Girolametto 1998.

Methods	Design: quasi‐RCT
Participants	Location: Toronto, Canada Setting: participants were drawn from early intervention services Child participants Sample size: 12 children (intervention: 6, control: 6) with Down syndrome Mean age: intervention: 39.2 months (range 29‐44 months), control: 37.2 months (range 32‐41 months) IQ: intervention: 59‐93, control: 65‐103 Inclusion criteria: children communicated using at least 10 single words or signs with no word combinations, had a confirmed diagnosis of trisomy 21, and English was the only language of the home Comorbid conditions: 3 children (intervention: 1, control: 2) had mild hearing losses but did not use hearing aids. Most children had hearing that was within normal limits as assessed by a paediatric audiologist. Number of children per family: 2.7 (average), with 1 singleton in each group Parent participants Sample size: 12 mothers Mean age: 32 years (range 23‐34 years), across both groups Education: all mothers had completed at least high school, with 10 completing additional postsecondary education Marital status: all families described as being 'intact' Occupation: 7 mothers were homemakers, and the remainder were employed outside the house on at least a part‐time basis Socioeconomic status: middle class
Interventions	The 12 participants were randomly assigned to 1 of 2 groups. Intervention group (n = 6): Hanen Parent Program adapted for a focused stimulation approach. The intervention taught parents to model language at their child's level during naturally occurring situations. There were 9 group sessions, each lasting 2.5 hours, and 4 individual home sessions (time unspecified) with videotaped feedback to coach mothers on their use of the techniques, which took place over a 13‐week period. The total intervention time was approximately 26.5 hours. Mothers also chose up to 20 target words for their children to learn and were taught how to set up routines to allow for opportunities to model the target words, and to use signs as they spoke to the children. Children in the intervention group did not participate in any other therapy during the parent program. Control group (n = 6): usual language intervention services. Families in the control group continued to receive language intervention through their regular preschool services.
Outcomes	The measures listed below were used to measure the outcomes over 2 × 90‐minute sessions within 3 weeks following the intervention. Mother‐child free play session to measure the child's use of 20 target words A semi‐structured probe on the child's expressive use of 20 target words A free‐play experimental probe to measure the use of 20 target words Mervis's adaptation of the Communicative Development Inventory Changes in maternal interactional behaviours, based on a 15‐minute sample of videotaped interaction rated for rate of talk (number of utterances/min), complexity of language input (MLU in morphemes and type token ratio) and use of labels (number of focused target words). This was also measured through a consumer questionnaire completed by mothers about their use of strategies and observations from the therapist based on the home visits. There was no report on adherence to the intervention by the clinician, although this was measured for the parent. Parental dosage (intervention fidelity) was not reported.
Notes	Study start and end dates: not reported Funding source: grant from the Hospital for Sick Children Foundation, Toronto, Ontario, Canada Conflict of interest: none reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: no indication of how randomisation was carried out
Allocation concealment (selection bias)	Unclear risk	Comment: did not report if this was conducted
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: not possible
Blinding of outcome assessment (detection bias) All outcomes	High risk	Comment: parent‐report measures were used and parents were not blind to group allocation. In addition, they did not report if the raters of the observational assessments were blind to group allocation.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: seemed to be no missing outcome data
Selective reporting (reporting bias)	High risk	Comment: the receptive scale of the SICD was completed after the intervention but was not reported.
Other bias	High risk	Comment: control group continued to receive their regular speech and language therapy input, and it was unclear how much parents were involved, but the intervention group did not. Target words were chosen for both groups to be measured after the intervention, but only the intervention group were made aware of these targets.