Chronic Low Back Pain and Yellow Flags, Why, How and What Physiotherapists Do: Part 2 Discussion


The literature review overviews the various different methods currently used in practice to give an overview of the background to which these decisions are being made.

Why these physiotherapists practice

The literature review showed the breadth of approaches available to clinicians and informed the interview process. This research sheds some light on how and why clinicians choose which of these interventions to use. The central theoretical code identified was the physiotherapists looked to do what was best for the patient from a variety of motivations. This is similar to previous grounded theory work, which found that all physiotherapist expressed a commitment to caring (Resnik and Jensen 2003). One of the key open codes supporting this concept was that of empathy. Interestingly a study of 1111 students found that physiotherapists in training had the highest levels of empathy compared with other allied health professions (Williams et al. 2014). This suggests empathy could have more influence over behaviour in physiotherapists than other professions. Empathy is known to increase helping behaviours and this has both strengths and weaknesses. The pioneering work of Goleman (1996) highlighted the many benefits of empathy such as “success” in multiple professions and more specifically in a healthcare setting treatment outcomes, satisfaction and job performance. Back pain patients report more satisfaction when treated by clinicians with more empathy (May 2001). However this increased desire to help others can lead to more empathetic individuals feeling more pressure and helplessness, leading to negative health outcomes. This is important for physiotherapists to be aware of as it can lead to being taken advantage of, helping colleagues and patients more than is necessary and feeling greater responsibility than those with less empathy. Thus this trait can lead to significant difficulties that can transcend the professional life and affect all areas of a physiotherapist’s life.

Human beings typically achieve more fulfilment in work than in our social activities (Csikszentmihalyi 1991). This comes as a surprise to many. Nonetheless in this research it seemed that a sense of personal pride in their work was central to reasons for why physiotherapist wanted to help patients. The finding of professional competitiveness was very surprising. KM strove to come from a clear mind as suggested by Descartes (1637) that would be ideal for an inductive piece of research. Despite this an affinity for the physiotherapists who gave up their time to help with the study developed. This made the emergence of the code all the more troubling. Initially there was an eagerness to dismiss this due to this affinity, however it’s repeated emergence meant it needed to be analysed and considered as shown in KM’s theoretical memo.

Memo 1st October 2015

The disdain for other professionals and repeated allusions to mistakes made by other professionals appears to be part of giving experienced physiotherapists their self-worth in practice. It may be that this comes from an innate competitive drive that shows its self in these statements. Further this may help push therapists to improve and enjoy their practice further.

This finding may be more particular to this group of physiotherapists. Perhaps due to their experience they were all confident. For example all felt comfortable with the patient having yellow flags and described their strategies for managing this. In contrast previous work has suggested physiotherapists may not feel that comfortable or well trained in managing the psychological aspects of patient care (Hemmings and Povey 2002).

This research cannot give direct recommendations for how to assess and potentially change a physiotherapist’s motivation to practice. However, perhaps it can be of value by encouraging the interested reader to reflect on their own motivations to practice and consider how this may affect decisions in practice on a day-to-day basis.

What’s technically best for the patient

The results of this study show that “how” physiotherapists try to help patients is by finding what is technically best for them. The model of how and what is interesting when viewed in the context of previous grounded theory work in to physiotherapists’ clinical reasoning. The clinical reasoning of the “what” would fit with Jensen et al.’s (2000) diagnostic reasoning. Jensen et al. describe a narrative reasoning, reasoning and procedure element within diagnosis, which would equate to the understanding, whilst their interactive and collaborative reasoning would equate to the rapport element highlighted in developing adherence. Jensen et al.’s predictive, ethical and teaching reasoning components are not included within the theory here, but are more the underlying processes of deciding how to empower the patient, based on the other elements. Jensen et al. split these elements across diagnosis and treatment; whilst the data gathered here suggested the clinical audit process clinicians used makes such distinctions difficult. Thus the integrative diagram in the previous article can be updated as seen below.

reasoning Jensen

Figure 1 Update cycle of Iterative cycle to help patients do what is best for them

The finding of how clinicians came to their clinical reasoning decisions may be a disappointment to policy makers and proponents of the contemporary descriptions of evidence based medicine. The term evidence based medicine is evolving and whilst all physiotherapists alluded to it or mentioned it directly there was broad variation in what they appeared to mean with the term. The physiotherapists interviewed used a combination of trial based evidence, experience and expert opinion as well as the patient’s wishes. This suggests these physiotherapist are generally practicing somewhere between traditional expert based medicine (Smith and Rennie 2014) and Sackett et al.’s (1996) definition of evidence-based medicine, which discussed the integration of the best evidence with what the patient presents and the clinician’s experience. Current definitions have progressed to focus on using the best available evidence to make decisions (Gray 2014), with an emphasis on a hierarchy of evidence prioritising meta-analyses, systematic reviews and higher quality trials over expert opinion and basic science (Moore and Jull 2006). Greenlagh et al. (2014) argue that the movement has had significant benefit but that there is too much evidence for clinicians to consume, guidelines are too large and rules have become inflexible and mechanical. Greenlagh et al. (2014) argue that clinicians should return to “real evidence-based medicine” using expert analysis of evidence to deliver patient centred care. They argue this should be done using higher levels of analysis and human intuition as described by the Dreyfus brothers (1987). Nonetheless physiotherapists struggle to apply trial based evidence to practice (Jette et al. 2003). In a sample of 488 American Physical Therapy Association members lack of time was seen as the major limiting factor. Similarly, Hannes et al. (2009) highlighted several barriers to the implementation of evidence-based practice by Belgian physiotherapist. They conducted a grounded theory approach analysis of 43 purposively sampled physiotherapists with varied interests and expertise in evidence based medicine. The problem tree they developed highlighted the key obstacles to the implementation of evidence-based practice as a lack of autonomy and authority to decide treatments. Interestingly questions were raised of the quality, applicability and accessibility of evidence. A lack of motivation from physiotherapists to use evidence-based practice was alluded to but not expanded on. However, this returns us to a fundamental human conundrum. Behavioural economics has shown us human intuition is not reliable and following simple algorithms is often more effective (Tversky and Khaneman 1981). Interestingly we prefer human error to computer error and would usually prefer to trust a human rather than an algorithm (Kahneman 2011). Nonetheless at this time reducing the treatment of CLBP patients with yellow flags to an algorithm is out of reach due to it’s complexity but it is not impossible that in the fullness of time we may be able to develop more useful “mechanical” guidelines.

At the superficial level it is interesting to compare “what” the different physiotherapists wanted to do with the patient in the vignette. The background and literature review highlighted the varying different approaches and emphasised their differences across the spectrum of physiotherapy. The interviews found a similar breadth of approaches likely extenuated by the purposive sampling. This was evident in sub-grouping where physiotherapists would use overt and covert sub-grouping. Sometimes using specific methods and other times interviewees were able to specify specific conditions under which they would give one style of intervention or use a specific questionnaire. Given the effectiveness shown by the classification-based approaches in the literature review it is not surprising most interviewees were using some form of classification. Clinicians often prefer classification-based or more individualised treatment-based approaches. However, Karayannis et al. (2015) showed that across the most prominent forms of classification there is both overlap and discord. Thus whilst there is utility in sub classification there are certainly discrepancies in application. It may be that this is because they allow more autonomy. The social sciences have shown as autonomy in our work decreases so does our motivation and performance (Cerasoli et al. 2014). Thus it makes sense that these approaches are more appealing in comparison to approaches that simplify treatment and perhaps suggest there is less skill involved in the work of physiotherapists. This area has become contentious in the physiotherapy literature as governing bodies push for more physiotherapists to follow guidelines and evidence begins to suggest all exercise has a fairly similar response (Searle et al. 2015). Thus clinicians desire for autonomy which may be central to the desire to continue to practice, appears at loggerheads with policy makers who may acknowledge the need for individual based treatment but focus guidelines and policy around delivering one size fits all programmes.

Whilst there was a dearth of literature that met the selection criteria considering the significance of the problem, those that did highlighted the benefits of passive, active; more comprehensive and simple education interventions. The physiotherapists interviewed all mentioned a process KM used the code “graded exposure” for. This was a process of getting the patient to do what they could from both activities of daily living perspective and an exercise perspective and then gradually increase it. Interviewees used their own paradigms to explain how this approach was having an effect; often using biomechanics based explanations, which may be at least partially accurate. However, it appears this approach can definitely work to reduce pain related fear, catastrophising and pain related disability (Vlaeyen et al. 2001). Macedo et al.’s (2010) systematic review suggested graded exposure was effective for reducing back pain of longer than 6-weeks duration in the short term compared to minimal interventions but no different to other exercise interventions. This is not a fair comparison as many of the different approaches coded as “graded exposure” would fit within the other exercise approaches compared in Macedo et al.’s review. This is a common problem in physiotherapy where semantics and how techniques are labelled leads to disagreements where there is much common ground.

The literature appeared to show that while passive interventions can improve pain, more whole body active approaches such as whole body weight training or walking may be necessary to positively impact the psychosocial aspects. Interviewees appeared conscious of this and almost conflicted about the role of passive approaches, in particular manual therapy. As demonstrated in the memo below:

Memo 27th September 2015

Physiotherapists appear conflicted over the role of manual therapy. They appear to want to use it and feel it will help, but are fearful patients will become dependent on it, and it will not lead to long-term recovery. All interviewees seem to feel that there may be specific manual restrictions that can significantly benefit from manual therapy and improve symptoms.

This is an interesting concept and when examined the potential for the development of dependency does not appear to have been well examined. Indeed it is not discussed in guidelines for the management of CLBP (Koes et al. 2010). Indeed for acute symptoms manipulation is actively encouraged (Koes et al. 2010), although it’s benefit for CLBP is at best no greater than other interventions and may not be superior to sham (Rubenstein et al. 2011). Furthermore manual therapy has been found to outperform exercise at one year follow up in patients with greater than 8-weeks of low back pain in a 49 patient randomised controlled trial (Aure et al. 2003). Culture was mentioned as one factor to sub-classify patients to receive manual therapy. The implication from the interviewees was that patients from certain culture expect manual therapy and respond favourably at least in part due to their expectations. Interestingly Bishop et al. (2011) investigated the role of patient expectations in manual therapy for CLBP. Their analysis suggested that patients expected active interventions and manual therapy to help their pain and disability more than medication, electrotherapy or rest. However, they found that if patient’s met the clinical prediction rule for a positive prognosis for a manipulative technique this superseded the patient’s beliefs. Thus whilst the use of manual therapy is a broad area the fears of the physiotherapists interviewed do not appear fully warranted.

The literature found good strength evidence demonstrating the benefits of comprehensive CBT and functional restoration programmes. However, whilst all interviewees indirectly mentioned functional restoration, they directly mentioned pain education and specifically “explain pain”. Approaches involving pain neurophysiology education and pain acceptance were both effective compared with more traditional back school and pain avoidance approaches respectively in the literature review. Physiotherapists interviewed described using a combination of these approaches in practice. Neuroscience education for musculoskeletal pain, disability, anxiety and stress has been found to be effective in a systematic review, which retrieved eight papers of good quality (Louw et al. 2011). Importantly whilst not all the papers reviewed were on CLBP some looked at longer term follow ups than the very short terms considered by the two studies included in the literature review. Thus demonstrating longer-term benefits to neuroscience education.

The code “clinical audit” was used to describe a process of assessment and re-assessment to see if a patient’s symptoms could be modified during the session. Liebenson (2010) describes this process as using active techniques to carry out the assessment, whilst some of the interviews used passive techniques, such as the general nociceptive screen outlined by one interviewee. This process coded as the clinical audit, is an interesting example of where a technique is used widely in practice but has not been studied extensively. This is likely due to the flexibility of the technique as any movement that reproduces the mechanical sensitivity can be used for the assessment and any technique that may reduce the mechanical sensitivity can be used for treatment. Thus clinicians are attracted to the flexibility and researchers and policy makers are afraid of the lack of reproducibility. Furthermore, the analysis showed that physiotherapists used this modification of symptoms as a way to get the patient to understand their symptoms could be changed and to see the benefit in the treatment used. In particular they suggested using this to get patient “buy-in”.

How to get the patient to do what’s technically best for them

The second aspect of “how” identified was how physiotherapists get patients to do what is technically best for them. The individual elements of the iterative flow identified of understanding, rapport, adapting, and empowering the patient have received some investigation in the literature but overall this half of the “how” receives far less attention. For example the clinical guidelines for the management of CLBP focus on the “what” and not how to get patients to do this (Koes et al. 2010). McCrum et al. (2015) highlight the importance of understanding CLBP patient’s perspectives in helping physiotherapists adapt their communication to gain compliance and empower patients to engage in self-management. A Cochrane systematic review of 42 studies looking at improving adherence to exercise in patients with musculoskeletal pain indicated that these interventions had moderate sized effects (Jordan et al 2010). One of the key criticisms of the studies was their lack of a relevant behaviour change theory that might explain the underlying processes. In fact only two were based on a relevant behaviour change theory. Therefore as Lonsdale et al. (2010) state the most effective methods for behaviour change and how they work remains to be ascertained. Work in this area is currently developing with pioneering work looking at the use of a theory based on a self-determination intervention to improve exercise adherence in CLBP currently underway (Lonsdale et al. 2012).

In the realm of adapting the empathy identified as a frequent trait in these physiotherapists and in student physiotherapists by Williams et al. (2014) may help in improving adaptability. Empathy is central to Goleman’s (1996, Goleman 1998a) concept of emotional intelligence (EI) and is theorised to play a central role in all communication. Potentially most important is managing feelings and reactions to others’ communication. This allows an individual to look for the real meaning, which is frequently not what is initially perceived. Individuals that are empathetic are attentive to emotional cues and listen well (Goleman 2000). Further they show sensitivity and understand other’s perspectives. This allows them to communicate in a way, which is effective for each individual and the group as a whole. The key to empathy is actively listening to effective communication. Goleman (1998b) states that the mark of having truly heard someone else is to respond appropriately, even if that means making some change in what you do.

Nonetheless until further high quality research is conducted this area of practice will continue to be led by clinicians working in practice. Reflecting on the model outlined here may prove useful for novice clinicians and to hep stimulate and challenge more experienced clinicians to develop improve their own ability to help patients do what is technically best for themselves.


This research has several limitations. Most notably that the research was conducted and analysed by a novice researcher. Six interviews were conducted and whilst theoretical saturation was achieved these findings are not generalisable. KM’s lack of experience relative to the more experienced clinicians may have limited the quality of interviews, meaning they were less challenging than they could be. This may have encouraged interviewees to make more comments about other healthcare professionals and increased the likelihood of the findings around egotistic drives. Whilst not invalidating this finding it may have come across differently with a more experienced clinician interviewing. Further the interviewees were all experienced clinicians and thus the results are not applicable to inexperienced clinicians, although they may be useful to stimulate reflection and generate ideas in novice therapists. KM struggled with the dichotomy of being objective in collecting data whilst getting close enough to obtain rich data. The use of a vignette means the “what” the findings are not generalisable around what to do with a patient cannot be generalised and more suitable as idea generators for clinicians and to help challenge current practice and physiotherapists own motivations. Furthermore, the other findings are likely influenced both by the context of the vignette and by KM’s involvement in the interview and analysis process. The interviews focused on the “what” to analyse the why and how, however this meant there was no questioning of how clinicians paradigms were formed. It would have been very informative to probe this area, by asking their own views and getting an overview of their formative physiotherapy experiences. The study relied solely on interviews where Glaser  (1992) stated that to have a proper grounded theory required observation. The implication of this would be that this research is focused more on the lived experiences of the participants rather than the social processes (McCann and Clark 2004). However, the by interviewing participants about the “what” and analysing this to uncover the social processes, the “why” and “how” this potential criticism is less valid.

Implications for practice

Given the limitations and methodology of the study it is hard to make any firm recommendations for practice. Nonetheless this research should provide a useful insight in to a framework of what experienced physiotherapists do to manage a typical patient with CLBP and yellow flags. The research may potentially be of more value by aiding clinicians in considering their own “why”, stimulating thought and reflection on their own motivations to practice. Furthermore, the two step “how” process of working out what is technically best for the patient and then how to get the patient to do this may help stimulate physiotherapists to consider their own processes. Additionally, the iterative flow highlighted in this research may provide a framework for therapists to consider how they get patients to do what is technically best for them and whether there are ways they could be better at doing this.


The importance of the researcher in a qualitative study such as this in co-creating the data means the inexperience of KM is a key inherent limitation to this research. The analysis produced the grounded theory; physiotherapists seek to help CLBP patients with yellow flags for altruistic and egotistic reasons, by establishing what they feel is technically best for the patient and working out how to get the patient to do that. What they do with patients relies on “Sackettian” reasoning, paradigm-based approaches, sub-classification, clinical auditing, graded exposure, education, and an uneasy relationships with manual therapy. This theory along with the framework physiotherapists used to decide what do with patients and the updated iterative flow used to help patients do what is best will provide points for reflection to help develop the practice of physiotherapists. This study would benefit from being repeated by another more experienced researcher and further understanding may be gained by repeating the research in novice clinicians. These findings will also be of use to policy makers as it may help them understand how clinicians are choosing treatment approaches and help the target guidelines more effectively. In particular it is hoped it will draw some attention to the need to focus on compliance and how physiotherapists achieve this. It is clearly an important, under researched area and is ignored within guidelines.

Trackback from your site.