Methods | Cluster-randomised trial to evaluate dissemination of a behavioUrally-based program to support women to stop smoking in pregnancy Study conducted in Queensland (Australia). Data collection dates not stated |
|
Participants |
Inclusion criteria: Public hospitals which provided antenatal and delivery care for 10 or more patients a year, had less than 50% Aboriginal and Torres Strait Islander population, and did not currently provide any antenatal smoking cessation care Exclusion criteria: Not further specified. Recruitment: Hospitals were matched on number of births, location of population centre (rural/metropolitan), and whether they had a specific antenatal clinic 80 (92% public hospitals) hospitals eligible. 10 omitted as they stopped providing antenatal care. 70 hospitals (35 pairs) included Baseline characteristics: Characteristics of individuals not reported. No outcomes included in study so not coded. |
|
Interventions |
Control: Received ‘awareness’ phase of intervention based in Rogers’ Diffusion of Innovation theory. Flyers were distributed to all hospitals Intervention: Control +‘Persuasion’ phase, which included an educational workshop and presentation. ‘Implementation phase’ where each hospital conducted the recommended program Main intervention strategy: Intensive dissemination vs less intensive intervention. No outcomes to include in analysis Intensity: NA |
|
Outcomes | Self-reported implementation of program at each hospital. Success was defined as the routine offer of an evidence-based smoking cessation program to at least 80% of the pregnant clients who smoke | |
Notes | ||
Risk of bias | ||
Bias | Authors’ judgement | Support for judgement |
Random sequence generation (selection bias) | Unclear risk | Report states hospitals were randomised into intervention and control groups, within matched pairs |
Allocation concealment (selection bias) | Unclear risk | Not reported. |
Incomplete outcome data (attrition bias) All outcomes |
High risk | Complete follow-up could not be obtained primarily due to the inability to contact either the medical superintendent or the director of nursing after a minimum of 3 attempts High attrition (37% hospitals), though those not responding were included in analysis as ‘not implemented’ |
Selective reporting (reporting bias) | Unclear risk | Smoking cessation rates not reported, but not included as an aim of this dissemination study |
Other bias | Low risk | No other bias detected. |
Biochemical validation of smoking abstinence (detection bias) | Unclear risk | Smoking status not assessed in this dissemination study. |
Blinding of participants and personnel (performance bias) All outcomes |
Unclear risk | Unclear whether control hospitals were blinded. |
Blinding of outcome assessment (detection bias) All outcomes |
Unclear risk | Not stated. |
Incomplete implementation | High risk | 37% reported as ‘not implemented’. |
Equal baseline characteristics in study arms | Low risk | Matching of the hospitals was successful as there were no differences in number of births, rurality, and whether they had a specialised antenatal service at baseline |
Contamination of control group | Low risk | Cluster design likely to minimise risk of contamination. |