Methods
Detailed descriptions of the methods underpinning this qualitative study were published previously 18. These methods are summarised below.
Reporting framework: The Standards for Reporting Qualitative Research (SRQR) were followed to ensure a rigorous reporting process of the study findings 19. The SRQR checklist is provided in Additional file 1.
Ethics: Ethics approval was received from the Health Research Ethics Committee of Stellenbosch University (S17/05/100). Written informed consent was obtained from the participants via an online form. All research team members were familiar with the South African context in which this study took place and were experienced in the conducting and reporting of qualitative research.
Sample: The participants were South African PTs from private, public or educator sectors (considered for analysis purposes as three independent clusters). Participants needed to be registered with the Health Professions Council of South Africa (HPCSA), and be practicing PT for at least 25 hours per week in any field of PT. The email list obtained from the HPCSA, was randomized, and groups of 30 participants were contacted at a time.
Data collection and saturation: Semi-structured individual interviews were conducted, audiotaped, and independently transcribed18. The question relevant to this paper was: “When considering your workload, is it easy or difficult to integrate CPG into your clinical load?” This question explored whether the participants perceived CPG utilisation in practice as adding to or decreasing their clinical workload. In the previously published paper18, the main theme that emerged from this questions was “lack of time”, with the subthemes being: few staff; high workload; access to CPGs; EBP as priority in clinical practice; “time is money” attitude; and knowledge on the use of CPGs. Interviews continued with PTs in each cluster until no new information was obtained in consecutive interviews.
Data analysis: For the purposes of this paper, an inductive thematic content analysis approach was taken where the transcript content was analysed by hand, and the themes emerged from the analysis of the relevant question 20. Data analysis followed an iterative process of data immersion and familiarization; theme identification; creating a codebook, data coding and categorizing; data mapping and interpretation, and then checking the findings against the original transcripts 20.
Establishing credibility: Member checking of transcriptions were performed by 28 of the 31 participants (three participants did not respond to requests to do this) 21.
Researcher bias : The interviewer was a practicing PT clinician and educator. She could relate to the participants and analyse the data within her understanding of South African PT practice. She brought a recognized bias to the process of analysis of the interviews in that she is enthusiastic about providing best-evidence PT care, and teaching PTs about CPGs.
Determining SoC classification: There is currently no way of establishing participant’s SoC for evidence uptake, thus the need arose to categorise in the following manner. The best-fit SoC for each interviewee was determined using an amalgamation of interview findings and socio-demographic data, which was layered onto the Transtheoretical model of the SoC 17.
The following steps were followed (See Figure 2):
  1. JS assessed whether the participants could provide a definition of CPGs that aligns with recognised and published definitions and whether they believed that CPGs were of benefit to the PT profession. This was achieved by comparing the definition of CPGs according to Treweek et al. 7 with the definition that each participant provided. The participants’ answers were divided into Clear or Unclear understanding of CPGs;
  2. Participants’ views (positive or negative) were explored to determine whether the use of CPGs added to, or reduced, their workload, and whether PTs perceived CPGs in their clinical practice as beneficial (or not), and overall to the PT profession.
  3. A preliminary allocation into a SoC category was made for each participant, by comparing participants’ understanding of CPGs, and their views on benefits, with the definition for each level of the SoC model.
  4. Exemplar quotations from the full interview transcript on understanding, uptake and benefits of CPGs were identified as an independent source of evidence for the preliminary categorisation;
  5. Differences in socio-demographic data (sectors of employment, years since graduation, and gender) were then determined for each SoC.
  6. The authors discussed and agreed on the final SoC categorisation for each participant.
>>Figure 2: Process of Stages of Change allocation about here>>