to offer strong evidence that doxycycline causes suicide/suicide attempts, as strong as the data for isotretinoin, bearing in mind that a proportion of these patients will not have been given doxycycline.
Doxycycline and Behavioural Events: Accounts
My interest in the issue of doxycycline and suicide was stimulated around 2012 by requests to write two reports for inquests after suicides on doxycycline and by witnessing a colleague given doxycycline, unbeknownst to me, become agitated on treatment.
This led to a publication in BMJ Case Reports headed Doxycycline and Suicidality. (Atigari OV, Hogan C, Healy D. BMJ Case Rep Published online: doi:10.1136/bcr-2013)
The cases are outlined here.
This is the case of a 19-year-old man, Mr. A, the youngest of four siblings from a happy home, with a stable social life. He was easy-going and intelligent. Mr. A had no history of mental disorder, and no history of alcohol or substance misuse. He had an early childhood history of mild asthma, allergic rhinitis, giardiasis, hives and allergic reactions with resultant swelling of his fingers and toes. The only family medical or mental history of note is that Mr. A’s eldest brother had reported severe anxiety symptoms after being commenced on doxycycline for malaria prophylaxis and another brother had developed similar symptoms on doxycycline for dermatitis which resolved after discontinuation of the medication.
Mr. A developed perioral dermatitis and was commenced on doxycycline 50mg BD six days prior to his death. He was not on any other medication. Four days after this, he went out socializing with friends and nothing unusual was noticed in his behaviour. On the night before his death, he engaged in sports with friends.
Later in the evening, he had a discussion with his parents about his future and they indicated they thought he was spending too much time on facebook and socializing. He accepted this and indicated he would limit his facebook usage in the next academic year to a set amount of time per day. His parents also had similar conversations with his brothers in relation their academic courses. Mr. A may have been a little bit more quiet than usual but nothing untoward was noticed.
On the day of his death, he had a normal morning at home with his family and exchanged innocuous text messages with his friends. He sent a text 17 minutes before his death to his friend to bring some whites for him to the cricket match. He then sent a text to the girl he liked, which read “wake up it’s nearly midday”. Within five minutes of this text, he jumped off a building at his former school, which is located close to his family home. Neither family nor friends had noticed any warning signs and no letter of intent to the possibility of a suicide has been discovered.
Following the death of Mr. A, the family came to an understanding that the medication prescribed in the cases of his older brothers, aforementioned, was possibly doxycycline.
In the autopsy report there was no alcohol or illicit substances detected in Mr. A’s blood or vitreous humour. Genetic testing of Mr. A revealed CYP2C19*2 heterozygote genotype associated with diminished cytochrome p450 enzyme activity. However, there was no method available in the laboratory for the analysis of doxycycline in the body.
This is the case of a 33yr old health professional in a stable relationship with her family including her sisters and mother. She had no history of mental disorder, alcohol or substance misuse. She has a history of mild acne for the last nine years. She has no other medical history of note.
Following a recurrence of her acne she was commenced on doxycycline 200mg stat, and then 100mg OD. The tablet gave effective relief for the acne. However, she developed nausea and headache on the second day of treatment. There was a feeling of disinhibition in her interaction with others, which she found liberating. This quickly developed into instability in her mood; fluctuating rapidly between irritability, anxiety and low mood. She became withdrawn, detached and began to ruminate over minor occurrences. By the third day she privately began to consider suicide as a solution to end her perceived problems.
Ms. B continued taking doxycycline for the next few days and the suicidal ideation, including thoughts of crashing her car got worse. She stopped treatment; she later attributed her stopping to luck. Ten days after discontinuation of doxycycline her mental state had returned to normal without any treatment with psychotropic medication. She remains well three months after discontinuation of the medication.
Afterwards she analyzed her circumstance with the help of family, and friends to determine if any occupational, family or social stressors could have precipitated or maintained the change in her mental state. None were identified.
Mr. C was initially put on doxycycline for mild acne; at the age of 18 years. He had no personal history of a mental disorder and was on no other medication. His sister was being treated for social anxiety. He had completed mechanics pre-apprenticeship and was employed part-time.
Mr. C took doxycycline 50mg per day over a period of a year. He felt unwell on it, and appeared “down” at times, and walked out of a job he appeared to be enjoying. This led his mother to talk things through with him but without either making a connection between his mood and his medication. He stopped treatment in August 2011. His mood brightened. He resumed work and began saving for a holiday.
He restarted doxycycline in February 2012, as his acne had returned. At this point he was apparently dispensed 100mg medication in 50mg packaging. He again began to have intermittent periods of low mood in the absence of any identifiable psychosocial stressors, which he reported to his mother. He again left his job. Eight weeks after restarting doxycycline and two days after leaving work he committed suicide by hanging.
With colleagues, since 2012, I have been running an adverse event reporting website – RxISK.org. In 2018, we had 16 reports of suicidality or agitation on doxycycline consistent with the reports in Cases A-C. In 2019, we had 11 such reports. This year (2020) we have had the following 7 reports in January and February.
Finally, there is currently a petition on Change.org called Doxycycline causes suicide in teens put there by Tara McCarthy after her 13-year- old son, Damein, committed suicide when put on this drug for acne. It has been signed by over 6,600 people.
Suicidality on Doxycycline
Doxycycline is an effective broad-spectrum antibiotic that is well tolerated by many people.
In the case of those who do not tolerate it, there appear to be two mechanisms that can lead to suicide. The most common mechanism is the induction of a toxic state that produces a delirium. By delirium here is meant a clouding of consciousness so that the individual is not registering their environment properly and behaves inappropriately in it.
These toxic states, whether caused by drugs or illnesses, are commonly accompanied by a degree of anxiety or agitation as can be seen in viral infections for instance. In the case of antidepressant and antipsychotic drugs, there is a word for this drug induced agitation – akathisia. But it is recognized that these drugs also cause a state more like AC which is commonly termed “psychotic”. The term delirium would probably be more appropriate in most instances than psychotic.
It is generally accepted that people in delirious states are not responsible for their actions. In the event of assaulting and even killing someone, while in such a state, such individuals have an absolute defense against a charge of murder. In the same way, death while in such a state would not ordinarily be viewed as a suicide any more than stepping out from a 55th floor window while under the influence of LSD would be regarded as a suicide.
The second mechanism involves an induction of emotional numbing that can lead to disinhibited or impulsive behaviour in someone who otherwise might appear to be in a state of clear consciousness. This appears to have been the element linked to the death of Case A above and the death of Damein McCarthy in the Change.org petition.
People who develop emotional numbness can also be agitated on treatment – somewhat confusingly perhaps there may be both anxiety and anxiolysis present at the same time.
Both delirium and emotional numbing may have been involved in AC’s case but a delirium seems to have been the predominant element.
Report to Regulator
The regulators for drugs, such as MHRA in UK, do not ordinarily investigate a death like AC’s in the manner that an inquest can. Regulators record suicidal events and their association with certain drugs without checking out the medical history and other details of the person in question and without making a judgement as to whether the death was more likely or not linked to the drug.
Reporting the results of an inquest to the regulator where a determination that it is more likely than not that this death was linked to this drug, if that for instance is the result of this inquest, can make an important contribution to wider patient safety.
It may encourage the regulator to request companies to add sufficient details to the label of their drug, in this case the many forms of doxycycline currently on the market. This is not a move aimed at reducing the use of doxycycline but rather aimed at putting a person who may be on treatment and adversely affected in a better position to understand what is happening to them. Or to put the relatives of someone affected such as her parents, in a better position to advise those who were looking after A not just to ask her to stop taking the drug but to ensure that all tablets were put beyond her reach.
This proposal aims at building confidence in the use of doxycycline rather than the reverse.
Mr Osborne’s Conclusion
How, when and where and for investigations where section 5 (2) of the Coroners and Justice Act 2009 applies, in what circumstances the deceased came by his or her death.
On the 25th of July 2019 the deceased open the door of a light aircraft that was flying from
Anjajavay to Antananarivo in Madagascar and fell to her death. She was a student from
Cambridge University carrying out research on the island. She had taken doxycycline as an
antimalarial medication and it is believed that she suffered a psychotic/delirium event that led to her behaviour and death.
Conclusion of the Coroner as to the death
The coroner has intimated he will be filing a regulation 28 report on this case, noting the effect of the drug.