| Methods |
Country: USA
Setting: well baby check
RCT |
| Participants |
103 mothers smoking ≥ 10 cigarettes/d with infants presenting to a well baby check |
| Interventions |
Urine was collected from all infants and analysed for cotinine.
Intervention: A report of the infant's urinary cotinine level along with a personalised letter to the mother to be signed was returned to the child's doctor. The letter outlined ways to reduce child ETS exposure (identify location of smoking, wash hands after smoking, ensure day care home is smoke‐free, ask friends to avoid smoking in the presence of the infant when visiting) but did not discuss cessation. The physician called the mother by telephone to further explain the results.
Control: usual care |
| Outcomes |
At 2 months, all participants were contacted to obtain a second urine sample from the infant for analysis. |
| Type of intervention |
Well‐child (child health check) |
| Notes |
Retention: 56/103 (54%) |
| Risk of bias |
| Bias |
Authors' judgement |
Support for judgement |
| Random sequence generation (selection bias) |
Low risk |
"randomly assigned by computer on an individual basis to intervention or control groups" |
| Allocation concealment (selection bias) |
Low risk |
See above. |
| Incomplete outcome data (attrition bias)
All outcomes |
Low risk |
High loss to follow up ‐ 43% control and 48% intervention; "however, it is unlikely that exclusion bias would mask a true impact of the intervention. Characteristics of those who complied were similar to those of the noncompliers... even with the reduced participation... the data were adequate to indicate that the response to the intervention was poor" |
| Blinding of outcome assessment (detection bias)
All outcomes |
Low risk |
Outcomes biochemically verified |
