Participants and setting

The workshop took place at the University Hospital Heidelberg, Germany. GPs were invited to participate in one of two ways: 1) via an existing GP research network including 86 practices, or 2) via a program for Continuing Medical Education for GP practice teams. All GPs gave their written informed consent before participation. Ethical approval was granted by the Ethics Committee of the University of Heidelberg (registration ID: S-414/2013).

Procedures

The participants were evenly divided between the two focus groups (9 and 10 participants) which were led by one of two trained facilitators (SK, JSt) using a semi-standardized interview guideline (Table 1). Initially the facilitators invited participants to share their ideas on the presented stratified care approach. Prior to the focus groups the semi-standardized interview questions were tested with a GP who was not a workshop participant [26]. Each facilitator was supported by a research assistant, who took care of organizational aspects of the meeting and took notes during the discussions.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Interview guideline.

https://doi.org/10.1371/journal.pone.0136119.t001

The discussions were audiotaped, completely transcribed and coded using a content analysis approach [27]. Transcription was carried out verbatim following predefined department standards by the two research assistants who supported the facilitators and were present during the focus group discussions. The transcripts were re-checked by the two facilitators.

Analysis began by the facilitators becoming immersed in the data through detailed multiple readings of the transcripts (SK, JSt). The main themes and subthemes which emerged from the transcripts were developed by one of the facilitators (SK), organized to align with the order of the interview guideline (SK), and then interrogated by the other (JSt). Through this iterative process the themes were refined collaboratively (SK, JSt). Quotes supporting each identified theme were chosen. As software R-QDA was used for the analysis (2014; http://rqda.r-forge.r-project.org/).

Applying the screening tool

Aspects which were discussed by the GPs in respect to implementing the SBST were: the most appropriate administering health professional, if the patient was already well known to the practice, the enquiry of the reason for the consultation, tool scoring, tool format, and its anticipated clinical impact.

In respect to organisational aspects GPs discussed if they themselves or their medical assistants would be most appropriate to administer the screening tool. Some GPs favoured delegating the task, whilst others were against this suggestion.

During the discussion it became clear that individual GP practices are organised very differently in respect to how much information GPs know beforehand about the main reason for the consultation. In some practices medical assistants routinely ask patients about the reason for their visit, whilst for others this was not considered appropriate.

The issue about patients who were known to the practice as having recurrent LBP problems also led to differences in opinion among different GPs regarding how best to administer the tool. Some GPs felt it was inappropriate to ask patients to complete the SBST during their first consultation for LBP, whereas others felt this would be a possibility.

Scoring the tool was discussed in context with time resources and the tool’s format. Individual participants reasoned that using the tool would require additional resources, as calculating the STarT-score and determining the patient’s risk-group might take time. Others opposed, suggesting that the instrument with only nine items was sufficiently short to be practical for use within short consultations. In order to improve implementation, some GPs suggested embedding the tool within the GP computer or having it available on a tablet or mobile device. If this was possible, then calculating the score and risk-group could be automatic and so save the GPs time.

The impact the tool might have on GP decision making was an important discussion point. Some physicians thought the tool was likely to be a useful way of assisting their treatment plans for an individual. GPs tended to view the tool as a potential resource that might help to speed up their clinical decision making process, and that it could support their explanations to patients about their likely prognosis and how best to manage their condition. Other GPs in contrast did not expect it to be of much assistance to their decision making and some felt it might have a negative influence on the consultation because it was impersonal and involved disturbing non-verbal communication methods between the patient and their doctor.

Psychologically informed therapy

In respect to implementing psychologically informed physiotherapy, the majority of GP comments were focussed on the physiotherapists as the professionals to deliver this care. Implementation of an effective treatment approach for complex high-risk patients was anticipated as beneficial. The GPs wished for support for treatment of these patients. In this context, the skills and qualifications of physiotherapists were discussed intensively, including their abilities to identify patients that might need formal psychological therapy with a specialist clinician. In addition, the need for some adaptation of physiotherapists’ current working conditions was discussed in order to ensure that more time was available for these patients.

A further controversial discussion point was the issue of implementing physiotherapists who deliver psychologically informed physiotherapy, especially since the participants agreed that treating high risk, complex patients could be very clinically challenging. Therefore, several GPs expressed the view that training physiotherapists to better manage psychosocial aspects among high-risk patients might be particularly beneficial. One GP reported in fairly isolated cases to already asking the physiotherapist to specifically consider psychosocial obstacles.

Overall, GP participants were open minded about the stratified care approach to LBP, but recommended that implementation should be coincided with a robust scientific evaluation of the impact of this new approach in the German system.

Another theme that was comprehensively discussed was whether physiotherapists needed formal educational skills and qualifications or simply could be expected to learn and gain experience by doing. GPs felt that physiotherapists had appropriate skills to manage complex biomedical needs and expected that psychological competencies were something which came to some extend with the right experience and so could be within their scope of practice. Potential limitations of even specifically trained physiotherapists, to address the needs of some patients with severe complex psychological problems were discussed (i.e. needing treatment by a psychologist or in a specialized pain clinic).

Uncertainties were however expressed, about the dimension to which psychosocial competencies were currently included within physiotherapy training programmes, and they also were uncertain about how knowledgeable physiotherapists were about evidence-based practice.

In respect to physiotherapists’ working arrangements and appointment times, there was broad recognition by the GPs for the need for some system modifications. The participants reflected on the current practice of communication between GPs and physiotherapist which, in nearly all cases is solely based on the referral form. For effectively implementing the STarT-approach, the flow of information would have to be improved. Additionally, the remuneration for the treatment of high-risk complex patients was a point of discussion. This was mentioned as a task of high value which should be acknowledged.

Management of low risk patients

The following themes emerged from the discussion about implementing the matched treatment for low-risk patients; the most appropriate professional to advise patients, the length of consultations, media that should be used to support information given, and information about local exercise venues. The alternative health professionals that were highlighted and considered in this discussion were GPs, physiotherapists, and medical assistants. In Germany, it is current practice that LBP patients receive advice and education from their GP, but some GPs were open to this task being shared with physiotherapists, on the condition that this type of appointment would be available within a very short waiting time.

In order for low-risk patients to receive their advice and support for self-management the issue of the spatial and organisational distance between physiotherapists and GPs was seen as a barrier, especially due to the need for patients to receive immediate advice. The medical assistant in the GP practice was seen as primarily an administrative role which could be useful.

Concerning the use of advice booklets, GPs had mixed experiences and the use of an informational video to help educate patients was rare. Their expectations about the success of such an approach varied. They also highlighted that providing patients with an advice/educational video would probably require additional structures which were currently not available within their practices, although, as a possible solution some suggested to put a computer in the waiting room. Alternatively different strategies like online videos or DVDs for patients to view at home were proposed.

The duration of the initial 30-minute assessment and advice appointment for patients in the STarT approach was identified as a potential barrier for GPs due to perceived time constraints under the current working conditions even though it did not seem impossible, especially if supported by the medical assistant.

Some GPs felt comfortable that they knew about local opportunities for patients to exercise, although others did not feel they had a comprehensive overview of such venues and were open to having support to collate this information. It was felt that it was important for GPs to be able to recommend local facilities if patients were to have confidence in them, although most GPs did not feel prepared to provide such an evaluation for their patients. Concerns were put forward regarding potential costs, although health insurance financed options were discussed as an alternative.

Inductively identified themes

In the analysis two additional themes with relevance to the implementation of the STarT Back approach in Germany were identified inductively: GPs views about physiotherapy and GPs perceptions about different types of patients.

For physiotherapy it was discussed that there is a lack of information about practice: “Well, that would be desirable, if you knew what the physiotherapist really does in this case” (Gr2_GP1). In Germany, due to the regulated practice of physiotherapists, a referral from a physician is necessary. The regulation process is complex and many physicians in ambulatory care fear a financial penalty as part of their liability as a result of over-referring. During the discussions it was stated, that physiotherapists should share financial responsibility: “Yes, the [physiotherapists] must sit in the budget-boat.” (Gr2_ HA8).

The GPs also felt that there was an important patient-type that was not specifically identified or given appropriate consideration in the STarT Back study. With “Rentenbegehren”(Gr1_GP1) patients are meant who have a strong desire to retire early and to get a pension due to their low back pain (Rente = pension, Begehren = desire). This specific German term was repeatedly mentioned.

Implications for practice organisation

Overall, participants agreed that the SBST’s brevity and ease of scoring make it feasible and practical for use in general practice [25]. However, they suggested that an electronic version of the tool with automated scoring would be a further facilitator for its implementation. This type of advancement has been undertaken in the UK [31] and is possible within German primary care. Moreover, a free app is available for mobile phones and could be translated into German [32].

Time constraints appear to be a perceived barrier to GPs feeling able to advise LBP patients during their consultation. However, the alternative of implementing physiotherapists to perform this task was considered to be impractical due to the geographical and organisational separation between physicians and physiotherapists. One proposed solution is for GPs to provide patients with some brief advice that is then followed up through a further session with a physiotherapist. This model would be similar to the procedure described by Sowden et al. for the IMPaCT Back implementation study. In this project patients who were classified as medium or high risk by the GP were sent to a physiotherapist, although those who improved and were classified as low risk after re-assessment at the physiotherapy practice, were only provided with a one-off assessment and advice session [33].

One alternative suggestion to increase the time available to give advice to patients within the GP practice was to involve GP medical assistants. In Germany such medical assistants are mainly administrative roles, but there are some attempts to broaden their competencies [20]. For example, it was felt appropriate that medical assistants could provide information about local exercise venues, written education material and to show patients an educational video. A possible strategy to overcome the limitations of physical space within the practice to deliver video material was suggestions to offer web-based material that patients could watch at home or on their mobile device. One notable finding was that not all GPs were aware of available informational material for LBP patients, and so ensuring such information was easily accessible would be an important component for a stratified care implementation program.

Training for GPs

The positive perceptions of the GPs concerning the STarT Back stratified care approach is a potential enabling factor for future GP training and implementation. GPs should be informed about the evidence on biopsychosocial treatment effects e.g. the impact on clinical outcomes [34, 35], but also need practical details on how this approach might influence patient care and clinical pathways both for themselves, and for related physiotherapy services and their ability to treat particularly the high risk complex patients group [10, 36, 15]. This also applies for video-based information [37, 10]. An important patient-group that might need more consideration in the STarT Back approach are those with a strong desire to retire early [38]. A specific training to help to manage these patients could be of benefit.

One important finding was that GPs could have difficulties with the idea of changing from existing stepped care models (waiting to see how patients symptoms develop) to the stratified care model (acting decisively based on patient’s initial SBST prognostic profile). As a result, this aspect should be addressed intensively within a training program.

Another change for GPs was the use of prognostic information during a LBP consultation, and it was apparent that here a training need also existed as GPs had the perception that use of such a tool during the GP consultation would make the patient interaction feel more impersonal. Accordingly, it would be relevant to explain that the items contained in the SBST may help patients to relate more honestly about the impact their pain is having on them, particularly in respect to their mood.

To address the concerns GPs raised about financial aspects it could be important to investigate the cost effectiveness of the STarT Back stratified care approach during its implementation into the German healthcare system. The UK implementation has demonstrated the approach to be broadly cost-neutral with extra resources used for relatively small group high-risk patients being made available from savings gained in the treatment of the large group of low-risk patients [12, 10].

Strength and weaknesses

The varying topics discussed in the focus groups highlighted that the GPs were able to grasp the STarT Back approach despite a relatively short introduction. Unclear points were readily clarified during discussions. Accordingly, the transcripts could be evaluated effectively according to the initial aims.

The results convey the impression that with regard to the described aim, a satisfying saturation was reached. For qualitative group approaches, Carlsen and Glenton identified recommendations ranging from two to five groups adding that the necessary number depends on the research question and the composition of the groups [39]. Recommendations concerning the number of participants range between 6 and 12 [40]. We are of the opinion that with two groups of altogether 19 participants, we received a saturation fitting our research purpose, which was satisfactory although the duration of the discussions was short. Reasons for reaching our goals with this compact design relate to preparing a clear introduction and discussion guideline.

Research agenda

Research on the psychometric properties of the German version of the SBST ran in parallel to the described work. Foster et al. propagate randomized trials as a preferred design for testing stratified care [9]. Correspondingly, the next step should be a pilot study examining the adapted approach and also providing information on the developed implementation strategy.

Patient expectations are believed to play an important role on treatment outcome [41]. The physicians participating in our study mentioned several ideas on how patients would react to different aspects of the STarT-approach if implemented in Germany. Those should be discussed with patients and additional aspects should be gathered.

Similarly, this applies to physiotherapists. Questions need to be discussed, e.g. their attitudes towards evidence-based practice, their qualifications and ability to address psychosocial factors or concerning the one-off clinic appointment, as described for the STarT-approach [10].