The hazards that actually injure and kill small children in this country are a short list, and it is not the list that occupies most of the worry. That mismatch is the useful thing to know about child safety, and it is the organising idea for everything gathered here.
Falls are the commonest accident at home by a wide margin, and the serious ones are falls between two levels rather than trips on the flat — off a bed, a changing mat, a highchair, out of a pram, down the stairs. Stairs account for a large share on their own, and they go on doing so well past the age at which a child is confident on them.
What kills rather than merely injures, in the under-fives, is suffocation, strangulation and choking. The specific causes are the part worth repeating, because they are not obvious and they do not appear in most parents' mental inventory of dangers. Looped blind cords have killed at least thirty-three young children since 2001. Nappy sacks have killed at least fourteen babies over the same period — they are thin, silent, unperforated, and usually stored within reach of the cot. Neither is a thing anyone worries about, which is precisely why they continue to do harm.
Then burns and scalds, where hot drinks cause most of the injuries to under-fives but bath water causes the most severe ones; poisoning, overwhelmingly from medicines, household products and cosmetics; fire; architectural glass; and drowning, which takes very little water and very little time.
Room by room the pattern is consistent: the living room produces the largest number of accidents, the kitchen the most serious ones, and the bathroom the worst scalds.
Age matters more than any single precaution, because the hazards change faster than household habits do. A baby who could safely be left on a bed last week can roll off it this week. A toddler who could not reach the worktop can now climb to it. A three-year-old walks downstairs alone. Most incidents involve a precaution that was perfectly adequate a fortnight earlier, which is an argument for revisiting the arrangements rather than setting them up once.
The counterweight deserves saying plainly. Children cannot be made safe by removing everything from their reach, and they learn to judge risk largely by being allowed some of it. The sensible aim is to engineer out the small number of hazards that cause real harm — cut the cords, turn down the hot water, fit the stair gate, lock the medicines away, use the right car seat — and then let a child climb things. A household organised entirely around hazard is not a safe one so much as an anxious one, and anxiety is not a protective factor.
The two most consequential items in this area are set out in full elsewhere: child car seats, and the sleep arrangements under safer sleep for babies.