Violence and Aggression

In most services that send people out alone, aggression accounts for more reported incidents than every physical hazard combined. It is also the category most workers do not report.

When violence and aggression also depends on reliable work records, available on this page can support time, attendance and workload review without being treated as the emergency response itself. Compare any workflow with Suzy Lamplugh Trust resources and keep alarm ownership, escalation and dispatch responsibilities explicit.

What it actually looks like

Verbal abuse and threats, which is the overwhelming majority and which people stop reporting because it is constant.

Intimidation: blocking a door, standing too close, following somebody to their vehicle.

Property damage during a visit.

Physical assault, which is rarer and is what everybody plans for.

And harassment afterwards — messages, social media, turning up — which is specific to services where the worker's identity is known and is almost never covered by a lone worker policy.

Where it comes from

Refusal. Enforcement, benefits, debt, tenancy, removals, age-restricted sales. The role involves saying no, and that is when it happens.

Distress. Bereavement, mental health crisis, a person whose care has been reduced.

Intoxication.

Third parties rather than the client: a relative, a partner, a neighbour. This is the largest blind spot, because the risk assessment is about the person being visited.

Waiting and unmet expectations, which is the one the organisation creates: a late visit, a cancelled service, a decision communicated badly.

The controls that work, in order

Not attending. The strongest and the one that requires permission to be real.

Attending in pairs, for flagged situations.

Attending at a different time or in a different place — a neutral venue, an office appointment, a daytime slot.

Information before arrival: what happened last time, who else is present, what the person is expecting.

De-escalation training, which has evidence behind it and which changes the outcome of the situations that do occur.

Then, and only then, the alarm.

The device sits sixth on a list of six, which is the honest ordering and the reverse of most programmes.

De-escalation, briefly

Not a technique but a set of habits: position, tone, distance, not arguing about facts, giving the person something to decide.

Taught through scenarios rather than slides, because the skill is behavioural.

And it includes recognising the point at which de-escalation has failed and leaving is the correct action, which is the part most courses underweight.

The organisation's own contribution

Unrealistic appointment windows produce angry people.

A decision communicated by letter and explained by the person at the door puts the worker in front of a reaction they did not cause.

Repeated changes of worker mean no relationship exists when it is needed.

And a complaints process that goes nowhere means the visit is where frustration lands.

Each of these is fixable by somebody other than the worker, and each is invisible in an incident report that records only what happened at the door.

Reporting, and why it does not happen

"It's part of the job" is the most common reason, and it is a cultural position that the organisation either challenges or endorses by silence.

Fear of the client being withdrawn, particularly in care, where the worker values the relationship.

Fear of being seen as unable to cope.

And a reporting form designed for physical injury, which does not fit verbal abuse and therefore does not get used.

Make verbal abuse and threats explicitly reportable, in one field, in thirty seconds, and the rate will rise sharply — which is the system starting to work rather than a deterioration.

Afterwards

Support offered actively, not signposted.

A choice about returning to that client or address, which is theirs.

Police reporting supported by the organisation rather than left to the individual, including the decision not to.

And a stated position that the organisation will not require somebody to attend where they have been assaulted, which sounds obvious and is frequently absent.

Flags, and their limits

A flag reduces risk only if it reaches whoever attends and says what to do.

"History of aggression" is a label; "two workers, daytime only" is an instruction.

And flags must be reviewed, because a permanent flag on somebody who had one bad incident three years ago is both unfair and eventually ignored.

The measure

Reported incidents by type, with verbal abuse counted separately, since a rise there is the measure of reporting rather than of risk.

Incidents by address and by round, which is where the clusters are.

Withdrawals, which should be non-zero.

And the proportion of incidents at addresses that were already flagged, which tells you whether the flagging is working or whether it is a record of things that already happened.

Recording incidents against the address

Rather than against the individual, where a third party was involved.

Because a relative who was abusive is not the client, and flagging the client is both inaccurate and unfair.

The address record carries the instruction; the client record carries only what relates to them.

Which is a small data design decision with a large effect on whether the flags are trusted and whether they survive a challenge from the person they describe.

The pattern worth looking for

Incidents concentrated at a point in a process: the week a decision letter goes out, the first visit after a service reduction, the day after a benefit change.

Which is the organisation's own calendar producing the risk, and it is predictable enough to schedule around.

Almost nobody looks for this correlation, and it is available from two datasets most services already hold.