Advice
Clinical and technical evidence
Clinical and technical evidence
A literature search was carried out for this briefing in accordance with the published process and methods statement. This briefing includes the most relevant/best publicly‑available evidence relating to the clinical and cost effectiveness of the technology. The literature search strategy, evidence selection methods and detailed data extraction tables are available on request by contacting medtech@nice.org.uk.
Published evidence
Two studies were selected for inclusion in this briefing, on the basis of the most relevant clinical outcomes. The first was a re‑analysis of data from 2 datasets, taken from participants in an ageing research project, and included 442 people (Greene et al. 2016). The study outcome was the prediction of falls risk. The study reported that fall‑risk assessment accuracy was greatest when QTUG was used alongside clinical risk factor assessment. Accuracy was greater when using QTUG alone compared with clinical risk assessment alone.
The second study was a cohort study assessing frailty in 399 people (Greene 2014). The evidence suggests that the assessment of frailty was most accurate when a TUG test with inertial sensors (such as those used in QTUG) was used.
Table 2 summarises the clinical evidence as well as its strengths and limitations.
Table 2: Summary of selected studies
|
Study |
Details of intervention and comparator |
Outcomes |
Strengths and limitations |
|
Study title: 'Fall risk assessment through automatic combination of clinical fall risk factors and body‑worn sensor data'. 442 (of 748 participants recruited). Retrospective analysis of 2 data sets. Single centre (Ireland). |
QTUG sensor based risk factor assessment. Clinical risk factor assessment (questionnaire). QTUG and clinical risk factor assessment (combined). Independent validation sample, QTUG sensor based risk factor assessment (n=22). |
The study favoured combined QTUG and clinical risk factor assessment compared with either QTUG alone or clinical risk assessment alone in terms of the percentage of participants correctly classified as being a 'faller' or 'non‑faller'. |
Limited details of participants provided. A large proportion of the people in the study did not have data recorded. It was unclear what questions were used in the clinical risk factor assessment questionnaire or whether this was validated. QTUG was validated by classifying participants into 'faller' or 'non‑faller' based on retrospective falls reports, not on prospective falls outcomes. The paper reported on 2 data sets but data is presented for data sets 1 and 1+2 combined (not data set 2 separately). Results differed between dataset 1 and dataset 1+2 combined, suggesting differences between datasets. Few measures of variance around estimates are presented in the paper. The study also reported clinical risk factor assessments for all 748 participants, with more favourable results than the subgroup. The study used self‑reported history of falls, which can be unreliable, to classify participants. The independent validation sample was small. |
|
Study title: 'Frailty status can be accurately assessed using. inertial sensors and the TUG test'. 399 (of 479 participants recruited). Cross‑sectional study. Single centre (Ireland). |
TUG sensor based assessment. Manual TUG based assessment. Maximum grip strength (lbs) from left and right hand (using a handheld dynamometer). All were compared with frailty category (as defined using modified Fried criteria). |
The study favoured TUG with inertial sensors compared with manual TUG in terms of the proportion of participants identified by the system as 'frail' or 'non‑frail'. The study favoured maximum grip strength compared with manual TUG or TUG with sensors when models are stratified by gender (not in the single regression model). |
Owing to missing data, approximately 17% of participants were not analysed. Significant differences between frailty subgroups at baseline were identified. 'Pre‑frail' and 'frail' categories were combined because of the small number of participants in the 'frail' category. Results from a single regression model of all participants and from the mean of models stratified by gender were reported. Results cited as headline were from the means of models stratified by gender. Few measures of variance around estimates were included. |
Strengths and limitations of the evidence
The 2 studies identified were both observational studies and no randomised studies were identified. Limited details of the studies were reported and as such there is a risk of bias that may influence the results. Data were missing for a large proportion of people in both studies. Neither of the studies describe any blinding of outcome assessors, therefore it is uncertain whether test results were interpreted without knowledge of the results of the other tests that were used. These studies were both carried out as part of a larger cohort study and it is possible that some of the same people were included in both studies. Neither of these studies included adults with physical disability or neurological diseases and the effectiveness of QTUG to predict frailty and falls in these populations is unclear. The lead author of both studies is a Director of Kinesis Health Technologies Ltd, which manufactures QTUG.