Mark Robert Smith · Prevention of Future Deaths report

On 5 April 2019, an investigation commenced into the death of Mark Robert Smith (referred to as Mark during the Inquest and below) who died aged 47 years. The inquest commenced on 2 March 2026 and concluded on 24 March 2026. The jury recorded in a narrative conclusion that the direct cause of Mark’s death was an epileptic seizure leading to cardiac arrest. They also found that albeit not a direct cause the following was a significant contributing factor namely; “toxicity related to pregabalin and sodium valproate level in someone with epilepsy together with a decline in physical and mental wellbeing.”

A summary of the jury’s key findings as recorded in the record of Inquest is detailed below:

Mark arrived at HMP Thameside on 8th January 2019, with a documented history of asthma, epilepsy, and depression. He was already prescribed medication before arrival, including pregabalin, dosulepin and sodium valproate. All medication was maintained by the prison GP as before, with the exception of co-dyramol, which was replaced by an as-and-when required dose of diazepam, and a change of route for pregabalin administration from tablet to liquid, on a dose of 50mg per day for the first 7 days of prison admission, increasing to 100mg per day thereafter.

There was a conflict between the pregabalin prescription shown on Systm 1 prison medical records (3100mls across 29 days), and the 50mg per dose recorded on the administration records. There was a failure to correct the prescription on Systm 1 despite it appearing a minimum of 6 occasions.

On 16th January, Mark was referred to the prison mental health team after a documented decline in his mental health. He was admitted to the prison inpatient unit (“IPU”) on that day. He also experienced seizures and hallucinations in his first week at HMP Thameside.

On 17th January at 07:00, Mark was observed on the floor of his cell, unresponsive and foaming at the mouth. An emergency code (code blue) was called but was stood down after he came round. He was later reviewed by the prison GP, who, based on Mark's vital signs, including a high temperature, tachycardia, sweating and generally delirious presentation, admitted him to hospital. He was transferred to Queen Elizabeth Hospital, Woolwich by 22:50. His medication upon arrival was kept to the same plan as prison, including the increase in pregabalin dose to 100mg after day 7.

Blood testing revealed no illicit drugs causing Mark's symptoms but found a kidney injury correlating with high toxins in the blood, and low electrolytes due to dehydration.

Across the next 11 days, Mark was subject to extensive testing to rule out neurological infection. While at Queen Elizabeth Hospital, Mark was administered 2.5ml of liquid pregabalin twice a day.

On 23rd January, a member of prison healthcare staff visited Mark in hospital, raised concerns about Mark's condition and requested a multidisciplinary team meeting with Queen Elizabeth Hospital ahead of Mark's discharge back to HMP Thameside. This did not take place.

All neurological conditions were excluded, and a lower respiratory tract infection was diagnosed. Mark was discharged back to HMP Thameside IPU on 28th January at 20:00 and seen by the prison GP at 22:54.

On 29th January, a multi-disciplinary team meeting between custodial and medical staff took place, where concerns that Mark had lost power in his limbs, had significant mobility issues, risk of falling and high levels of confusion were raised.

The decision was made to place Mark into a disabled cell, which allowed more room and reduced risk of serious injury. However, this meant that constant observations through a closed door were not possible, and so Mark was subject to intermittent observations every 15 minutes. Due to concerns about the adequacy of 15-minute observations, healthcare staff requested approval from prison staff for a constant watch. Prison staff carried out a risk assessment, where the request was refused on the grounds of risk of safety to others.

On 30th January, Mark was observed to be having seizures on and off for 3 to 4 hours, urinating on himself and being generally incoherent. He was reviewed again by the Prison GP, who administered diazepam (without effect), and decided to readmit Mark to Queen Elizabeth Hospital at 17:05 on that same day.

On 31st January, Mark had a lumbar puncture, which later came back negative of any infection. On the same day, multiple members of prison healthcare staff attended Queen Elizabeth Hospital again and met doctors who had insufficient briefing on Mark's history, including a lack of NHS patient number. They were only able to discuss Mark's case with those junior staff.

On the same day, Mark was seen by the neurology consultant, who observed him shaking but in full consciousness. During this admission at Queen Elizabeth Hospital, Mark was administered 2.5ml of liquid pregabalin twice daily. Based on the negative lumbar puncture result, no further neurological findings, and positive vital signs, Mark was discharged from Queen Elizabeth Hospital back to HMP Thameside on 4th February.

This discharge was agreed based on a discussion between a junior doctor on the ward and the neurology SHO. Mark was given his medication on the morning of 4th February, however, no medication was given upon his readmission to HMP Thameside after discharge. Mark appeared restless, but talkative, and deemed fit to return to his cell. The IPU staff at HMP Thameside were not adequately prepared for Mark's return due to insufficient briefing and a declined MDT meeting with Queen Elizabeth Hospital, despite numerous requests.

During the night of 4th and 5th February, Mark was observed sitting on the floor acting in an erratic manner. Prison healthcare staff requested a cell unlock due to Mark's erratic behaviour and water on the floor of the cell, but this was declined due to it being a non-emergency.

During the morning of 5th February, a second internal prison multidisciplinary team meeting took place, where it was decided to keep Mark on the 15-minute observations. Around 9:00, Mark was observed by the prison psychiatrist to be sleeping, and therefore was unable to carry out an assessment. At 09:50 he was woken by healthcare staff for food and medication. He was observed by healthcare and prison staff to be wet through his clothes and bedding, shivering, with a temperature of 34.0 to 34.9 degrees celsius, which is hypothermic. He was assisted back to a dry bed in dry clothes. Thereafter he was observed to be sleeping, but 11 further observations by healthcare staff were missed. A prison GP came to review Mark around 12:00 but didn't carry this out as Mark was asleep.

At 17:30, Mark was woken for food, drink, medication and a vitals check. He was agitated and refused the vitals check and the tablets. It is not clear if Mark ingested pregabalin. At 18:00, Mark was found again agitated and this time out of bed, on the floor lifting and dropping his head, making contact with the floor. With assistance of prison staff, he was helped back to bed. A similar out-of-bed incident reoccurred around 18:45 to 19:00.

There was an inadequate handover between day healthcare staff and night healthcare staff regarding Mark's condition through the day and insufficient instructions given for the evening. At 20:30, Mark was observed on the floor in his cell and was helped back to bed by prison and healthcare staff. This was the last undisputed time that Mark was seen alive.

There were several observations made by prison staff across the evening. Despite the nursing observation charts being fully completed, there were falsified entries made at the following times: 19:15, 19:45, 21:30, 21:45, 22:15, 23:15 and 23:30.
Despite requesting support for the shift on the night of 5th and 6th of February, the duty nurse took unauthorised leave from the prison site from 20:55 to 21:24. With only a healthcare assistant remaining, this left the IPU inadequately staffed.

At 23:39, the nurse observed Mark laying prone in his cell, unable to observe breathing. After completing the full observation round, the same nurse returned 5 minutes later to observe Mark once more and left. A minute later, the nurse returned with a prison custody officer (‘PCO’) and observed through the panel again, expressing the need to open the cell door. At 23:47, the PCO called Victor 2 (senior officer) to request permission to open the cell door. At 23:51, a second PCO arrived and the cell door was opened, with a code blue called at 23:55. The second PCO retrieved the defibrillator from the prison management office but found it to be inadequately equipped without batteries or pads. Both prison CPOs performed CPR before the ambulance service arrival at 00:03. Following further attempts of resuscitation by paramedics for 34 minutes, Mark was pronounced dead at 00:38.

Mark died as a result of an epileptic seizure which led to cardiac arrest. Secondary contributory factors were toxicity related to pregabalin and low sodium valproate level, together with an overall decline of clinical state and metabolic condition, described as a combination of a lack of food and drink, restorative sleep, anxiety and inability to take medication.

The nature and extent of medical care and clinical observation by prison healthcare staff between 5th and 6th February possibly made a material contribution to Mark's death. The inadequacy of handover and basic observation, and failings in sufficient record-keeping by the prison healthcare staff during the evening of 5 February after 20:30 meant that signs and symptoms of an epileptic seizure were not observed and therefore an opportunity to perform life-saving measures in either administering medication to stop the seizure and/or, ultimately, timely CPR, was missed.

Mark was administered doses of liquid pregabalin of more than 2.5ml by healthcare staff while at HMP Thameside and this probably made a material contribution to Mark's death. On the balance of probability, this is the most likely cause of the levels of pregabalin seen in the toxicology postmortem results.

(1) The wrong dose of medication could be prescribed and/or administered with life threatening consequences. Albeit I have seen evidence of significant improvements in prescribing practice at HMP Thameside since 2019 (e.g HMIP report February 2026); as recently as 2024 to 2025 HMIP and IMB reports noted “significant risks with management of medicines” and “prescribing errors.” Whilst recent internal audits in 2025 show significant improvements, medication incidents (datix) are recorded in late 2025 and the principal pharmacist notes a very busy site with multiple prescriptions screened daily.
Further, during the inquest it proved difficult to establish how Systm 1 (the medical note system) operated and whether there were risks inherent in the system itself. For example, it was suggested the system would convert mg into ml or pre-populate entries such as 100ml, in contradiction to the subsequent PFD evidence provided. (Practice Plus Group)
(2) I remain concerned a similar situation could arise as did in Mark’s case regarding a discharge back to prison from QEH without an MDT meeting despite repeated requests from prison healthcare and there could be a risk to life. I note PFD evidence provided regarding improved liaison between QEH and the prison GP but note no reference to liaison between QEH and the head of the healthcare in prison, who visited QEH twice in Mark’s case.
Further, as recently as 2025 an incident occurred when there was a lack of understanding at QEH around the limited provision in healthcare in prison. I am concerned as to how awareness of the limits of prison healthcare will be disseminated on a continuing basis to new and locum staff at QEH. (QEH and Practice Plus Group)
(3) Healthcare staff may be unable to enter a prisoner’s cell at night and monitor them in a situation which may not constitute a “medical emergency” but in which a patient nevertheless requires attention; a patient could decline and the situation become life threatening. I have received detail from Serco of “an escalation process” if Oscar 1 and Hotel 1 cannot agree about entry to a cell and I have been told this has been “recommunicated” to all staff. However, this is not recorded in a policy. I am concerned about awareness of the process for all prison and healthcare staff including on an ongoing basis. (Serco and Practice Plus Group).
(4) There are no larger disabled cells (which can accommodate hospital beds and wheelchairs) adapted to also facilitate a constant watch. Security concerns, in this might not be possible, and a similar situation might occur to that in Mark’s case. (HMPPS)
(5) I am concerned there were numerous missed and falsified observation entries in Mark’s case and how similar incidents can be prevented given the reliance on handwritten sheets being scanned onto the system at a later stage. I appreciate the efforts made with training and audits. (Practice Plus Group)