Table 3 Summary of the Evidence
Study | Quality Rating | Factors affecting qualityStrengths (+)/Weaknesses(-) | Recommendation* | Best Estimates |
|---|---|---|---|---|
Aldred et al. (2005) | High | + RCT + Independent randomization + Allocation Concealment/blinding + Clear outcome reporting + Low risk of bias + Valid outcome measures | Weak + | Improved symptom severity, social communication skills and parent perception of child's communication skills |
Drew et al. (2002) | Moderate | + RCT + Appropriate randomization + Clear outcome reporting + Low risk of bias +Valid outcome measures - poorly matched groups (gender) - lack of blinding | Weak + | Improved Communication Skills |
Howlin & Rutter (1987) | Low | + Appropriately matched control group + Allocation Concealment/blinding + Clear outcome reporting + Low risk of bias + Valid outcome measures - Quasi-Experimental design | Weak + | Improved mother/child speech and interaction |
Koegel et al. (1996) | High | + RCT + Appropriate randomization + Allocation Concealment/blinding + Clear outcome reporting + Low risk of bias + Valid outcome measures | Weak + | Improved parents' global style of interaction with their child |
McConachie et al. (2005) | Low | + RCT + Appropriately matched control group + Allocation Concealment/blinding + Clear outcome reporting + Low risk of bias + Valid outcome measures - Quasi-Experimental design | Weak + | Improved children's vocabulary and parents' use of facilitative strategies |
Neef (1997) | Very Low | + Low risk of bias - Quasi-Experimental design - Inappropriately matched control group - Allocation Concealment/blinding - Incomplete outcome reporting - Undetermined validity of outcome measures | Weak + | Improved parent teaching skills and child skills development |
Ozonoff & Cathcart (1998) | Very Low | + Appropriately matched control group + Low risk of bias + Valid outcome measures - Quasi-Experimental design - lack of blinding - Incomplete outcome reporting | Weak + | Improved overall development, imitation, fine and gross motor skills and nonverbal conceptual skills |
Rickards et al. (2000) | High | + RCT + Appropriate randomization + Allocation Concealment/blinding + Clear outcome reporting + Low risk of bias + Valid outcome measures | Weak + | Improved children's IQ and behaviour problems |
Schreibman et al. (1995) | Moderate | + RCT + Appropriate randomization + Allocation Concealment/blinding + Low risk of bias + Valid outcome measures - Incomplete outcome reporting | Weak + | Improved parental affect while interacting with their child |
Tonge et al. (2001) | High | + RCT + Independent randomization + Allocation Concealment/blinding + Clear outcome reporting + Low risk of bias + Valid outcome measures | Weak + | Improved self-reported parent mental health |
Wang (2008) | High | + RCT + Appropriate randomization + Allocation Concealment/blinding + Clear outcome reporting + Low risk of bias + Valid outcome measures | Weak + | Improved the quality of parent's interactions with their child |