To address the process of ATD selection, participants in a longitudinal (prospective) multi-cohort study completed the Assistive Technology Device Predisposition Assessment (ATD PA) at baseline and 6 months follow-up. The ATD PA items identified consumer predispositions to ATD use as well as AT and user match. It is concluded that the ATD PA is a valid measure of predisposition to use an ATD and the subsequent match of ATD and user. Practitioners who use the ATD PA will may achieve enhanced evidence-based practice and ATD service delivery outcomes.
Assistive technology device, evidence-based practice, outcomes research
The Consortium on Assistive Technology Outcomes Research (CATOR) developed the Framework for the Conceptual Modelling of Assistive Technology Device (ATD) Outcomes (1). The framework assumes antecedent factors to the procurement of a particular ATD, which is the starting point in the CATOR framework . This paper discusses factors found to influence consumer predispositions to ATD selection and subsequent match with the selected ATD.
Outcomes of ATD use are important indicators of a quality service delivery process (1-3). From the perspective of consumers, equally or more important is achieving an optimal match of user and ATD at the outset and being involved in ATD selection (4,5). Recent developments in assessment research confirm the importance of "therapeutic assessment"(6,7) and an appropriate early assessment of consumer needs for ATD. As shown in Figure 1, the decision that a particular ATD is an appropriate and desirable support for an individual is proposed to be the result of a process that is affected by unique consumer perspectives predisposing the individual to the selection of a particular ATD.
To determine the adequacy of this supposition, a study was conducted to address the role of (a) consumer expectations and preferences and (b) personal characteristics on their (c) ATD predisposition or "subjective need" for an ATD as well as subsequent match of ATD and user at 6 months.
The 213 respondents at baseline were all participants in the Locomotion Device Outcomes Study (LDOS), a longitudinal investigation of assistive technology device users following the onset of a complex medical condition, hip fracture, or stroke and who were provided with one or more ATDs for mobility. The goal of this research is to gain a better understanding of the factors contributing to ATD abandonment and discontinuance. The LDOS includes baseline data collected with the first month post discharge from acute rehabilitation as well as repeated measurements at 6-, 12-, and 24-months follow-up on a variety of variables related to device effectiveness and subjective well-being.
The Assistive Technology Device Predisposition Assessment (ATD PA) Person and Device Forms was one of the measures used in the LDOS study (8). Section A of the ATD PA Person Form consists of 9 items addressing subjective functioning. Sections B consists of 12 items addressing quality of life and subjective well-being. Section C is composed of 33 items characterizing consumer personal characteristics divided into eight subscales as follows: Mood, Self-Esteem, Self-determination, Autonomy, Family Support, Friend Support, Therapist and Program Reliance, and Motivation to Use Support. The ATD PA Device Form has 12 items asking respondents to rate their expectations of the AT under consideration (or realization of benefit from ATD use at follow-up). Sections A, B, and D use 5-point Likert scales whereas the Section C items are yes/no.
It is important to know how clear, nonambiguous, reliable and consistent responses are among the items on the ATD PA Forms. The reliability of each scale was empirically examined through the calculation of Cronbach's Alpha coefficients. The greater the consistency and reliability, the higher coefficient alpha will be. Generally, scales that obtain alpha levels of 0.70 or greater are considered to be reliable. The 12 Section B items were considered to be a single scale, as were the 12 Device Form items. Ten items from Section C indicating a negative mood (9) were used in the Cronbach's alpha calculation.
A score of 60 on the ATD PA Device Form means the individual expects to experience maximum benefit from use of the AT (12 items x maximum score of 5 points = 60). Thus, a score of 60 would indicate a highly favorable predisposition to use the particular ATD. In reality, however, respondents vary in their predispositions to use a selected ATD. A 3-group categorical dependent variable was statistically created (weak, moderate, strong ATD Predisposition ) as determined by the percentile score distribution computed by SPSS 11.0. Discriminant analyses were performed to see which of the Section B and Section C items on the ATD PA Person form predicted ATD Predisposition and AT and User Match at 6 months .
Based on the results of measurement standards applied to date, and through a number of validation studies, the ATD PA has been determined to have reasonable inter-rater reliability and validity (9, 10).
The general research protocol was reviewed and approved by the Committee on Investigations Involving Human Subjects, Boston University.
Cronbach's Alpha scores for each of the sections of the ATD PA were calculated at baseline and the obtained value for the 22 items on Sections B and C of the Person Form was .79 and it was .74 for the 12-item Device Form. All values of Alpha are above .70 and we can generally conclude that the subscales of the ATD PA Forms are reliable.
As expected, respondents varied in their predispositions to the ATD according to their ATD PA Device Form total scores (weak, moderate, strong predisposition at baseline and match at 6-month follow-up). This 3-group categorical dependent variable was statistically computed by SPSS 11.0 based on percentile score of the ATD PA Device Form total value. Only cases with complete data were used (a missing value on any variable disqualified the case from inclusion).
Discriminant function analyses were used to determine which predictor variables discriminate among the three predisposition groups. The predictors were all 45 Section B and Section C items on the ATD PA Person form. For both baseline and 6-months follow-up, the overall Wilks' Lambda was significant at the p =.01 level or better, indicating that the 45 predictors (45 Section B and Section C items) differentiated the three groups at both baseline and 6 months follow-up. The findings gain further strength when prediction of group membership was calculated. The percent correctly predicted into one of the three groups (weak, moderate, strong ATD Predisposition) at 6 months follow-up was 98.7%, which indicates strong accurate prediction.
The more positively respondents scored their subjective quality of life and their expectations and preferences and personal characteristics, the more favorably predisposed they were to AT use and the better their match with the selected ATD at 6 months follow up. Thus, the hypothesis guiding the analyses has been shown to be correct: specific personal and psychosocial variables exist which serve as predictors of a person's predisposition to using, and subsequent match with, a given ATD. Specifically, it can be said that a quantifiable relationship exists between the ATD PA's measure of subjective quality of life and well-being (Section B items) and personal characteristics (Section C items) such that it is possible to identify a consumer's predisposition to use a particular ATD as measured by the 12-item ATD PA device form and that these same factors are predictive of the subsequent AT and User Match .
The authors are all members of the Consortium on Assistive Technology Outcomes Research (CATOR, http://www.AToutcomes.com). This work was supported in part by grant H133A010401 from the National Institute on Disability and Rehabilitation Research.
This should be in the right column.