Development of an activity-based therapy tracking tool: Item generation and reduction using the Delphi method

Abstract

Context/Objective

Activity-based therapies (ABT) are increasingly used in rehabilitation after spinal cord injury or disease (SCI/D). However, the absence of standardized tools to track the details of an ABT program hinders the collection of data needed for client-tailored programming and resource allocation. The objective of this study is to determine the content to include in an ABT tracking tool for people living with SCI/D.

Design

Cross-sectional e-survey.

Setting

Community.

Participants

The 60 participants from Canada and the United States who had knowledge and/or experience with ABT included: individuals with SCI/D; hospital clinicians (i.e. physical and occupational therapists/assistants); community-based clinicians; hospital or community clinic administrators; researchers; and funders, advocates and policy makers.

Interventions

None.

Outcome Measures

A Delphi e-survey comprised 16 types of ABT (e.g. treadmill training) and 4 types of technology (e.g. virtual reality). Participants rated the importance of including each item on a tracking tool and the feasibility to track each item using a 9-point Likert scale.

Results

After two survey rounds, nine types of ABT and one technology were identified as important to include in a tracking tool. All items rated as important were considered feasible for clinicians and people with SCI/D to track, except crawling.

Conclusion

This study identified the types of ABT and technology to include in an ABT tracking tool. Such a tool may provide details of an ABT program that can support decision-making at the individual, program and health system levels and aid the development of best practice guidelines.

Introduction

The landscape of rehabilitation for people impacted by spinal cord injury or disease (SCI/D) has undergone a transformation over recent years with an increasing focus on maximizing function through neurorecovery. Movement-based therapies that promote neurorecovery, called activity-based therapy (ABT), involve “repetitive neuromuscular activation below the level of spinal injury, typically achieved through intensive, task-specific movement practice” (1). Task-specific movements are discrete, goal-directed movements of the limbs, such as walking, transferring from sitting to standing and grasping objects. Beyond recovery of motor function, ABT may help to reduce spasticity (2, 3) and the risk of metabolic and cardiovascular diseases (2, 4). ABT may also improve bone health (4–7), body composition (2), cardiorespiratory fitness (8, 9), and bowel, bladder and sexual function (10, 11).

Although ABT has shown promise as a viable intervention to optimize recovery for people with SCI/D, implementation challenges exist, such as a lack of standardized guidelines, lack of knowledge regarding the optimal timing and dosage and lack of tools to track the details of a session or program (12). The Canadian ABT Community of Practice (CoP) was formed in 2019 in response to the growing need for greater access to and quality of ABT in Canada. The diverse members of the CoP, which includes individuals with SCI/D, initially identified five priority areas that were later expanded to eight key priorities for advancing ABT research and practice in Canada (1, 13). One of these priorities is the development of a tool to track participation in ABT across the continuum of care. A tool of this nature could document important details of an individual’s therapy, such as the duration, frequency and intensity of their sessions, the types of exercises they perform and equipment they use, and how they progress so they are continuously challenged (14). When included in research studies, the tool may facilitate the collection of information needed to help determine optimal dosing and how dosing may be influenced by demographic and injury characteristics (15–17). This evidence may inform decision-making at all levels of the healthcare system and lead to the development of best practice guidelines (18).

A review of the literature identified a few tools for tracking participation and/or performance in SCI rehabilitation. Groups from Europe (Spinal Cord Injury-Interventions Classification System (SCI-ICS)) (19–21) and the United States (SCIRehab Project) (22–25) each created a taxonomy to document the details of SCI rehabilitation. Although both tools are comprehensive, they include activities unrelated to ABT and lack the necessary detail to track exercise intensity (e.g. rate of perceived exertion or heart rate) and challenge level (e.g. number of repetitions, speed, distance); both fundamental aspects of an ABT program. Currently, there is an absence of tools for clinicians and individuals with SCI/D to use to specifically track the details of an ABT session and program across the continuum of care. Therefore, we elected to develop an ABT tracking tool that could be used to document specific details about an individual’s ABT session, such as the duration, types of activities performed, technology used and exercise intensity, to name a few.

Typically, the development of a measurement tool involves an iterative process that includes item generation and reduction (26, 27). Similar methods were followed to add strength and rigor to the development of the ABT tracking tool. Previous work has been undertaken to generate items through a two-step process. First, a scoping review was conducted to identify the characteristics of ABT across the care continuum for individuals with SCI/D (28). Next, experts from six key interest groups were consulted through focus group interviews to understand their perspectives on the details they thought were important to track (29). The current study focused on item generation and reduction using the Delphi method. This multi-stage approach is ideal when seeking consensus from a large and diverse group of experts over a wide geographical area where anonymity is preserved (30–32). As such, the objective of this study is to determine the content to include in an ABT tracking tool for people living with SCI/D using the Delphi method.

Materials and methods

Study design

This quantitative, cross-sectional study used the Delphi method to reach a consensus on the items to include in an ABT tracking tool through an online survey. Using an e-Delphi method was advantageous as we strived for a broad geographical reach of experts and completed this study during the COVID-19 pandemic (33). The study received ethical approval from the Research Ethics Board of the University Health Network (#20-5382). The checklist for reporting survey-based research, as outlined by Artino and colleagues (34), was followed.

Participants

Individuals living in Canada or the United States were recruited to participate in the study through an email campaign. Recruitment emails were sent to the members of the Canadian ABT CoP, the North American SCI Consortium (NASCIC) and community ABT clinics that were members of the Canadian ABT CoP. Interested individuals then contacted the study coordinator (KC). The individuals had to have knowledge of and/or experience with ABT (i.e. experience participating in, supervising, researching, funding or advocating for ABT) and belong to at least one of the following key interest groups: (1) researchers, (2) people living with SCI/D, (3) administrators in hospitals or community clinics, (4) hospital clinicians (i.e. physical and occupational therapists and therapy assistants), (5) community-based clinicians (e.g. occupational therapist, physical therapist, kinesiologist, exercise therapist), and (6) funders, advocates and policy makers.

There are no established guidelines for determining sample size in Delphi surveys, however, they typically range from 10 to 100 experts (35). Since this study included a heterogeneous sample of participants, we aimed to recruit 100 individuals in order to secure a sample of 60 participants (10 belonging to each group) that met inclusion for the study.

Survey development

Items for the initial survey were generated through the findings from (1) a scoping review that identified the characteristics of ABT across the continuum of care (28) and (2) a qualitative study using focus group interviews with experts of ABT to understand their perspective on tracking ABT across the care continuum (29). The review identified 16 types of ABT (e.g. treadmill training, load-bearing exercise, muscle strengthening exercises), four types of technology that were often combined with a type of ABT (e.g. transcutaneous electrical stimulation, virtual reality) and 84 parameters (e.g. walking speed, distance, resistance). The focus group interviews were conducted with 48 experts of ABT from across Canada who belonged to the same six key groups recruited for this study. In general, there was agreement in types of ABT, technology and parameters reported in the review and qualitative study. Due to the volume and specificity of the parameters reported in the scoping review, we restricted the Delphi survey to reaching a consensus on the types of ABT and technologies to include in an ABT tracking tool. As such, the initial survey developed contained 16 types of ABT and 4 types of technology (see Supplement 1).

Procedure

Guidelines provided by Hasson and colleagues were followed for conducting a Delphi survey (36). The four key criteria of the Delphi approach, namely anonymity, iteration, controlled feedback and statistical analyses of group responses, were incorporated into the current study (37, 38). The survey was hosted on a web-based platform (LimeSurvey version 3). Participants were asked to complete the survey if they agreed to participate after reading the consent form and responded ‘yes’ to the screening question ‘Do you have experience participating in, supervising, researching, funding or advocating for ABT?’. The Delphi survey consisted of two sections: demographic information and a series of questions querying the importance and feasibility for clinicians and people with SCI/D to track the details of different types of ABT and technology used in ABT. Demographic questions included the primary group they belonged to, sex, gender and age. Participants with SCI/D were also asked to indicate injury-related information and years of experience participating in ABT, while hospital and community clinicians were prompted to indicate their years of clinical experience and years delivering ABT.

Section two contained 20 questions, with each question describing one type of ABT or technology. A definition with a link to a picture for each type of ABT and technology was included in the survey. For each question, participants had to rate (1) the importance of including the type of ABT or technology in an ABT tracking tool and the feasibility for (2) clinicians supervising ABT and (3) people living with SCI/D to record information related to the type of ABT or technology using a 9-point Likert scale (1 = not important or not feasible and 9 = very important or highly feasible) (39). To assist participants with rating feasibility, examples of the information that may be tracked for each activity were provided. For example, for ergometer training participants were told that the revolutions per minute, work load and bout duration may be recorded (see Supplement 1). There was also an option to indicate if they were unfamiliar with that type of ABT and/or technology and therefore, would not provide the three ratings for that item. At the end of the survey, there was an open text section where participants could provide comments and/or suggest additional types of ABT and/or technology that were not listed, but would be important and feasible to include in an ABT tracking tool. The survey was piloted among the research team to ensure clarity.

Participant email addresses were collected in order to contact them for subsequent rounds. The first round of the survey was accessible for six weeks and email reminders were sent out every two weeks. Subsequent surveys were accessible for four weeks with email reminders sent out at the two- and three-week timepoints. All participants who completed the first survey were provided with a unique user ID to use for successive rounds and ensure anonymity. Following analysis of the first survey, items that reached consensus for the importance to include in an ABT tracking tool, as this was the primary aim of the study, were removed and the second survey was created with the remaining items that did not reach consensus. Group mean and standard deviation of the importance ratings were included, as well as any additional types of ABT or technology suggested by participants in the initial survey. Participants were asked to consider this feedback as they re-rated items on the importance to include in an ABT tracking tool for the second survey. For any new items suggested in the initial survey that were added to the second survey, participants had to rate both the importance to include in a tool and the feasibility to track for clinicians and people living with SCI/D. The process was repeated until consensus for the importance of each type of ABT and technology was achieved or a point of diminishing return was reached (30, 39).

Since the primary aim of the study was to identify the content to include in an ABT tracking tool, a high majority of participants would need to be in agreement on the importance of an item. As such, in order to achieve consensus on the types of ABT and technology that were deemed important to include in an ABT tracking tool, an a priori threshold of at least 70% of participants were required to rate that specific type of ABT or technology a seven or higher (39). Types of ABT and technology were considered feasible to track if more than 50% of one group gave it a rating of seven or more (38). Group data were summarized using descriptive statistics (39).

Results

Participants and response rates

Eighty-three individuals responded to the initial invite to participate in the study. Of those who responded, 60 individuals completed the initial survey. Reasons for exclusion were: (1) survey partially completed (n = 10), (2) no experience participating in, supervising, researching, funding or advocating for ABT (n = 9), and (3) survey opened but not started (n = 4) (see Fig. 1). Only two groups (people living with SCI/D and community-based exercise trainers), reached the sample size target of 10 participants. Three participants selected ‘other’ and identified belonging to multiple groups including researcher, engineer, person living with SCI/D, hospital clinician, community-based exercise trainer, funder and advocate. Except for administrators who were all women, the groups represented a mixture of different sexes, genders and ages. There was an even split between men and women among persons living with SCI/D, the majority of whom had complete tetraplegia of traumatic origin. Years of experience living with SCI/D and participating in ABT varied within this group (i.e. <1 year to >20 years, and <3 months to >5 years, respectively) (see Table 1). Most hospital and community clinicians had at least three years of experience as a clinician and practicing ABT. There was 15% attrition in the second survey as nine participants who completed the initial survey did not complete the subsequent survey. They included a researcher, two people living with SCI/D, two community clinic administrators, one hospital clinician and three community-based clinicians.

Figure 1.

Participant flow diagram.

Table 1.

 Participant demographics.

Group	N	Age (Yrs)	Sex	Gender	Clinical or SCI/D experience (Yrs)	ABT participation (Yrs)
SCI/D*	16	3 (18–29) 2 (30–39) 3 (40–49) 5 (50–59) 2 (60–69) 1 undisclosed	8 Male 8 Female	8 Man 8 Woman	1 (<1) 2 (1–5) 4 (6–10) 5 (11–15) 4 (20+)	3 (<0.25) 2 (0.6–0.9) 2 (1–2) 3 (3–5) 6 (5+)
Hospital clinician	9	5 (30–39) 3 (40–49) 1 undisclosed	2 Male 6 Female 1 undisclosed	2 Man 6 Woman 1 undisclosed	1 (3–5) 7 (5+) 1 not reported	1 (<0.5) 1 (0.6–0.9) 2 (3–5) 4 (5+) 1 not reported
Community-based clinician	19	9 (18–29) 7 (30–39) 3 (40–49)	5 Male 14 Female	5 Man 14 Woman	1 (<0.5) 1 (0.6–0.9) 3 (1–2) 4 (3–5) 9 (5+) 1 not reported	2 (0.6–0.9) 2 (1–2) 5 (3–5) 9 (5+) 1 not reported
Hospital and community administrator	5	1 (18–29) 1 (30–39) 2 (40–49) 1 (50–59)	5 Female	5 Woman	Not applicable	Not applicable
Researcher	8	1 (18–29) 1 (30–39) 4 (40–49) 1 (50–59) 1 (60–69)	4 Male 4 Female	4 Man 4 Woman	Not applicable	Not applicable
Other**	3	1 (18–29) 1 (30–39) 1 (40–49)	1 Male 2 Female	1 Man 2 Woman	Not applicable	Not applicable

This table includes demographics of participants from round 1 of the survey. N, number; Yrs, years; SCI/D, spinal cord injury or disease; ABT, activity-based therapy.

*Of the 16 participants with SCI/D, 13 had traumatic SCI and three had non-traumatic SCI, 10 had tetraplegia and six had paraplegia, and the American Spinal Injury Association Impairment Scale (AIS) distribution was AIS A (n = 10), B (n = 1), C (n = 2), D (n = 1), and unsure (n = 2).

**Participants who selected ‘other’ identified themselves as belonging to more than one group (e.g. one participant identified themself as a researcher, clinician and funder).

Survey round 1

Most respondents were familiar with the types of ABT and technologies included in the survey. The number of respondents who indicated they were unfamiliar with a specific type of ABT or technology in survey 1 ranged from 0 (0%) to 9 (15%) for the 16 types of ABT and 1 (1.7%) to 15 (25%) for the four technologies queried. The types of ABT and technologies participants were least familiar with (i.e. >10% of respondents were unfamiliar) were pool therapy (n = 9, 15%), running (n = 8, 13.3%), virtual reality (n = 11, 18.3%) and spinal stimulation (n = 15, 25%). After the initial round of the survey, eight types of ABT and one technology reached consensus for importance to include in an ABT tracking tool (73–92% of respondents rated the importance a 7 or higher). The types of ABT included muscle strengthening, balance training, load-bearing exercise, task-specific movement, transfer training, ergometer training, overground walking and treadmill training. The single technology that reached consensus by respondents was transcutaneous neuromuscular electrical stimulation (NMES) (see Table 2). All items were deemed feasible to track by clinicians (54–95%), whereas five items were not considered feasible to track by individuals with SCI/D (see Table 3). These items included plyometrics, crawling, pool therapy, vibration and virtual reality; none of which were rated sufficiently important to include in an ABT tracking tool. In the free text provided, respondents suggested including activities of daily living as an additional item. This item was added to the second survey.

Table 2.

 Survey ratings of importance.

	Range	Median	Mean	SD	% Respondents scoring ≥7
SURVEY ROUND 1
Type of ABT
Muscle strengthening (also known as resistance training, includes active assisted exercise, with or without weight machines and free weights, and may be robotic-assisted)	5–9	9	8.4	1.1	91.7
Balance training (includes activities that challenge stability in seated or upright positions)	5–9	9	8.3	1.1	90.0
Load-bearing exercise (any activity that involves supporting your own body weight, includes standing (e.g. standing frame, squat rack), tilt table standing, tall kneeling or holding a quadruped/4-point position)	2–9	9	8.3	1.6	85.0
Task-specific movement (includes discrete, goal-directed movements of the arms or legs, such as reaching and grasping and kicking a ball)	5–9	9	8.2	1.3	85.0
Transfer training (involves moving from one surface to another, such as from a wheelchair to a bed, or moving from one position to another, such as sitting to standing)	1–9	9	7.8	1.9	81.4
Ergometer training (includes arm crank, hand cycle, leg cycle, stationary bike, spin bike, arm and leg ergometer, and recumbent tricycle)	1–9	8	7.7	1.8	80.0
Overground walking (any type of walking overground, whether with or without body weight support, gait aids or an exoskeleton)	2–9	9	7.9	1.6	79.7
Walking on a treadmill (walking with or without body weight support, robot-assisted treadmill training, and walking on an anti-gravity treadmill)	3–9	8	7.6	1.7	73.2
Crawling (includes moving on hands and knees (i.e. quadruped crawling) or pulling the body forward with the arms (i.e. belly crawling))	1–9	7	7.1	2.1	67.8
Wheeling in a wheelchair (may involve wheeling on a treadmill or overground)	1–9	8	6.8	2.6	66.0
Plyometrics (a type of exercise training that uses speed and force of different movements to build muscle power (e.g. pushups, throwing, jumping, kicking))	2–9	8	7.0	2.2	61.0
Cross training (includes exercise machines that target whole body movements in an upright position, such as an elliptical trainer (with or without body weight support), or recumbent position (e.g. the NuStep))	1–9	7	6.8	2.1	58.9
Stair training (involves stepping up or down one or more steps with or without the use of a hand rail and/or gait aid)	1–9	7	6.4	2.2	52.7
Rowing	1–9	6.5	6.4	1.9	50.0
Pool therapy (any water-based movement performed as a therapy or exercise in a community or personal pool, includes walking or moving the arms and legs in water)	1–9	6	6.2	2.2	47.1
Running (may be performed on a treadmill or overground)	1–9	5	5.6	2.8	44.2
Technology used with ABT
Transcutaneous NMES (involves applying an electrical current to the body to cause muscles to contract or to cause a sensory response, stimulation is non-invasive (i.e. electrodes applied over top of the skin))	3–9	9	7.8	1.6	83.1
Spinal cord stimulation (involves applying an electrical current to the spinal cord either through the skin overlying the cord or through insertion of electrodes in the epidural space)	1–9	7	6.5	2.4	57.8
Vibration (includes whole body vibration during load-bearing exercises or vibration plate exercises)	1–9	7	6.4	2.5	50.9
Virtual reality (involves all levels of immersion, including computer games, video games and virtual reality headsets)	1–9	5	5.5	1.9	28.6
SURVEY ROUND 2
Type of ABT
Crawling	1–9	8	7.2	2.0	71.4
Wheeling in a wheelchair	1–9	7.5	6.5	2.7	60.0
Plyometrics	2–9	6.5	6.4	2.0	50.0
Cross training	2–9	7	6.8	1.8	60.4
Stair training	1–9	7	6.6	2.0	55.3
Rowing	3–9	6	6.1	1.8	34.0
Pool therapy	2–9	6	6.4	1.9	42.9
Running	1–9	5	5.5	2.3	34.0
Activities of daily living*	1–9	7	6.4	2.8	61.2
Technology used with ABT
Spinal cord stimulation	1–9	7	6.9	2.1	66.0
Vibration	1–9	7	6.4	2.3	51.0
Virtual reality	1–9	5.5	5.4	2.4	37.0

This table includes the survey results from round 1 and 2 of the ratings of importance for types of ABT and technology to be included in an ABT tracking tool. Additional descriptions about the types of ABT and technology that were provided in the Delphi survey are included in brackets in the left-most column. Participants were not provided with any additional details about rowing. Numbers in bold indicate items where ≥70% of respondents rated it as important. SD, standard deviation; ABT, activity-based therapy; NMES, neuromuscular electrical stimulation.

*This type of ABT was suggested for inclusion by a participant in survey round 1.

Table 3.

 Survey ratings of feasibility for clinicians and people with SCI/D.

		Range	Median	Mean	SD	% Respondents scoring ≥7
Type of ABT
Muscle strengthening	C	6–9	9	8.3	1.0	94.6
	SCI/D	3–9	8	7.9	1.3	84.2
Balance training	C	3–9	8	7.6	1.6	80.0
	SCI/D	1–9	7	6.8	1.8	59.7
Load-bearing exercise	C	2–9	9	7.9	1.7	81.8
	SCI/D	2–9	8	7.2	2.0	68.4
Task-specific movement	C	5–9	8	7.8	1.3	80.4
	SCI/D	3–9	7	7.1	1.6	66.7
Transfer training	C	2–9	8	7.6	1.6	85.5
	SCI/D	1–9	7	6.9	2.2	66.1
Ergometer training	C	3–9	8	8.0	1.3	91.1
	SCI/D	4–9	8	7.5	1.4	75.4
Overground walking	C	4–9	8	7.6	1.6	76.4
	SCI/D	2–9	7	6.8	2.0	57.1
Walking on a treadmill	C	3–9	9	7.9	1.7	83.3
	SCI/D	3–9	7	6.9	1.9	66.7
Crawling	C	1–9	8	7.3	2.0	74.6
	SCI/D	1–9	6	6.3	2.1	48.2
Wheeling in a wheelchair	C	1–9	7	7.1	2.0	72.6
	SCI/D	1–9	7	6.6	2.2	55.8
Plyometrics	C	2–9	7	7.2	1.7	65.5
	SCI/D	1–9	6.5	6.6	2.1	50.0
Cross training	C	3–9	8	7.8	1.5	84.6
	SCI/D	3–9	7	7.1	1.7	64.2
Stair training	C	1–9	8	7.5	1.7	78.4
	SCI/D	1–9	7	6.9	1.8	60.4
Rowing	C	3–9	8	7.9	1.4	84.9
	SCI/D	3–9	8	7.3	1.7	66.7
Pool therapy	C	2–9	7	6.6	1.9	54.4
	SCI/D	1–9	6	5.9	2.1	34.0
Running	C	1–9	8	7.1	2.0	73.5
	SCI/D	1–9	7	6.6	2.3	56.9
Activities of daily living*	C	1–9	6	5.6	2.6	41.7
	SCI/D	2–9	7	6.6	1.9	55.1
Technology used with ABT
Transcutaneous NMES	C	3–9	8.5	7.8	1.6	80.4
	SCI/D	2–9	7	6.7	2.1	56.9
Spinal cord stimulation	C	3–9	8	7.5	1.9	75.6
	SCI/D	2–9	7	6.6	2.3	51.2
Vibration	C	1–9	8	7.0	2.2	66.0
	SCI/D	1–9	6.5	6.5	2.5	50.0
Virtual reality	C	1–9	7	6.4	2.1	54.6
	SCI/D	1–9	5	6.0	2.3	41.9

This table includes the survey results of the ratings of feasibility for types of ABT and technology to track. Numbers in bold indicate items where >50% of respondents rated it as feasible. SD, standard deviation; ABT, activity-based therapy; C, clinician; SCI/D, spinal cord injury or disease; NMES, neuromuscular electrical stimulation.

*This type of ABT was suggested for inclusion by a participant in survey round 1.

Survey round 2

The second survey included nine types of ABT (i.e. eight types of ABT that did not reach consensus in survey 1 along with activities of daily living) and three types of technology. Following this round, only one additional item (crawling, 71.4%) reached consensus for importance to include in an ABT tracking tool (see Table 2). Most respondents (74.6%) deemed crawling feasible for clinicians to track; however, only 48.2% of respondents considered this type of ABT feasible for people with SCI/D to track (see Table 3). After round 2, the survey was terminated as it was determined that a point of diminishing return was reached.

Discussion

This study sought to identify the content to include in a tool that could be used by both clinicians and people living with SCI/D to track the details of participation in an ABT program. Researchers, people living with SCI/D, hospital and community clinic administrators and clinicians and funders, advocates and policy makers, provided input on the types of ABT and technologies to include through a Delphi e-survey. After two survey rounds, nine types of ABT and one technology were identified as important to include in an ABT tracking tool. All types of ABT and technology that were recognized as important to include in an ABT tracking tool were also considered to be feasible for both clinicians and people with SCI/D to track, with the exception of crawling. The feasibility of tracking was rated more highly for clinicians than for people with SCI/D for every type of ABT and technology aside from activities of daily living.

The types of ABT and technology that were identified as important to include in an ABT tracking tool in this study aligned with those most reported in a published scoping review (28) and qualitative study (29) (see Table 4). One exception was task-specific movement, which was identified as important in the Delphi survey, but infrequently mentioned by focus group participants in the prior qualitative study (29). The term ‘task-specific movement’ has been broadly defined in the literature and may include multiple types of ABT, such as walking, transfers and grasping (2, 4, 40). Although a clear definition of task-specific movement was provided in the Delphi survey (i.e. discrete, goal-directed movements of the arms or legs, such as reaching and grasping and kicking a ball), to delineate it from the other types of ABT that could also be considered task-specific movement, it was left to interpretation by participants in the focus group interviews which may account for the discrepancy. Aligning with the findings from this study, Gauthier and colleagues (41) conducted a survey on equipment use for ABT in Canada and found surface electrical stimulation devices to be the most reported piece of technology. Triangulation of the findings of this study with previous work highlights agreement from a large body of evidence on the types of ABT and technologies that should be included in future tools that are developed to track ABT.

Table 4.

 Triangulation of ABT Delphi survey, scoping review (28) and qualitative study (29).

	Delphi survey		Scoping review (28)*	Qualitative study (29)**
	Round 1 (%)	Round 2 (%)	N (%)	N (%)
Type of ABT
Muscle strengthening	91.7	–	49 (35)	39 (81.3)
Balance training	90.0	–	22 (15.7)	12 (25)
Load-bearing exercise (with crawling)	85.0 67.8 (crawling)	71.4 (crawling)	25 (17.9)	35 (72.9)
Task-specific movement	85.0	–	19 (13.6)	6 (12.5)
Transfer training	81.4	–	11 (7.9)	10 (20.8)
Ergometer training	80	–	34 (24.3)	23 (47.9)
Overground walking	79.7	–	44 (31.4)	30 (62.5)
Walking on a treadmill	73.2	–	80 (57.1 )	35 (72.9)
Wheeling in a wheelchair	66.0	60.0	0 (0)	1 (2.1)
Plyometrics	61.0	50.0	1 (0.7)	0 (0)
Cross training	58.9	60.4	4 (2.9)	9 (18.8)
Stair training	52.7	55.3	8 (5.7)	3 (6.3)
Rowing	50.0	34.0	3 (2.1)	1 (2.1)
Pool therapy	47.1	42.9	3 (2.1)	3 (6.3)
Running	44.2	34.0	0 (0)	0 (0)
Activities of daily living	–	61.2	***	1 (2.1)
Technology used with ABT
Transcutaneous NMES	83.1	–	80 (57.1)	33 (68.8)
Spinal cord stimulation	57.8	66.0	11 (7.9)	4 (8.3)
Vibration	50.9	51.0	6 (4.3)	5 (10.4)
Virtual reality	28.6	37.0	8 (5.7)	3 (6.3)

This table triangulates the findings between three separate studies: a Delphi survey, a scoping review and a qualitative study. N, number; ABT, activity-based therapy; NMES, neuromuscular electrical stimulation.

*Data from the scoping review (28) was extracted from Supplement 5 and includes the number of studies (N) from a total of 140 studies using the type of ABT or technology.

**Data from the qualitative study (29) was extracted from the transcripts and reports the number of participants (N) from a total of 48 participants who stated they engaged, practiced, or valued a type of ABT or technology.

***Activities of daily living were not identified as a type of ABT in the scoping review.

Activities of daily living (ADL), which includes feeding, dressing, grooming, toileting and ambulating (42), was suggested by a participant to include in round 2 of the Delphi e-survey, however, it did not reach consensus for inclusion in an ABT tracking tool. Although some aspects of ADL, such as ambulation or task-specific movement of the upper extremities for feeding and grooming, could be considered ABT, it would depend on the way in which these movements were trained; for example, whether the training incorporated neuromuscular activation below the injury level and repetition of movement. This suggestion highlights the challenge in understanding what constitutes as ABT and the lack of consensus in how ABT is defined, even among individuals who felt knowledgeable about ABT. As a future recommendation, clinicians and people with SCI/D may benefit from the addition of a supplementary toolkit that could accompany and support the proper use of an ABT tracking tool by providing a definition of ABT, descriptions and demonstrations of types of ABT and technology as well as appropriate use of the technology.

Since the anticipated use of an ABT tracking tool is across all settings (acute care, inpatient and outpatient rehabilitation, community clinic and home) it is important that both clinicians and people with SCI/D are able to independently use the tool. Hence, it was important to consider the feasibility in this study. Unsurprisingly, feasibility to track the different types of ABT and technologies was rated higher for clinicians than for people living with SCI/D. Higher feasibility scores for clinicians may suggest participants believe that clinicians have specialized knowledge about the ABT or technology, the parameters to track and how to progress the activity. This knowledge facilitates documentation about the activity. Participants may also think it would be easier and more accurate if a clinician recorded data during an ABT session while an individual was engaged in ABT as opposed to a person with SCI/D relying on recall to record their information after the session or activity. For the nine types of ABT and one technology that were rated important in this study, only crawling scored low for feasibility to track by people with SCI/D. Survey respondents may have considered some details of crawling, such as the pattern of interlimb coordination adopted, difficult for individuals with SCI/D to interpret and document.

The medium used to track ABT may also play a role in the feasibility of tracking for clinicians and people with SCI/D. A previous qualitative study that explored tracking ABT from the perspective of clinicians, administrators, researchers, advocates, funders, policy makers and people with SCI/D found participants preferred digital tracking tools, such as an app, especially for individuals with limited hand function (29). However, some individuals may lack access to technology or feel ill-equipped to use technology that supports an app. In addition, participants recognized that a paper-based tool was still needed as some hospital charting systems are not digital and unable to manage this type of data. Future research could consider exploring the perceptions of clinicians and people with SCI/D on the barriers and facilitators regarding the feasibility of tracking ABT to improve our understanding of this issue. A toolkit, as mentioned previously, may also increase the feasibility of tracking ABT by providing clear and detailed instructions to enhance user knowledge and training of the tool.

Limitations

Convenience sampling (43) was used in this study. Although we primarily targeted community ABT clinics and the Canadian ABT CoP to ensure we recruited individuals with expertise in ABT, we also broadened our reach to NASCIC, which has a network reach of over 400 members across Canada and the United States. As a result, our recruitment message likely reached individuals without ABT experience. Participants self-identified as having experience participating in, supervising, researching, funding or advocating for ABT, which is a limitation. Despite the broad recruitment reach, we did not achieve our target of 10 participants within each group. It is possible that there is a limited number of individuals within these groups (e.g. administrators and funders, advocates and policy experts) who have expertise in ABT, or that there is scarce membership of these groups from the organizations we recruited from. The lack of input from administrators and funders, advocates and policy experts, combined with the over-representation of individuals with SCI/D and community clinicians, may have biased the data. In addition, as this was a North American-based study, we cannot confirm generalizability of these findings to other jurisdictions. Participants who were hospital and community clinic clinicians or people with SCI/D also had to indicate their years of experience with ABT. Since these queries were self-reported we had no way of determining the accuracy of the information provided or confirming whether participants had sufficient knowledge of ABT to respond appropriately to the survey questions, which may have affected the validity of the data collected. Lastly, some participants indicated they were unfamiliar with a specific type of ABT or technology; however, only two types of ABT and two types of technology were unfamiliar to >10% of participants. Even if all participants who indicated they were unfamiliar with a type of ABT or technology rated it as important (i.e. ≥7), none of those items would have reached the threshold to be considered important to include in an ABT tracking tool. Hence, we do not think the unfamiliarity with some types of ABT and technology affected the results.

Conclusions

This study identified the types of activities and technologies to include in an ABT tracking tool. Tools that track the details of an ABT session or program may provide the information required to support decision-making at all levels of the healthcare system and may aid in the development of best practice guidelines.

Disclaimer statements

Contributors None.

Funding This research was funded by the Canadian Institutes of Health Research Catalyst Grant to K.E.M. and A.K. and a Vanier Canada Graduate Scholarship and a KITE-Toronto Rehab’s TD Graduate Scholarship for People with Disabilities to A.K. K.E.M. holds a Canada Research Chair (Tier 2) in Multi-morbidity and Complex Rehabilitation.

Conflicts of interest Authors have no conflict of interest to declare.

Supplemental data

Supplemental data for this article can be accessed on the publisher’s website.