Skip to main content
. 2018 Sep 8;8(9):e022560. doi: 10.1136/bmjopen-2018-022560

Table 2.

Design and outcome characteristics of primary-school based studies

Study (% quality rating) Study design Measures Follow-up Effects/outcomes
Attwood et al 30
(34%)		 Randomised pre-post intervention evaluation using opportunistic sample. No blinding or randomisation procedure reported. ‘cCBT’ (n=6) × control group (n=7).

  • SCAS – Parent & Child version.

  • SDQ – parent version.

  • Focus groups (n=8).

 Baseline; 6 weeks postintervention. Significant reduction in SCAS-C ‘social’ (d=0.49*) and ‘general anxiety’ (d=0.48*) subscales (note: intervention group significantly higher on SCAS at baseline). No effects on parent rated measures.
Berry et al 31
(68.8%)		 Randomised controlled trial; web randomisation system.
29 schools ‘PATHS’ intervention × 27 schools WL control.†

  • SDQ – teacher version.

  • PATHS teacher rating scale (PTRS).

  • T-POT.

 Baseline; 12-month postintervention; 24-month postintervention. No differences on SDQ at 12-month follow-uup. Some significant results on subscales of PTRS at 12-month follow-up (social competence: d=0.09*; aggression: d=0.14*; inattention: d=−0.06*; peer relations: −0.10*). Not maintained at 24-month follow-up.
Collins et al 32
(46.9%)		 Randomised 3×3 mixed design. No randomisation procedure reported.
Psychologist-led anxiety intervention (n=103) × teacher-led anxiety intervention (n=79) × controls (n=135).

  • CSI

  • SCAS – Child version administered by teachers.

 Baseline; postintervention; (within 3 weeks of end); 6-month follow-up. Improvement in psychologist-led and teacher-led groups on SCAS-C (d=0.41*; d=0.31*) and CSI ‘Avoidance’ (d=0.31*; d=0.31*) and ‘problem solving’ (d=−0.66*; d=0.52*) subscales. No difference between psychologist or teacher-led groups. SCAS-C outcomes maintained at 6-month follow-up (d=0.39*; d=0.39*). Noted: those lost to follow-up (n=155) were not included in analysis.
Stallard et al 33
(43.4%)		 Pre-post evaluation of pupils (n=106) from three schools taking part in the FRIENDS intervention.
No controls employed.

  • SCAS-Child version.

  • CFSEQ.

 ‘T1’: 6 months prior; ‘T2’: prior to intervention; ‘T3’: 3-month follow-up. Improvements in SCAS (d=−0.50*) and CFSEQ (d=0.58*) from T1 to T3 for whole sample; not between T2 and T3 (across intervention). Improvements on both measures (d=−1.26*; d=−1.27*) for ‘high risk’ group between T2 and T3.
Stallard et al 34
(75%)		 Cluster randomised controlled trial randomised through computer tool.
Health-led FRIENDS (n=489) × school-led FRIENDS (n=472) × controls (n=401).†

  • RCADS 30 – child & parent.

  • Penn State Worry Questionnaire.

  • RSES.

  • Bully/victim questionnaire.

  • Subjective well-being assessment.

  • SDQ – Parent version; teachers completed ‘Impact scale’.

 Baseline; 6-month follow-up; 12-month follow-up. Improvement on total RCADS (d=0.20*) and social (d=−0.09*) and general anxiety subscales (d=−0.20*) – not depression. Smaller effect sizes in school-led group (d=0.02*; d=0.11*; d=0.01*). No statistical improvements on secondary outcome measures or teacher/parent rating scales.

*Significant at p<0.5 level.

†Study sufficiently powered to detect change.

CFSEQ, Culture-Free Self-esteem Questionnaire; CSI, Coping Strategy Indicator; PATHS, Promoting Alternative Thinking Strategies; PTRS, PATHS teacher rating scale; RCADS, Revised Child Anxiety and Depression Scale; RSES, Rosenberg Self-Esteem Scale; SCAS, Spence Children’s Anxiety Scale; SDQ, Strength and Difficulties Questionnaire; T-POT, Teacher Pupil Observation Tool.