Table 2

Quality of included studies

First authorAllocation concealmentBlindingFollow upClinically important outcomes
*Archival data used for objective outcome assessments.
Henggeler4 (USA) 1992Adequate (sealed envelope)Yes*100% for archival dataReported
66% for psychosocial measures
Chamberlain5 (USA) 1998Adequate (coin toss)Yes*87% for archival dataReported
Borduin17 (USA) 1995Adequate (coin toss)Yes*94.5% for archival dataReported
71.5% for psychosocial measures
Henggeler19 (USA) 1997Adequate (sealed envelope)Yes*90% for archival and psychosocial measuresReported
Alexander15 (USA) 1973Inadequate (some subjects not randomised)Yes*86% for archival dataReported
Bank16 (USA) 1990UnclearYes*88% for archival dataReported
Emshoff18 (USA) 1983UnclearYes*Attrition data not givenReported
Raue20 (Aust) 1985UnclearNo86% for psychosocial measuresNot reported