Table 2

Effectiveness of drowning interventions using swimming lessons and water safety (included studies)

Interventions
Study characteristics (quality)Intervention descriptionKey elementsMeasuresResults
Asher et al16 Water safety training and swimming lessons
Randomised trial no control

Evidence level: III-3

Swimming skills and water safety in two groups

24–42 months
USA
Swimming skills and water safety lessons delivered in two groups twice weekly for 8 weeks (n=48) or 12 weeks (n=61) duration.
Participants recruited by letters sent to Child Care Centres located near public pools in middle-income Seattle area 1990s.
Trained instructors. Lessons adapted from American Red Cross programme

Instruction provided in groups of six children in pool with parent. Lesson time not published.

Evaluation by blinded observers and follow-up phone calls over 18-month period.

Participants offered $50.
Three skills sets:
1. Out of water safety behaviour (deck behaviour—e.g. running around pool, pushing others, entering water without an adult).
2. Swimming ability (face underwater, recover from prone, roll back to front, propulsive kicking, beginner stroke, independently enter and exit pool, jump into pool independently).
3. In-water safety skills (water recovery ability to stand up when dropped from above water and ability to jump in and swim to edge of pool).

Swimming ability assessed by instructors. Water safety skills measured by independent blinded observers (instructors) four times in two groups.
Self-report surveys for parental demographics and child development.
Evidence that swimming lessons improve swimming ability in children aged 2–3 years.

Swimming ability significantly improved in both 8- and 12-week groups p<0.001*

Deck behaviour not improved p<0.03.†

Water recovery showed significant improvement in both groups p<0.001.*
Jump and swim improved over time in both groups p<0.005.*

No control group—effect could be explained by water familiarity.
Impact of skills effecting parent vigilance not studied and no children <2 years. Recommends swimming part of comprehensive approach including barriers, PFDs adult supervision and safety awareness.
Found no support for concern that water safety instruction increases young children's risk of drowning
Incentive payment. Simulated risk as proxy for drowning. No children <24 months included.
Brenner et al18 Swimming lessons (formal and informal) and drowning assessed
Case–control population-based

Evidence level: III-2
Formal lessons vs informal/no lessons

1–19 years

USA
Fatal drowning among 1–19 years, April 2003–September 2005 (2.5 years).
Cases (n=88) (1–4 years n=61; and 5–19 years n=27) fatalities identified from Coroner's data.
Controls (n=213) matched to cases on age, sex, residence and presence of pool at their home.
Telephone interviews by random digit dialling or mail with next of kin.
$25 inducement fee paid to participants.
Formal and informal swimming lessons assessed for association with fatal drowning.
Definitions:
Formal lessons: child received paid lessons or through day care, school or camp. Informal lessons: child not received pointers about swimming or water safety.
Non swimmers: no exposure to water or pointers or tips.
Outcome fatal drowning.
Exposure to water (≤1 time, or more), swimming ability (Y/N float on back, float on stomach, jump in pool and swim 5 ft back to wall), participation in formal and informal swimming lessons (Y/N).

Parental self-report child development, temperament (1–4 years), risk taking, medical conditions and household characteristics.
Swimming lessons: 1–4 years. Cases less likely than controls 3% vs 26% to have participated in previous formal swimming lessons (adjusted OR=0.12 95% CI 0.01 to 0.97).‡
No significant association between formal swimming lessons and drowning in older children 5–19 years (OR=0.36, 95% CI 0.01 to 1.51)†
No significant associations were observed between informal swimming instruction and drowning in either age group.
1–4 years: cases less skilled at floating on back for 10 s 5% vs 18% controls p=0.01‡
Cases were also more likely to be non-white p=0.4; low income p=0.03, low education p=0.003; risk takers p=0.03; temperament p=0.06; other medical condition p=0.07.

5–19 years: less likely to swim > 1 min cases 42% vs controls 16% p=0.01†

Cases more likely to be non-white, low income or have seizure disorder p<0.01.

Limited by small sample numbers. Incentive payment.
  • *Statistically significant improvement.

  • †No improvement not statistically significant/no change.

  • ‡Improvement but not statistically significant.

  • PFD, personal flotation devices.