What happened
Patrick Clarke, 68, a mental health staff nurse from Caterham in Surrey, was on duty at Southleigh Community Hospital on Brighton Road in Croydon in May 2016 when a patient attacked him with a knife. He was stabbed several times. A post-mortem examination gave the cause of death as a stab wound to the chest.
The patient was Jimmy Jedson, 40, of Croydon. The Metropolitan Police officer who led the investigation, Det Insp John Massey, described Clarke as “a hardworking man who dedicated his life to helping others”.
The outcome
Jedson was charged with murder and appeared at Bromley Magistrates’ Court. He was later found guilty of manslaughter on the grounds of diminished responsibility and, at the Old Bailey, was given a hospital order with no fixed end date.
Southleigh Community Hospital said after the verdict: “Our deepest sympathies are with [Mr Clarke’s] family, friends and colleagues. He was a valued and respected member of staff who will be deeply missed.” The senior investigating officer said the family had behaved with “the upmost dignity throughout this very harrowing ordeal”.
What went wrong
Nothing in the public record describes the hospital’s systems. There is no employer investigation on the record, no coroner’s prevention of future deaths report, and no finding about staffing, alarms or observation on the ward that day.
What the sources do describe is the setting. Southleigh was a private community hospital, the kind of step-down facility that cares for patients well enough to leave a large psychiatric unit and not well enough to be at home. The record’s reading is that facilities of this type carry the same unpredictability as a bigger unit with fewer staff on duty, and that a single clinician can end up covering a ward alone.
The sources do not say how many staff were on duty when Jedson produced the knife, and this page does not guess.
What it teaches
A community ward is not a low-risk ward. The patient who kills a nurse in a step-down setting is usually known to the service, often under formal care, and the risk is that familiarity does the work a risk assessment should be doing.
That matters most for a single clinician. A panic alarm is not a control until someone has tested it on that ward, at that hour, and knows what turns up. Positional awareness in a corridor, an exit that is not the way you came in, and an agreed signal to a colleague are the practical version of the same thing.
Questions for the room
- On your ward, when was the panic alarm last tested, and who arrived when it was?
- If you were the only clinician on duty and a patient produced a knife, what is your route out, and does it lead anywhere safer?
- What in a patient’s notes would change how you run a medication round on your own?
- Who reviews a step-down placement when a patient’s presentation changes?
Evidence and limits
Confidence: confirmed. Two independent sources were read in full, BBC News and the Guardian, both reporting on 15 May 2016. They agree on the central facts. Neither carries an inquest conclusion, an employer disciplinary finding, or the wording of the hospital order, and this page states none of those. The record dates the sentencing to 2017; the sources read give no sentencing date, so the year rests on the record alone and the body prints no year. The point about single-clinician cover is the record’s reading of the case, not a finding by any court.
Our lone working training covers the assessment and the backup protocol a single clinician needs on shift.
Sources
- BBC News, May 2016 · https://www.bbc.com/news/uk-england-london-39817051
- The Guardian, May 2016 · https://www.theguardian.com/uk-news/2016/may/15/man-jimmy-jedson-charged-stabbing-death-nurse-croydon
Reviewed 20 October 2026.