Table 2: Quality evaluation of data driven studies for question 2.

First Author/yearDesignEvidence Level [16]DurationDosageSettingCountrySample sizeAge meanAge rangeCharacteristicsGenderSamplingKey DVKey measureResultsScore (max27)* [15]

Spenko [28] 2006Pilot Intervention studyIV§§ Residential careUSA“several”§ § § § § § Distance from wall in meters before the smart-walker changed directionMotor performanceGood performance outcome using both PAMM technologies. Subjective measures gave less confidence in robotic-controlled walkers to manually controlled walkers.4
Saeki [32] 2008Case studyIV<6 mths20 mins 2 d/wk for 4 mthsHomeJapan148§ Neurological01§ Oxygenated/ deoxygenated/ total haemoglobin level in motor cortex assessed by fMRI techniqueMotor assessment scale (modified Ashworth Scale)Improvement in modified Ashworth motor assessment scale (2 to 5); reduced score on modified Ashworth scale (3 to 2); no change in wrist and fine motor tasks; direct activation of motor area in affected hemisphere.4
Shimada [40] 2009Intervention studyIII-2<6 mths  mins/wk for 3 mthsRetirement villageJapan1578.372–85Healthy 015Conven-ienceWalking speed5 m walk testIncrease in walking speed; reduction in energy consumption (lower glucose metabolism)12
Carlson and Demiris [51] 2012Intervention studyIII-2<1 d  minsSimulated “home” environmentEngland21§ 17–47Healthy§ § Conven-ienceWheel-chair control around a simulated home environmentCollision and cognitive perceptionLess collisions with robotic assistance and lower scores on perceptions of concentration13

§Not given.
*The last question of the Downs and Black assessment tool was excluded due to ambiguity of the question. This meant that the maximum score possible was 27.