Table 1

Risk of bias in individual studies.

Paper

Blinding of assessors?

Treatment and control groups similar at baseline?

Percentage drop out at post intervention measure?

Analyzed by intention to treat

Subgroup analyses reported?

Statement of study funding

Included in meta-analysis?


Bodenmann et al. [32]

No

Yes

Triple P 5%

CCET 8%

Control 23%

(at long term follow up)

No

Yes

Yes. Gebert Ruef Foundation (Switzerland)

Yes

Connell et al. [50]

No

More females in control group

Intervention 0%

Control 8%

No

Yes

No

Yes

Gallart & Matthey [26]

No

Yes (not tabulated)

Not stated (9% overall)

No

Yes

No

Yes

Hahlweg et al. [51]

No

Yes

Intervention mothers 14%

Control mothers 3%

All fathers 19% (unable to distinguish intervention & control attrition)

No

No

Yes. Deutsche Forschungsgemeinschaft

Yes

Hahlweg et al. [13] (data also reported in [52])

No (parents and teachers)

Yes (observers)

More parents in control group were single in comparison to the intervention group: 34% and 15.6%, respectively. Baseline differences between groups for two-parent households

Intervention 0.5%

Control 1%

Yes

Yes

Yes. Deutsche Forschungsgemeinschaft

Yes

Hoath & Sanders [53]

No (parents)

Not known (teachers)

Control group had lower family income

Intervention 10%

Control 0%

No

No

No

Yes

Joachim et al. [54]

No

Higher proportion of male children in control group

Intervention 15%

Control 10%

Yes

Yes

No

Yes

Leung et al. [55]

No

Yes

Intervention 28%

Control 20%

Yes - but only per protocol results tabulated

No

No

Yes

Markie-Dadds & Sanders [56]

No

Yes

Intervention 3%

Control 0%

No

Yes

No

Yes

Markie-Dadds & Sanders [57]

No

Yes

Intervention 28%

Control 23%

Yes - but only per protocol results tabulated

Yes

Queensland Health and the National Health and Medical Research Council

Yes

Martin & Sanders [58]

No

Treatment group had lower ECBI scores

Intervention 30%

Control 50%

No

Yes

No

Yes

Matsumoto et al. [59]

No

Yes

Intervention 0%

Control 0%

Yes (in effect)

No

No

Yes

Matsumoto et al. [60]

No

No. ECBI scores substantially higher in intervention group

Not stated

No

No

No

Yes

McTaggart & Sanders [21]

No

Yes

Not known

No

Yes

No

Not ECBI/CBCL

Morawska & Sanders [61]

No (parents)

Yes (observers)

Yes

Intervention 12%

Control 10%

Yes - but only per protocol results tabulated

Yes

No

Yes

Morawska & Sanders [62]

No

No. ECBI scores substantially higher in intervention group

Intervention 11%

Control 3%

Yes - but only per protocol results tabulated

Yes

Yes. Telstra

Foundation.

Yes

Morawska et al. [63]

No

Yes

Intervention 18%

Control 18%

Yes - but only per protocol results tabulated

Yes

No

Yes

Nicholson & Sanders [28]

No (parents and step parents),

Possibly (teenager's self-report)

Yes

40% therapist-delivered

45% self-delivered

5% waiting list control

No

yes

Yes. National Health and Medical Research Council

Not ECBI/CBCL

Plant & Sanders [64]

Yes (video observations)

No (parent report)

Yes

Nil in all three groups

Yes (in effect)

Yes

Yes. Australian Research Council and Apex Foundation

ECBI only used as entry screener

Prinz et al. [5]

Not clear

Not clear (five year average data presented)

Not known

Yes (in effect)

No

Yes. US CDC

Not ECBI/CBCL

Roberts et al. [33]

Yes (video observations)

No (parent report)

In some scales

37% intervention

35% control

No

Yes

Yes. Western Australian Health Promotion Foundation

Not ECBI/CBCL

Sanders et al. [65]

Yes (video observations)

No (parent report)

No data presented

EBFI 23%;

SBFI 17%

SDBFI 18%;

control 8%

No

Yes

Yes. Grants from Queensland Health and the National Health and Medical Research Council

Yes

Sanders et al. [66]

No

Yes

Not stated

Not clear

Yes

Partial - acknowledged source of TV programs and funding for distribution of video material

Yes

Sanders et al. [6]

No

No. Intervention area sample younger, poorer, less well educated and more likely to be single

Not applicable

Not applicable

Yes

Yes. Several funders

Not ECBI/CBCL

Sanders et al. [27]

No

No data presented except baseline measures

Intervention 23%

Control 12%

Yes

No

Yes. Australia Research Council

Yes

Stallman & Ralph [25]

No (parents)

Possibly (teenager's self-report)

Yes

Intervention 19%

Control 11%

Yes, but only per protocol results tabulated

Yes

Yes. Australian Rotary Health Research Fund, grant

Not ECBI/CBCL

Turner et al. [67]

No

Yes

Intervention 23%

Control 28%

No

Yes

Yes. Queensland Health and Queensland Department of Premier and Cabinet

Yes

Turner & Sanders [68]

Yes (video observations)

No (parent report)

Yes

Intervention 19%

Control 14%

For measures with a significant univariate condition effect at post-assessment

Yes

No

Yes

Turner et al. [29]

Yes (video observations)

No (parent report)

Yes

Intervention 0%

Control 11%

No

Yes

Yes. National Health and Medical Research Council of Australia

Not ECBI/CBCL

West et al. [22]

No

Yes

Intervention 21%

Control 6%

Yes

Yes

Yes. Telstra

Foundation

Not ECBI/CBCL

Whittingham et al. [24]

No

Yes

Intervention 0%

Control 10%

Yes

Yes

Yes. School of Psychology University of Queensland

Yes

Wiggins et al. [23]

No

Yes

Intervention 10%

Control 26%

Yes

Yes

No

Yes

Zubrick et al. [20]

No

No. Intervention area sample had younger children, less highly educated parents, more parenting problems and higher child ECBI scores. Different recruitment methods in intervention and control areas

Intervention 14%

Control 4%

Not applicable

Yes

Yes. Western Australian Department of Health

No - Not randomized, and uncorrected outcome data for control group not given


CBCL, Child Behavior Checklist; CCET, Couples Coping Enhancement Training; EBCI, Eyberg Child Behavior Inventory; EBFI, Enhanced Behavioural Family Intervention (level 5); SBFI, Standard Behavioural Family Intervention (level 4); SDBFI, Self-directed Behavioural Family Intervention (level 4).

Wilson et al. BMC Medicine 2012 10:130   doi:10.1186/1741-7015-10-130

Open Data