A Core Outcome Set for Randomized Controlled Trials of Physical Activity Interventions: Development and Challenges | BMC Public Health

The results of the literature review (first stage) and the Delphi surveys (second and third stage) are presented below, along with the final core set of results.

First step – literature review to identify physical activity outcome domains

A review of the literature to identify potential outcome domains of physical activity, followed by discussion among the authors, led to the final classification of the following 13 outcome domains: adverse events; biophysical health; profitability; health-related quality of life; device-based physical activity level; self-reported level of physical activity; motivation; other health behaviors; fitness; physical function; satisfaction with the intervention; sedentary behavior; and self-efficacy. Of particular note, and given the debates surrounding the measurement of physical activity level, it was decided to create two outcome domains for “physical activity level”, one focusing on data collected through self-report measures and the other on device-based capital measures. Each domain was defined by the authors to support the consensus-building process (see Table 2 for definitions).

Table 2 Classification of physical activity outcome domains

Step two – determine result areas

Forty-one people (36 professionals and five laypersons) were asked to participate in a Delphi survey to give their opinions on the outcome domains to be included in the core outcome set. Twenty people (16 professionals and four laypersons) agreed to participate and completed the first round. All participants came from high-income countries. The majority of professional participants were academic researchers interested in physical activity, outcome measurement, or both. Two academics were current or past healthcare professionals, and one participant was a healthcare professional only. In terms of career stage, academics were at various stages, but there was a bias in favor of those in higher positions. Of the four lay participants, three were female and one male, and their ages ranged from “18-29” to “60-69”.

A summary of participants’ ratings of the importance of the outcome domains can be found in Table 3. At least 70% of participants rated ‘level of physical activity (device-based)’ and ‘quality of life related to health” as areas of critical importance. importance, meeting the criteria for inclusion in the core result set. No consensus was reached for the remaining 11 outcome areas and therefore a second round was conducted. No new outcome domains were added for consideration in the second round.

Table 3 Second stage Delphi survey results, cycles 1 and 2

Thirty-six people (32 professionals and four laypersons) were invited to participate in a second round (seven participants had opted out of the research and were not invited to participate in the second round, and two newly identified participants were invited (academic researchers interested in physical activity)). The other 11 outcome areas were presented to participants to assess their importance. Twenty-one people took part (17 professionals and 4 lay people), 17 of whom had taken part in the first round. No consensus was reached for any of the 11 outcome domains (see Table 3), however, “satisfaction with the intervention” was close to reaching consensus, with 66.7% of participants rating it as extremely important.

Third step – determining the measurement instruments for the agreed result areas

Thirty-six people were invited to participate in a second Delphi survey (the same participants were invited to participate as those who were invited in the second round of the first Delphi survey). Of the 36 participants invited to participate, 19 (15 professionals and four laypersons) responded to the second Delphi survey to give their opinion on the best measurement tools and approaches for the identified domains. The distribution of professional roles was the same as second step. A total of 15 participants (11 professionals and four laypersons) took part in all stages of the two Delphi surveys. As “satisfaction with the intervention” was close to the consensus in second step, further opinions on the inclusion of this domain in the core result set were solicited as part of this survey. The results for each result area are presented in turn.

Device-based physical activity level

Overall, respondents rated the accelerometer (average score = 3.89, score range 1-5) as the best device to include in the core result set for measuring physical activity level , followed by a multi-sensor device to include an accelerometer (average rating=3.32), an accelerometer plus heart rate monitor (average rating=3.11), and a multi-sensor device without an accelerometer (average rating= 2.11). Nine respondents provided reasons for their ratings. Several respondents highlighted the importance of device acceptability to trial participants, with compliance likely to be increased through the use of only accelerometers, particularly those that tolerate immersion in water. water. Several heart rate sensors and monitors were found to be less acceptable to participants, especially when worn for long periods of time. Moreover, one participant considered that the added value of a multi-sensor device was limited, while another warned that their use on a larger scale could be too expensive.

There was support (68.4%, not= 13) for raw accelerometer data to be collected and made publicly available, however, this did not meet the minimum level of agreement for consensus. There was limited agreement (42.1%, not= 8) that energy expenditure related to physical activity (PAEE) should be considered as the key concept to be measured by devices to assess the level of physical activity.

Health-related quality of life

There was consensus (73.7%, not= 14) that the EQ-5D should be included in the core set of outcomes for assessing health-related quality of life, with only five respondents (26.3%) disagreeing and suggesting alternative measures, such as than the SF-36 [17]SF-12 [18]or the Warwick-Edinburgh Mental Well-Being Scale (WEMWBS) [19].

Satisfied with the intervention

While many (68.4%, not= 13) considered “satisfaction with the intervention” to be an essential domain to include in the core set of results, the minimum level of agreement for consensus was not reached. Among those who wanted to include the domain, there was agreement (84.6%, not= 11) that there are currently no appropriate measures available to assess the domain.

The final baseline result set is shown in Table 4.

Table 4 The final core result set for physical activity trials

About Mark A. Tomlin

Check Also

Roadmap for a literature review

By Prof. RA Seetha Bandara, Board Member of the Sri …