FOR REGISTERED CLINICIANS ONLY — an educational resource, not a substitute for clinical judgement, your named/designated safeguarding professional, or local child-protection procedures.

Safeguarding · child physical abuse

The injury that does not fit the child, the story, or the stage of development.

Non-accidental injury is a way of seeing before it is a diagnosis. This is a teaching module — there is no decision tool, because the decision is never a doctor's alone to make. It covers how UK national guidance says to recognise inflicted injury, the evidence behind the signs you are taught to fear, and the law that governs what you must do next — set out for all four UK nations.

NICE CG89 (2009, upd. 2025)
England · Wales · Scotland · NI
Teaching · Evidence · Law

“… reasonable cause to suspect that a child … is suffering, or is likely to suffer, significant harm … the authority shall make, or cause to be made, such enquiries as they consider necessary …”

Children Act 1989, s.47(1)(b) — the statutory threshold is suspicion, not proof. Your job is to reach it, and to refer.

A broken bone, a bruise, a head injury — the same findings that are unremarkable in one child are the most important thing you will see all shift in another. What changes is not the injury but its fit: with the child's age and development, with the history offered, and with the other things you find when you look. This chapter is about learning to see the mismatch, and about the reflex that must follow it — not to accuse, but to suspect, refer, and keep the child safe while others establish what happened.

NICE UK national guidance (CG89), verbatim Evidence systematic review or cohort study Statute primary legislation, verbatim Teaching standard child-protection teaching, not guidance
Who this is about. The children most often harmed, and most often missed, are the ones who cannot tell you what happened: infants and pre-mobile babies. The younger and the less mobile the child, the higher the concern any injury should raise — because the younger and less mobile the child, the fewer injuries they can plausibly inflict on themselves. Every threshold below drops as age drops. A bruise on a cruising toddler's shin and the same bruise on a six-week-old are not the same finding.
A case to hold in mind · illustrative, not a real patient

A 5-month-old is brought to the ED at 22:00 by one parent. There is a bruise over the left cheek and a smaller one behind the ear. The history is that the baby "rolled off the sofa" this afternoon; the injury was noticed at bath time. The baby is alert, feeding, and settles quickly. The parent is calm, apologetic, and keen to get home. Observations are normal. A colleague suggests documenting "accidental bruise, safety-netting given" and discharging.

By the end of this chapter you should be able to say why almost nothing in that paragraph is reassuring, which single fact matters most, and what the law then requires of you. The answers are at the bottom.

The whole idea

Why we do what we do, in one table

The recognition principles this chapter is built on, and where each comes from.

The principleWhyBasis
Any injury in a non-mobile baby is concerning until explainedA child who cannot roll, crawl or cruise cannot generate the forces to bruise or fracture themselves. Bruising is present in under 1% of non-mobile infants.CG89 1.1.2 Sugar 1999 · Maguire 2005
Judge the injury against the developmental stage, not the calendar age"Cruising" — not a birthday — is what makes accidental bruising plausible. "Those who don't cruise rarely bruise."Sugar 1999
The history matters as much as the injuryAn absent, vague, changing or developmentally impossible account, or a delay in presenting, is itself an alerting feature.CG89 (unsuitable explanation) Teaching
Certain patterns are specific: hand/ligature/implement marks, non-bony sites, clustersThey carry the imprint of the mechanism; accidental bruising is small, over bony prominences, on the front of the body.CG89 1.1.1–1.1.2 Pierce 2010/2021
Suspect abuse with a fracture unless there is a medical cause or suitable explanationNo single fracture proves abuse, but rib fractures in infants carry the highest probability, and occult fractures are common.CG89 1.1.9 Kemp 2008
Think abusive head trauma in the under-3 with intracranial injury and no major accidentRetinal haemorrhages, rib/long-bone fractures or multiple subdurals alongside the head injury raise the concern.CG89 1.1.10–1.1.11 PredAHT
One injury is a reason to look for othersInflicted injuries cluster and are often of different ages; the visible one may be the least of them.CG89 1.1.9 (occult) skeletal survey
"Suspect" triggers referral; "consider" triggers looking harderNICE separates the two deliberately. Neither requires you to be certain, and neither is a diagnosis.CG89 terms used
Your duty is to refer a suspicion, not to prove a caseThe statutory threshold — "reasonable cause to suspect … significant harm" — is low by design, and it is the social worker's job, not yours, to establish what happened.CA 1989 s.47
Chapter 1

Non-accidental injury is a way of seeing, not a diagnosis

Why the label you are reaching for is "concern", not "abuse".

"Non-accidental injury" — child physical abuse — is not something you diagnose at the front door. It is something you suspect, and a suspicion is enough to act on. The whole edifice of UK child protection is built to take the burden of certainty off the person who first sees the child: you flag a concern, and a multi-agency process — social care, the police, paediatrics, sometimes the courts — establishes what happened. If you wait until you are sure, you have waited too long, and often you will never be sure at all.

NICE captures this in two words it uses with precision throughout CG89. Consider maltreatment means it is one possible explanation for the finding, and you should look for other alerting features and gather more information. Suspect maltreatment means there is a serious level of concern but not proof — and it is the trigger to refer the child to children's social care. CG89 Neither word asks you to be right. Both ask you to act.

The single most useful question

For every injury in a child, ask: does this injury, in this child, with this history, make sense? If the mechanism offered could not produce this injury, or this child could not have done what is described, or the story keeps changing, or nobody sought help for hours — the injury does not fit, and a non-fitting injury is the finding. The injury itself may be trivial. The mismatch is not.

Two facts about scale keep the reflex sharp. First, physical abuse concentrates in the youngest: inflicted fractures are most common in infants under one and toddlers up to three Kemp 2008, and abusive head trauma is overwhelmingly a disease of babies. Second, it is missed — repeatedly, and often more than once, before the injury that finally kills or maims. The "sentinel injury" — a minor bruise or intra-oral injury in a young infant, seen and dismissed weeks before a catastrophic presentation — is a recurring theme of serious case reviews. Teaching The bruise you are tempted to wave through may be the warning that was there all along.

Chapter 2

The history is half the examination

Why "unsuitable explanation" is a clinical finding, not a value judgement.

NICE ties almost every physical alerting feature to the same qualifier: the injury raises concern where the explanation is absent or unsuitable. CG89 An explanation is unsuitable when it does not account for the injury, is not consistent with the child's development, keeps changing, is given differently by different carers, or when there was an unexplained delay in seeking help. None of these, on its own, means abuse. All of them mean the story and the injury do not line up, which is exactly what you are meant to notice.

Take the history without cross-examining. Your role in the ED is to record — not to interview the child forensically, and not to confront the carer. Ask open questions, write down the answers in the words used (with quotation marks), note who gave the history and when, and note the time between the injury and the presentation. A carefully documented history taken at 22:00 is evidence; a leading question or an accusation is not, and can compromise both the child's protection and any later legal process. Teaching

  • No history at all for a real injury in a young child — the account is "I just found it" — is itself an alerting feature.
  • A mechanism the child could not perform — "he rolled over" in a baby who cannot yet roll — is a developmental impossibility, and the most useful thing you can document.
  • A shifting story — the mechanism changes between the triage note, your history, and the nurse's — matters more than any single version.
  • Delay in presenting a painful or frightening injury, without a reason, is an alerting feature in its own right (explicit for visceral injury in CG89 1.1.13).
Chapter 3

Bruising: the commonest sign, and the one most often missed

Why a bruise on a baby who cannot yet move is the whole subject in one finding.

Bruising is the most common — and the most commonly overlooked — presentation of physical abuse, and it is where the developmental-stage rule does its clearest work. In a landmark community study of nearly a thousand normal children, bruises were found in only 0.6% of babies under six months and in just 2.2% of children not yet cruising, rising to 17.8% of cruisers and 51.9% of walkers. Sugar 1999 The line the authors drew has become the aphorism every ED clinician should carry: those who don't cruise rarely bruise. A bruise in a child who cannot yet get about under their own steam is abnormal until proven otherwise.

<1%of non-independently-mobile infants have any bruise (Maguire 2005 systematic review)
0.6%of babies under 6 months had any bruise in a community cohort (Sugar 1999)
96%sensitivity of the TEN-4-FACESp bruising rule for abuse in under-4s (Pierce 2021)

The pattern and site separate accidental from inflicted bruising. Accidental bruises are small, sit over bony prominences (shins, knees, forehead in a toddler), and are on the front of the body. Inflicted bruises are away from bony prominences, cluster, are commonly multiple, may carry the imprint of an implement, and favour the head and neck — especially the face — then the buttocks, trunk and arms. Maguire 2005 NICE turns this into instructions to suspect maltreatment: bruising shaped like a hand, ligature, stick, teeth, grip or implement; or bruising that is unexplained and falls into examples including bruising in a child who is not independently mobile, multiple bruises or clusters, bruises of similar shape and size, bruises on non-bony parts including the eyes, ears and buttocks, bruises on the neck like attempted strangulation, and bruises on the ankles and wrists like ligature marks. CG89 1.1.1–1.1.2

TEN-4-FACESp — a memory aid, not a rule to discharge on

The bruising clinical decision rule refined and validated by Pierce and colleagues flags bruising to the Torso, Ear or Neck in a child ≤4 years, or any bruising in an infant ≤4 months, extended to the Frenulum, Angle of jaw, Cheek, Eyelid and Subconjunctivae, plus any patterned bruising. Pierce 2021 It is a screen to prompt evaluation, derived and validated in US paediatric settings — useful for teaching the high-risk sites, not a substitute for the NICE alerting features or your local pathway. A negative rule does not clear a worrying child.

One more caution: do not attempt to age a bruise by its colour. The old teaching that colour reliably dates a bruise has not survived scrutiny, and a confident "this is about four days old" written in the notes can be discredited in court and undermine the whole assessment. Describe what you see — site, size, shape, colour — and leave the dating to those with the evidence base for it (there largely isn't one). Teaching

Chapter 4

Fractures: which ones, and why the skeletal survey exists

Why no fracture proves abuse, and no single film clears it.

NICE says to suspect maltreatment when a child has one or more fractures, in the absence of a bone-fragility condition (such as osteogenesis imperfecta or osteopenia of prematurity) or a suitable explanation — and it names two presentations in particular: fractures of different ages, and occult fractures found on X-ray that were not clinically evident, "for example, rib fractures in infants". CG89 1.1.9 Those two phrases contain the whole reason a skeletal survey exists.

The systematic review behind the concern is Kemp's. Across 32 studies it found inflicted fractures throughout the skeleton, commonest in infants and toddlers, with multiple fractures more common in abuse. Once major trauma was excluded, rib fractures carried the highest probability of abuse (0.71, 95% CI 0.42–0.91); humeral fractures 0.48–0.54, femoral 0.28–0.43 (with the child's developmental stage an important discriminator), and skull fractures 0.30. Its central conclusion is the one to remember: no fracture, on its own, can distinguish an abusive from a non-abusive cause. Kemp 2008 Probability is not proof, in either direction.

Some fracture types are, nonetheless, strongly associated with inflicted injury by their mechanism:

  • Posterior rib fractures in an infant. They result from anteroposterior compression of the chest — a squeeze — not from the anterior/lateral pattern of the rare CPR-related rib injury, and not from normal handling. In the Kemp data they carried the highest abuse probability.
  • Classic metaphyseal lesions (the "corner" or "bucket-handle" fracture) at the ends of long bones, produced by shearing/tractional forces on the infant metaphysis — grabbing and pulling or twisting a limb. They are subtle on plain film and highly specific for infant abuse. Teaching
  • Multiple fractures, fractures of different ages, and occult fractures — the constellation the skeletal survey is designed to reveal.

When physical abuse is suspected in a young child (UK practice: under two years), a full skeletal survey — a defined series of dedicated radiographs, not a single "babygram" — is the first-line imaging, with a follow-up survey around 11–14 days later to catch healing fractures invisible on the first, performed and reported to the RCR/RCPCH national standards. This is a paediatric and radiology decision, made with the safeguarding team, not something initiated and interpreted single-handed in the ED. RCR/RCPCH standard Teaching

Chapter 5

The head: abusive head trauma

The most lethal presentation, and the one whose signs are inside the skull.

Abusive head trauma (the older term is "shaken baby syndrome") is the leading cause of fatal physical abuse in infants, and the one where the crucial findings are not on the skin. Teaching NICE says to suspect maltreatment when a child has an intracranial injury with no major confirmed accidental trauma or medical cause, in any of these circumstances: the explanation is absent or unsuitable; the child is under three years; there are also retinal haemorrhages, or rib or long-bone fractures, or other associated inflicted injuries; or there are multiple subdural haemorrhages with or without subarachnoid haemorrhage or hypoxic–ischaemic damage. CG89 1.1.10 Separately, it says to suspect maltreatment for retinal haemorrhages or eye injury with no major accident or known medical cause including birth. CG89 1.1.11

Axial non-contrast head CT of an infant; a red arrow marks a focal intraparenchymal haemorrhage with an overlying skull defect in the right parieto-temporal region.
Abusive head trauma on CT. Axial non-contrast head CT of an infant. The arrow marks a focal intraparenchymal haemorrhage with an overlying skull fracture. In practice the more typical and more specific finding is subdural haemorrhage — often thin, multiple, and over the convexities or interhemispheric — frequently with retinal haemorrhages and encephalopathy. A young infant may present only with irritability, poor feeding, vomiting, apnoea or seizures, with no external mark at all.
"An intraparenchymal bleed with overlying skull fracture from abusive head trauma." James Heilman, MD, CC BY-SA 4.0, via Wikimedia Commons. Resized, otherwise unaltered.

The presentation is treacherous precisely because it can be occult. A shaken infant may arrive with nothing but a history of "not himself" — vomiting, floppiness, a brief apnoea, a seizure — and normal external examination. The threshold for imaging the head of a young infant with unexplained neurological symptoms, or with any other inflicted injury, has to be low. Teaching

PredAHT — the evidence turned into a probability, with a health warning

The Cardiff group's Predicting Abusive Head Trauma (PredAHT) tool estimates the probability of abusive head trauma from a combination of six features: head or neck bruising, seizures, apnoea, retinal haemorrhage, rib fracture and long-bone fracture. Cowley 2018 It is an adjunct to specialist judgement in a child already being investigated for AHT — not an ED rule-out, and its own evaluation found professionals wary of a low score giving false reassurance and unsure of its standing in court. It belongs to the paediatric/child-protection assessment, referenced here so you recognise the six features, not so you calculate a number at triage.

Retinal haemorrhages deserve one line of caution: they are found in some accidental trauma and in medical conditions, and a competent examination for them is an ophthalmologist's dilated fundoscopy, not a direct-ophthalmoscope glance in the resus bay. Their character — multiple, multilayered, extending to the periphery — is what makes them concerning, and that is a specialist call. Teaching

Chapter 6

Burns, the mouth, the abdomen, the spine — and the cold

The rest of CG89's physical features, each with its own tell.

Thermal injury

Suspect maltreatment with a burn or scald if the explanation is absent or unsuitable, if the child is not independently mobile, on any soft-tissue area that would not normally contact a hot object in an accident (backs of hands, soles, buttocks, back), in the shape of an implement (cigarette, iron), or with the hallmarks of forced immersion: scalds to buttocks, perineum and lower limbs; a glove-or-stocking distribution; symmetry; and sharply delineated borders. CG89 1.1.6 The accidental scald is asymmetrical, has splash marks and an irregular edge, and fits a reach-and-pull mechanism; the immersion injury is symmetrical, uniform-depth, and sharply bordered.

Bites, lacerations and scars

Suspect maltreatment with a human bite mark unlikely to have been made by a young child CG89 1.1.3, and with lacerations, abrasions or scars where the explanation is unsuitable — including on a non-mobile child, multiple, symmetrical, on areas usually protected by clothing, on the eyes/ears/sides of face, or on the neck/ankles/wrists like ligature marks. CG89 1.1.5

Oral injury

Consider maltreatment with an oral injury and an absent or unsuitable explanation. CG89 1.1.14 A torn frenulum in a young infant — sometimes attributed to forced feeding, though the mechanism is debated — is a small injury that has repeatedly turned out to be a sentinel one; look inside the mouth.

Visceral injury

Suspect maltreatment with an intra-abdominal or intrathoracic injury and no major confirmed accidental trauma, where the explanation is absent or unsuitable or there is a delay in presentation — and note NICE's warning that there may be no external bruising or other injury. CG89 1.1.13 Abusive abdominal trauma is the second commonest cause of fatal physical abuse after head injury, and a normal-looking abdominal wall does not exclude it. Teaching

Spine, cold and the general injury

Suspect physical abuse with signs of spinal injury and no major accident — which may appear as a skeletal-survey or MRI finding, as cervical injury with inflicted head injury, or as thoracolumbar injury with focal neurology or unexplained kyphosis. CG89 1.1.12 Consider maltreatment with cold injuries (swollen red hands or feet) or unexplained hypothermia CG89 1.1.7–1.1.8, and with any serious or unusual injury for which there is no suitable explanation. CG89 1.1.15

Chapter 7

One injury is a reason to look for others

Why the visible injury may be the least of them.

Inflicted injuries cluster, and they are often of different ages. The bruise that brought the child in may sit over rib fractures you cannot feel, or accompany a subdural you cannot see. This is the logic of the skeletal survey and of considering neuroimaging in the infant, and it is why NICE explicitly lists occult fractures among the presentations that should make you suspect maltreatment. CG89 1.1.9

Two extensions of the same principle belong in the ED clinician's head:

  • Examine the whole child, undressed. You cannot find what you do not look for. A full skin survey, the mouth, behind the ears, the frenulum, the buttocks and the soles are all named sites for a reason.
  • Think about the siblings. Where one child in a household is being harmed, others may be at risk. The referral you make is about the family's children, not only the one in front of you — and the statutory duty to make enquiries is framed around the child at risk, whoever presents. Teaching
Chapter 8

What to do once you have a concern

The steps between "this does not fit" and a safe child — the bridge to the Legal tab.

Recognition is only useful if it is followed by the right actions, in the right order, without confrontation and without delay. The sequence is the same everywhere in the UK; the statutory routes it feeds into differ by nation and are set out under Legal.

  1. Make the child medically safe first. Resuscitate and treat the injury on its clinical merits. Safeguarding never displaces acute care; it runs alongside it.
  2. Examine fully and document contemporaneously. Record the history verbatim with quotation marks, who gave it and when, the injury by site/size/shape/colour, and the time from injury to presentation. Body maps and (per local policy and consent) clinical photography help. Do not date bruises by colour; do not editorialise.
  3. Do not interrogate or accuse. Your job is to record and to refer, not to investigate or to secure a confession. Confrontation can place the child at greater risk and can compromise a later legal process.
  4. Be honest with the family, within reason. Explain, in general terms, that when a child has an injury like this you are required to seek advice from colleagues and social care. Openness is the default; the exception is where telling the family would put the child at greater risk (for example a risk of removal or of the injury being concealed), when advice is taken first.
  5. Discuss with your named or designated safeguarding professional and the on-call paediatric team early. You do not carry this alone, and you should not.
  6. Refer to children's social care when you suspect maltreatment — this is the operative step, and the threshold is suspicion, not proof (CG89; the statutory threshold is the s.47 "reasonable cause to suspect … significant harm"). A telephone referral is confirmed in writing per local procedure. If a child is in immediate danger, that is a police matter, now.
  7. Keep the child safe in the meantime. Admission is often the safest holding position while assessment proceeds; a child must not be discharged into possible danger to expedite the department.
The one line to carry

You are not required to be certain, and you are not required to prove anything. You are required to recognise a concern, act to keep the child safe, and pass the concern to the people whose job it is to investigate it. The law makes the threshold suspicion for exactly this reason — see Legal.

Watch out

Ten pitfalls

The recurring ways an inflicted injury gets waved through — or a family gets wrongly accused.

01Judging by the injury's severity, not its fit

The trap
A tiny bruise looks trivial, so it is dismissed. But a tiny bruise on a non-mobile baby is a major finding.
Instead
Judge every injury against the child's development and the history, not by how bad it looks.

02Forgetting to undress the child

The trap
The presenting injury is examined; the frenulum, the ears, the buttocks, the soles and the back are not.
Instead
A full skin survey of the undressed child, mouth included, in any injury that raises concern.

03Accepting the first history and stopping

The trap
A plausible story is written down and not tested against the child's development, or against what other carers say.
Instead
Ask who was present, what exactly happened, and record each account. Inconsistency is the finding.

04Trying to date a bruise by its colour

The trap
A confident age written in the notes that the evidence cannot support, and that is later discredited.
Instead
Describe colour and appearance; leave dating to those with an evidence base — largely, no one.

05Treating a normal scan or a negative rule as "cleared"

The trap
A negative TEN-4-FACESp, a normal CT or a normal single X-ray is read as reassurance.
Instead
These reduce but do not exclude concern; occult fractures and evolving injuries are the point of follow-up imaging.

06Waiting to be certain before referring

The trap
"I'm not sure it's abuse" becomes a reason not to refer. The threshold is suspicion, not certainty.
Instead
Refer the concern. Establishing what happened is social care's job, not yours.

07Confronting or interrogating the carer

The trap
An accusation in the resus bay that raises the child's risk and taints any later process.
Instead
Record, stay neutral, be honest in general terms, and route the concern through social care and the police.

08Discharging to clear the department

The trap
A settled baby and a calm parent make discharge feel safe and efficient at 3am.
Instead
Admission is often the safest holding position while assessment proceeds. Do not send a child into possible danger.

09Missing the sibling and the household

The trap
The concern is confined to the child in front of you; other children at home are not considered.
Instead
Safeguarding covers the household's children; say so in the referral.

10Anchoring on abuse and missing a medical cause — or the reverse

The trap
Either a bleeding or bone disorder is missed because "it's abuse", or real abuse is explained away as "probably a clotting problem".
Instead
Consider mimics and abuse together; a coagulation screen or a bone opinion does not remove a safeguarding concern, and vice versa.
Not everything is abuse

Mimics and red herrings

Conditions and marks that can look like inflicted injury, and inflicted injuries that can look benign. Naming them is part of a fair assessment — but none of them excludes abuse, and abuse and a medical condition can coexist.

Looks like bruising
Dermal melanocytosis

The blue-grey "Mongolian" patches over the buttocks/back, common in darker-skinned infants, are congenital, non-tender and stable — not bruises. Document them so they are not later mistaken for injury.

Looks like bruising
Coagulopathy & ITP

Bleeding disorders, ITP, vitamin K deficiency and leukaemia cause easy bruising and petechiae. Investigate — but a clotting result does not, by itself, discharge a safeguarding concern.

Looks like burns
Coining & cupping

Cultural healing practices leave patterned marks. They are relevant context — but the presence of a cultural explanation does not remove the duty to consider harm.

Looks like fractures
Osteogenesis imperfecta

Brittle-bone disease causes fractures with minimal force. NICE names bone-fragility conditions as the explicit exception to the "suspect with a fracture" rule — but it is a diagnosis, not an assumption.

Looks like periosteal injury
Caffey disease / rickets

Infantile cortical hyperostosis, rickets, copper deficiency and physiological periosteal reaction can mimic healing fractures on film. A paediatric radiology opinion is what separates them.

Looks accidental
Birth-related injury

Clavicle fractures, cephalhaematoma and some retinal haemorrhages occur at birth and resolve on a known timeline. Age and history place them — but only within the neonatal window.

Looks benign
The "trivial" sentinel injury

The opposite red herring: a torn frenulum, a small bruise or a single rib fracture in a young infant, dismissed as minor, that is in fact the warning before a catastrophic injury. Small does not mean safe.

Back to the start

The case, revisited

The 5-month-old with a cheek bruise and a bruise behind the ear, "rolled off the sofa":

  1. The child is not independently mobile. CG89 counts a child as independently mobile only once they can crawl, bottom-shuffle, pull to stand, cruise, climb or walk — rolling does not count. So even if this baby can roll off a sofa, they remain in the highest-concern group and cannot inflict a facial bruise on themselves. This single fact makes any bruise an alerting feature — <1% of non-mobile infants have any bruise at all. CG89 1.1.2 Sugar 1999
  2. The sites are high-risk. Cheek and behind/around the ear are exactly the TEN-4-FACESp regions; ear bruising in particular is rarely accidental. Pierce 2021
  3. The mechanism may be developmentally implausible for the injuries described, and "noticed at bath time" hours later hints at delay.
  4. There are two bruises, not one — and the visible bruises are a reason to look for occult injury (skeletal survey, consideration of neuroimaging), not to discharge. CG89 1.1.9–1.1.10
  5. "Accidental bruise, safety-netting given" is the wrong disposition. The correct one is: full undressed examination, contemporaneous documentation, discussion with the paediatric and safeguarding team, and referral to children's social care — the threshold is suspicion, which is comfortably met. CG89 CA 1989 s.47

Nothing here required you to decide that this parent harmed this child. It required you to notice that the injury did not fit, and to act — which is the whole of the job.

The evidence behind the signs is under Evidence; the law that governs what you must do next, for all four UK nations, is under Legal; scope, limits and sources are under About.