Table 8 Summary of quality appraisal of included studies
| Â | Author/Year | Key Sources of Bias |
|---|---|---|
Summary of quality assessment for RCTs | ||
1 | Research Units on Pediatric Psychopharmacology Autism Network. 2005 [20] | • parent raters • process for randomisation, recruitment and sampling unclear • short follow up period |
2 | Ahmed et al. 2000 [21] | • selection bias, not blinded, no allocation concealment, process for randomisation, recruitment and sampling unclear • no reporting of PRN prescribing and administration, non-pharmacological interventions, level of support, co morbidities, level of ID • baseline characteristics of experimental and control groups uneven • short follow up period |
3 | de Kuijper et al. 2014 [23] | • selection bias, not blinded, no allocation concealment, process for randomisation, recruitment and sampling insufficient. • no reporting of other psychotropic medication prescribing, PRN prescribing and administration, non-pharmacological interventions, level of support, co morbidities • short follow up period |
4 | de Kuijper, G., et al. 2013 [24] | • side arm of previous study • outcomes are statistically significant but unclear if clinically significant. • lack of evaluation of confounding factors e.g. changes in diet and exercise • measurements and results were not AP specifically reported |
5 | de Kuijper et al. 2014 [25] | • side arm of previous study • confounding factors that could have affected the results include linking effects to the actual AP eg risperidone has greater effect on prolactin than others in the sample, olanzapine has a greater effect on weight gain. • measurements and results were not AP specifically reported |
6 | Haessler et al. 2007 [28] | • recruitment, randomisation and blinding process unclear • no power calculation • baseline comparability unclear • short follow up period • unclear if outcomes were discontinuation effects or reduced effects of placebo • no tapering • no reporting of other psychotropic medication prescribing, PRN prescribing and administration, non-pharmacological interventions, level of support, co morbidities |
7 | Hassler et al. 2011 [29] | • see no 6 Haessler, F., et al. 2007 • small sample • not blinded |
8 | Heistad et al. 1982 [30] | • no power calculation • rating scales not specified • rate of discontinuation unclear • process of randomisation unclear • simultaneous withdrawal of antiparkinsonian medication, • no reporting of PRN prescribing and administration, non-pharmacological interventions, level of support, co morbidities • short follow up period |
9 | McNamara et al. 2017 [31] | • significantly underpowered • trial finished prematurely and reported as pilot |
10 | Ramerman et al. 2019 [32] | • no allocation concealment, • no power calculation • no reporting of other psychotropic medication prescribing, PRN prescribing and administration, non-pharmacological interventions, level of support, co morbidities |
11 | Smith et al. 2002 [22] | • See no 2 Ahmed, Z., et al. 2000 |
Summary of quality assessment for non randomised controlled trials (CTs) | ||
1 | Aman et al. 1985 [35] | • subjective outcome measurements • sampling and recruitment process unclear • no power calculation • short follow up period |
2 | Carpenter et al. 1990 [36] | • selection method unclear • exposure inadequately ascertained • causality inadequately ascertained • short follow up period |
3 | Gerrard et al. 2019 [37] | • recruitment and allocation process unclear • length of follow up not reported. |
4 | Swanson et al. 1996 [39] | • selection bias • control group inadequately matched • inadequate blinding • statistics or statistical tests inadequately reported or inappropriate • institutional setting • intervention poorly defined |
5 | Wigal et al. 1993 [40] | • selection bias • statistics or statistical tests inadequately reported or inappropriate • missing baseline information • intervention poorly defined |
6 | Wigal et al. 1994 [41] | • selection bias • statistics or statistical tests inadequately reported or inappropriate • missing baseline information • intervention poorly defined |
7 | Zuddas et al. 2000 [42] | • no power calculation and small number of participants • sampling and recruitment unclear • confounding factors include psychological, behavioural and environmental interventions |
Summary of quality assessment for non randomised no control Pre Post Studies (PPSs) | ||
1 | Brahm et al. 2003 [43] | • missing baseline information • variable deprescribing schedules |
2 | Branford 1996 [44] | • patients living with relatives, those in unsupervised accommodation, and those in accommodation where staff were unwilling to engage excluded from study • selection bias • use of unvalidated measures or non-standard assessment tools • missing baseline information • intervention poorly defined • selective reporting or incomplete |
3 | de Kuijper et al. 2018 [33] | • sampling and recruitment unclear • rater reliability |
4 | de Kuijper et al. 2018 [33] | • as above |
5 | Ellenor et al. 1977 [46] | • intervention poorly defined • outcomes measures unclear |
6 | Ferguson et al. 1982 [47] | • duration of intervention variable • length of follow up not reported |
7 | Fielding et al. 1980 [48] | • unvalidated outcome measures |
8 | Findholt et al. 1990 [49] | • high turnover of medical staff delivering the intervention |
9 | Gerrard 2020 [38] | • author / researcher is the clinician delivering the intervention |
10 | Howerton et al. 2002 [50] | • differing referral rates from the various primary care providers • poor follow up rates |
11 | Inoue et al. 1982 [51] | • limited baseline information |
12 | Janowsky et al. 2006 [52] | • selection bias • missing baseline information • intervention poorly defined |
13 | Janowsky et al. 2008 [53] | • selection bias • missing baseline information • intervention poorly defined |
14 | Jauernig et al. 1995 [54] | • selection bias • use of unvalidated measures or non-standard assessment tools • missing baseline information • intervention poorly defined |
15 | LaMendola et al. 1980 [55] | • intervention poorly defined • duration of intervention and length of follow up not reported • missing baseline and outcomes information |
16 | Lindsay et al. 2004 [56] | • poorly defined methodology • small sample size • inconsistent weighing scales • no BMI measurements • missing data • no reporting of dietary modification, environmental and behavioural interventions |
17 | Luchins et al. 2004 [57] | • poor reporting of duration of intervention and length of follow up |
18 | Marcoux 1985 [60] | • intervention poorly defined |
19 | Marholin et al. 1979 [58] | • selection method unclear • causality not adequately ascertained • short follow up |
20 | Matthews et al. 2003 [59] | • duration of intervention and length of follow up missing • outcomes poorly reported |
21 | May et al. 1995 [61] | • small sample size • selection bias • use of unvalidated measures or non-standard assessment tools • statistics or statistical tests inadequately reported or inappropriate • missing baseline information • selective reporting or incomplete outcome data |
22 | Newell et al. 2000 [62] | • small sample size • selection bias • use of unvalidated tools • missing baseline and outcome data |
23 | Newell et al. 2001 [63] | • selection bias |
24 | Newell et al. 2002 [64] | • selection bias • inadequate blinding • use of unvalidated measures or non-standard assessment tools • statistics or statistical tests inadequately reported or inappropriate • missing baseline information |
25 | Ramerman et al. 2019 [32] | • weak methodology of combining studies with different designs |
26 | Shankar et al. 2019 [65] | • unvalidated outcome tools |
27 | Spreat et al. 1993 [66] | • selection bias • institutional setting • missing baseline information • intervention poorly defined |
28 | Stevenson et al. 2004 [67] | • weak methodology • use of non standardised assessment tools • subjective outcome measurements |
Summary of quality assessment of case studies | ||
1 | Adams and Sawhney 2017 [68] | • selection method unclear |
2 | Bastiampillai et al. 2014 [69] | – |
3 | Brahm et al. 2009 [70] | – |
4 | Branford 2019 [14] | • selection method unclear |
5 | Connor D 1998 [71] | Â |
6 | Dillon J 1990 [72] | • outcome and causality inadequately ascertained |
7 | Faisal et al. [73] | • |
8 | Ghaziuddin et al. 1990 [74] | – |
9 | Lee et al. 2019 [75] | • selection method unclear |
10 | McLennan J 2019 [76] | – |