Twenty clinicians were interviewed, with interviews lasting 23–52 min. The distribution of participants’ roles are summarised in Table 1. Most interviewees (85%) worked in both in-hours and out-of-hours general practice at the time of interview, and nearly all had previous experience of in-hours general practice (90%). Interviews were undertaken face-toface at the University of Bristol (n = 3), the clinicians’ workplace (n = 3), or by telephone (n = 14) (data not shown).
Benefits of the CG intervention
The benefits of CG pertained to themes of supporting clinician learning, ensuring clinician competence, and organisational quality assurance.
Supporting clinician learning
a) Peer feedback levels: many of those with experience of in-hours general practice noted peer feedback to be infrequent in that setting. By comparison, CG was felt to have a positive impact on feedback frequency, circumventing many identified causes of infrequent feedback such as time, clinical isolation, professional hierarchy, and avoidance of conflict:
‘ [CG is] the only feedback I really get on my documentation … my colleagues are probably too nice, and they’re not in a rush to offer me that feedback.’
(GP8)
b) Identification of learning needs: clinicians recognised inconsistencies in their own practice and how minimising this could improve care:
‘… there’s a high rate of variation even in my own practice so I think anything we can do where we spot inconsistencies, could potentially improve quality.’
(GP1)
In contrast with existing, predominantly reactive quality assurance structures, CG supported proactive identification of clinicians’ learning needs:
‘At the moment, the way people’s unknown unknowns get picked up is some sort of significant event or complaint when something has gone wrong. Otherwise, it just passes under the radar … ’
(GP7)
Owing to its frequency and detail, CG facilitated feedback regarding issues that would otherwise not be highlighted, with potential patient benefit through optimisation of existing clinical management:
‘ [Without CG] you don’t get the sort of small bits of feedback or slight nudges to improve like, “make sure you document your safety netting.” I mean who is ever going to spot that otherwise?’
(GPT2)
c) Validation of practice: CG was identified as a means of validating clinician practice and benchmarking with peers . The use of CG to reflect on how clinicians’ actions correlated with patient outcomes provided a mechanism to reinforce positive practice or trigger learning .
Ensuring clinician competence.
a) Clinician supervision: CG was felt to ensure a minimum standard of supervision for a professionally diverse and often transient out-of-hours general practice workforce:
‘… [CG is] really crucial because you’re employing a [range] of clinicians, and a lot of them are nurses and paramedics. So, how on Earth do you check somebody is alright? You have a responsibility to the patients to ensure … you’re checking up on standards … ’
(CGPRT1)
Owing to infrequent peer feedback in in-hours general practice, many clinicians noted CG feedback on their out-of-hours role was the only form of clinical supervision they received. Many felt the accountability provided by such supervision was likely to improve clinical practice:
‘… there could be an element of … if you know that someone is checking your work, you might be a bit more thorough. There shouldn’t be, but there probably is.’
(GP7)
There were concerns that CG could be seen as a replacement for supervision that should be more formalised and detailed:
‘I think the problem with doing Clinical Guardian is it could too easily become a substitute for something that should be a lot better.’
(GP6)
Inadequate supervision was recognised to have greater consequences for those with less clinical experience, and in the context of an increasingly multidisciplinary general practice workforce, risked deploying clinicians outside of their competence.
Acknowledging those concerns, participants emphasised competence to be a function of clinical practice rather than something to be inferred by professional title, and recognised value in supervision structures that apply equally to all clinicians:
‘ ... it doesn’t actually matter whether it’s a GP or not because we are all doing the same job, so we all have to be competent … ’
(GP6)
While CG may not provide a gold-standard of supervision for all clinicians, there was recognition that it represented an improvement on the perceived lack of consistent approaches in in-hours general practice:
‘ … [CG] gives a … solidity to the service in terms of that there is a running check of records of every clinician. In-hours general practice is 55 miles from that.’
(SMT2)
b) Identification of clinicians in need of support: while the number of performance outliers were noted to be low, CG was felt to be an effective mechanism to identify clinicians in need of support:
‘If you get … someone who is really not doing the things they should do, it’s a really effective way of just picking that up.’
(CGPRT1)
Continuous sampling of clinician practice was seen as a strength of CG, enabling the identification of patterns of behaviour that may highlight a need for support, and the universal application of CG ensured supervision of those who might not otherwise seek it.
c) Appraisal and revalidation: most clinicians supported the rationale for nationally standardised structures to ensure clinician competence, such as appraisal and revalidation. However, many did not feel such mechanisms were effective, and the reliance of appraisal on largely self-collated information was noted to create adverse incentives that may undermine it:
‘… I am much less likely to pull out cases where to be honest I am pretty sure I haven’t done the best thing [in my appraisal], than I am … with my colleagues.’
(GP6)
Clinicians indicated CG enhanced the quality of evidence they could submit for appraisal, and was potentially less biased than other means of assessing clinician competence, owing to its risk-based case sampling and equal application to all clinician groups.
Organisational quality assurance
a) Clinical governance: participants with experience of working for >1 service provider noted interorganisational variation in clinical governance culture and practices, with patient safety incident reporting identified as an area of inconsistency. CG was seen to help standardise such approaches.
Participants indicated clinical governance could be better integrated between organisations, and standardising approaches with the support of systems such as CG was seen as a way to facilitate interorganisational learning.
b) Organisational learning culture: for many, CG supported a positive learning culture and sense of organisational connectedness, which was noted to be harder to achieve in larger organisations. Some felt CG introduced a sense of hierarchy; however, this viewpoint was not widely held, and sharing learning from CG reviews at an organisational level via meetings and emails facilitated the perception of CG learning as a team exercise. Investment in CG was felt to communicate organisational values to clinicians, external organisations, and patients.
Factors limiting the usefulness of CG
Factors limiting the usefulness of CG pertained to the intervention itself, the clinician, and organisational context.
Intervention factors
a) Feedback quality and frequency: feedback containing written comments, rather than categorical grading, was perceived as being most useful for clinician learning. CG was recognised to focus on those most in need of support; therefore, many did not receive detailed feedback regularly, attenuating its use to them as a learning tool.
Constructive feedback was recognised to be of greatest value in supporting learning, but a potential source of anxiety and defensiveness:
‘It’s really easy to give positive feedback. It’s quite difficult to broach the thorny issue of trying to suggest somebody does things in a different way.’
(CGPRT1)
These concerns were rationalised in terms of concerns about the opinion of peers; that their actions could have caused patient harm and the associated medicolegal implications.
b) Selection of clinical cases: CGs’ use of random consultation record sampling was felt to reduce bias; however, some highlighted learning-points may be more efficiently identified through more purposive case selection. Some noted that when they had self-selected challenging cases for peer review this was highly valued.
c) Limitations of clinical note reviews: the CG peer-review process is principally conducted using consultation records. Interviewees recognised these to be a subjective representation of a clinical interaction from the clinician perspective, and therefore a potential limitation:
‘… notes aren’t the whole picture, it’s the way that the GP documents the kind of transaction … ’
(GP5)
Consequently, clinical records were observed to be vulnerable to unintentional or intentional misrepresentation.
Overemphasis on the clinical note quality may neglect other important aspects of practice such as consulting speed, breadth of competence, and quality of communication. Notes may also not adequately reflect the context of a clinical encounter, where difficult decisions may be made on a busy shift:
‘… there’s always going to be a limitation where there’s only one person at a distance looking at something you’ve done without the context or the business of the shift … ’
(GPT1)
Despite these limitations, the strength of notes reviews to appraise the key points of a clinical interaction supported its role as a quality assurance tool:
‘… [CG is] not really digging into what’s actually going on in the consultation … but yet it’s useful because you can see … that they’re taking good sets of records, taking appropriate clinical observations, and taking an appropriate course of action for a defined clinical problem.’
(SMT1)
Clinician factors
a) Clinician experience: less experienced clinicians appeared to find CG more useful as a consequence of having a greater proportion of their consultations reviewed than more experienced colleagues and being trained through a similar culture:
‘… younger GPs who’ve come through … training where they’re used to reflecting, being observed, getting lots of feedback … in general seem to find [CG] more useful ... ’
(SMT3)
Some more experienced clinicians felt they should be audited less; however, most reflected they would value more feedback, and experience did not negate the need for scrutiny:
‘There’s lots of GPs who maybe qualified 40, 45 years ago, don’t write very detailed notes. There has been a change in what is considered appropriate so hopefully [CG] encourages that.’
(GP7)
b) Clinician motivation: Clinicians were noted to be broadly receptive to peer-review interventions owing to common professional traits:
‘… doctors tend to get … very reflective … and very self-critical. They’re high achievers who’ve got quite high standards.’
(CGPRT1)
Interviewees recognised the focus of CG was to support clinicians, rather than catch them out. Most reflected a preference for more peer review of their practice , with those most committed to self-development finding CG most useful.
Organisational factors
a) Organisational performance: the potential impact of CG was noted to be affected by the strength of wider governance structures and existing organisational performance.
b) Learning culture: while some felt monitoring their practice through CG was within the spectrum of assurance processes they would expect in any clinical service, others noted the potential impacts of such interventions on the health and retention of already stretched clinicians:
‘... you have to balance, don’t you, patient safety and doctor morale ... being overly watched and scored is a big factor in doctor morale and burnout and stress ... ’
(GP3)
These tensions highlighted the importance of learning culture in influencing perception of such interventions. While CG was reported to promote many positive aspects of organisational learning culture, it was emphasised as a tool to support this, rather than a substitute, and the critical role of organisational leadership in setting such a culture was recurrently noted:
‘… clinical governance [is] … about culture and climate and permission to fail … it is the senior people who set that climate for better or worse. If [CG] is ever going to be taken further, that has to be a focus.’
(SMT2)
