Quantitative results
In total, locums accounted for 11% of 1 511 787 consultations analysed. Acute sore throat (26% of all identified antibiotic prescriptions) and acute cough (38%) were the two main conditions for which antibiotics were prescribed (data not shown). The number of patients receiving antibiotics for these conditions was 4% higher (on an absolute scale) when consulting with locums compared with when consulting with other GPs (Figure 1). A similar difference was observed for asthma exacerbations and acute bronchitis, while prescribing percentages were more similar for other conditions (Figures 1 and 2). An exception was impetigo, for which other GPs prescribed antibiotics more often than locums (54%, 95% confidence interval [CI] = 53% to 55% versus 47%, 95% CI = 45% to 48%) (Figure 2).
Figure 1. Proportion of patients without relevant comorbidities receiving antibiotics when consulting with acute respiratory conditions with locum GPs, other GPs, or nurse prescribers in primary care. AOM = acute otitis media.
Figure 2. Proportion of patients without relevant comorbidities receiving antibiotics when consulting with non-respiratory infections with locum GPs, other GPs, or nurse prescribers in primary care.
For several conditions, nurses prescribed antibiotics in a smaller proportion of consultations (Figures 1 and 2) than locums and other GPs. However, this pattern was not consistent across all conditions: a higher percentage of children (aged 2–18 years) consulting with acute otitis media received antibiotics in consultations with nurse prescribers compared with locums or other GPs (Figure 1).
Qualitative results
Nineteen locums were interviewed (Table 1). Interviews lasted 38–64 (mean 49) minutes. Four themes were identified to capture locums’ experiences and perceptions of influences on antibiotic prescribing and AMS. Findings are supported by quotes below, with additional quotes in Supplementary Box S1.
Theme 1. Antibiotic prescribing as a complex but individual issue
Interviewees described antibiotic prescribing as complex clinical decisions influenced by individual-level factors, such as the GP’s experience, skills, confidence, the patient’s clinical presentation (for example, symptoms), and expectations/behaviour. Some described an increasing awareness of AMS among GPs and patients; others argued that more change is still needed. They perceived locums as having an important role in AMS because they constitute many prescribers, but that role was seen as similar to all GPs’ responsibility for appropriate prescribing:
‘It’s very much an individual GP responsibility really. If a practice stressed to me the importance of avoiding inappropriate antibiotics, I’d almost feel like they were stating the obvious […] all GPs should understand what appropriate antibiotic prescribing is, that’s real basic bread and butter general practice.’
(Locum [L]1, male [M], 18 years since qualifying as a GP)
Participants also reported varied experiences and views of using AMS strategies (for example, guidelines, leaflets, delayed prescriptions, and clinical scores). Most did not know their individual antibiotic prescribing rates nor receive feedback on it, but they believed that they were low/prudent prescribers. They were interested in feedback on their prescribing and thought that it would be useful to locums and all GPs, and could be incorporated into GP appraisals. While most participants did not know the prescribing rates in practices they worked in, they had little or no interest in this, because as locums they described having no ‘vested interest’ in the practices:
‘I’d be more interested in my own prescribing in comparison to the other doctors within the practice […] I don’t have a vested interest in any of the practices that I’m working in, it wouldn’t be that useful to me to know whether they’re high or low prescribers ...’
(L9, M, 17 years)
However, they were also unsure how locums’ prescribing could be identified, as partner GPs’ names were often on prescriptions issued by locums.
Theme 2. Nature and patterns of locum work
The overall nature and patterns of locum work seemed to influence locums’ antibiotic prescribing. Participants worked across different practices and areas, which varied considerably in IT systems, workflows, and prescribing guidelines. This variation was challenging for locums who, as a result, might not always follow the local guidelines or workflows. For example, some participants reported following familiar prescribing guidelines from a different area to where they currently worked, and relied on IT prompts to indicate non-concordance with local guidelines. Participants suggested that working in one local area and regular, longer-term practices helped to minimise this challenge, and that adopting similar guidelines and approaches would make appropriate prescribing easier for locums.
Participants described how locums have more control and flexibility over their work. Thus, some requested longer appointments or catch-up slots to ensure sufficient time to provide good-quality care (for example, discussing antibiotics and safety-netting). Others perceived locums as being under more pressure from patients and time owing to hourly payment. This could lead to quickly ‘closing’ consultations with antibiotic prescriptions (rather than taking time to discuss patients’ perceptions about antibiotic treatment) to avoid running over:
‘… the pressure that patients put on locums to prescribe, and the pressure of time on the locum […] it is to do with the time that we have, so you’re paid by the hour. You don’t necessarily want to run over […] think that makes a lot of locums more likely to not want to have that discussion to change patient perceptions about their use of antibiotics, and that it’s just easier to give antibiotics.’
(L16, female [F], 4 years)
Participants also described another aspect of control and flexibility: being able to choose the practices where they work. For example, they reported avoiding practices perceived as being disorganised and struggling with demand and insufficient staff (and therefore having more staff turnover and less control over prescribing quality). They perceived appropriate prescribing and good-quality care as being more difficult in such practices.
Locums reported consulting more patients with acute infections who were more likely to require antibiotics (while patients with chronic illnesses were seen by regular doctors). They also reported that seeing unfamiliar patients and not being able to follow patients up can make locums more likely to prescribe antibiotics to avoid the risks of not prescribing, such as complaints and additional work for other GPs if patients re-consulted. However, participants also described how seeing unfamiliar patients may put locums in a better position to suggest a ‘new’ no-antibiotic approach, and that locums might be less concerned about potential negative impact of not prescribing on doctor–patient relationship:
‘… because I’m not their normal GP […] I have the time and I have the fresh pair of eyes to go, “Actually, you know, things are changing a bit and I read something recently or I’ve been to an education session and how about trying without antibiotics this time?”’
(L3, F, 22 years)
Theme 3. Relationships between practices and locums
Participants’ relationships and communication with practices varied considerably. They reported no communication about practices’ initiatives or approaches related to prescribing and AMS (for example, antibiotic-related targets and priorities), and said that they could participate or contribute if they were told about them (for example, during inductions). There was generally little to no feedback between practices and locums (unless concerning safety issues), and even less scrutiny over, and accountability for, locums’ prescribing than for regular clinicians. Participants reported that receiving feedback from practices would be helpful for them to improve and/or feel appreciated.
Participants described how locums were generally perceived, and felt to be ‘ just to see patients’ (L10, M, 3 years) and not a part of practice teams. Thus, some reported that locums have neither influence on, nor a role in influencing, practices’ antibiotic prescribing, or AMS initiatives:
‘… you’re not part of the team and they don’t make you feel part of a team, you’re just there to come in and cover the session and that’s all you’re there to do. You’re not involved in discussions about prescribing or the processes in the practice […] it’s not really my role as a locum to get involved in trying to change processes that don’t seem to be working.’
(L3, F, 22 years)
Others noted that locums working across many different practices have opportunities to observe, compare what works well with what does not, and identify potential improvements. Although locums’ feedback to practices was rare, some reported contributing to improvements in practices where they had good relationships.
Although participants reported that they practise similarly regardless of their role or where they work, the influence of different organisational cultures was apparent. Patient notes (that is, when and how consistently other GPs prescribed) and patients’ expectations for antibiotics gave locums a sense of the practices’ approaches to prescribing. Some reflected on being more inclined to prescribe antibiotics in higher-prescribing practices and where they would not feel supported by other GPs when not prescribing:
‘I saw a patient today who has COPD and you look back and see that’s what they do and I’m more inclined to prescribe antibiotics just in case because that’s what they do, rather than making my own judgment, I’m just following what the practice are doing.’
(L11, F, 5 years)
Theme 4. Professional isolation
Participants described GPs as ‘working in silos’ (L10, M, 3 years) and locums as even more professionally isolated than practice-based GPs. This was exacerbated by limited or no communication from commissioners, being less connected to professional groups and networks, and having to participate in professional training and meetings during unpaid time. Consequently, locums found keeping up to date with guidelines, evidence, and training more challenging, and having fewer opportunities for peer learning. This could contribute to less appropriate prescribing:
‘ [Locums] are a little bit outside […] the mainstream GPs who are going to all the regular CPDs [continuing professional developments] and GP updates and maybe are more aware of the problems around antibiotic stewardship and resistance […] as a locum you can go to no CPD meetings […] you can be far away from the nourishing flow of information […] and you can see how you can have a very different viewpoint about antibiotic prescribing.’
(L10, M, 3 years)
Nevertheless, some participants described locums as well trained and aware of evidence as a result of being more proactive about their professional development. Many reported being proactive about ensuring they had access to relevant resources (for example, bookmarking online guidelines/tools) and communicating with practices and peers (for example, asking about training opportunities and joining meetings) to provide good care.
Participants discussed how practices and commissioners should better integrate and support locums; for example, by circulating information, updates, and training opportunities to all GPs registered on Performers Lists. Some suggested that AMS-related training should be mandatory, and that local peer groups for locums should be encouraged. Finally, some suggested that locums should be better recognised as a considerable professional group, and involved in wider policy development:
‘We’re a significant amount of the workforce and we prescribe lots of antibiotics […] However, because we can’t influence local policy, perhaps our involvement would be limited only based on our personal experience […] our voices are minimally heard in general when we talk about health policy, just because locums and sessional GPs is just this nebulous group that aren’t really organised very well. So I think you’re missing probably lots of people that could add value to any sort of policy discussion.’
(L5, F, 7 years)
Box 1 summarises the main influences reported by the interviewees on locums’ antibiotic prescribing, and suggestions for potential improvements.
Challenges and reasons for higher antibiotic prescribing | Opportunities and reasons for lower antibiotic prescribing | Strategies used by locums to manage challenges | Suggestions |
---|
See more acute patients See more unfamiliar patients with limited follow up (and wanting to avoid work for others) May feel less accountable for their prescribing (no audit or feedback) May feel less invested in or concerned by antibiotic prescribing in practices where they work as locums Less (access to) training and peer learning May be under more pressure from patients seeking antibiotics May feel under more time pressure (antibiotic prescribing is seen as quicker than not prescribing) Less aware of practices’ AMS initiatives May feel influenced by practices’ high-prescribing culture and feel unsupported when not prescribing antibiotics (want to avoid risks and complaints)
|
No pre-existing relationship and expectations from patients (easier to suggest a ‘new’ no-antibiotic approach and less worried about impact on the relationship) Well trained and aware of the evidence May work more flexibly and take longer in consultations if needed to provide good care
|
Use typical AMS strategies (for example, guidelines and clinical scores) Select practices that are ‘good’ to work in, and avoid practices perceived as more disorganised and with higher staff turnover Work locally and in regular, longer-term practices Ensure extra time to familiarise with new practices Keep own notes/information/links related to local guidelines, processes, and patients to follow up Agree/request sufficient time for good-quality care Initiate communication with colleagues and take time to develop good relationships Ask for support when needed Rely on IT prompts for first-line antibiotic Ask practices for information about relevant training or meetings and attend them Join local GP groups or locum organisations
|
Audit locums’ prescribing Enable locums to issue prescriptions signed with their names, and link locums’ prescribing to their roles Provide feedback to locums, especially on individual antibiotic prescribing; invite locums’ feedback/suggestions for improvements to practices Use appraisal/revalidation to influence antibiotic prescribing (for example, require antibiotic prescribing audit and training) Adopt similar IT systems, guidelines, and processes across regions Improve inductions, including information about practice’s AMS approach and support for prudent antibiotic prescribing Use IT prompts and solutions to promote appropriate prescribing Organise locum peer groups, or include locums in local GP groups Provide free access to and encourage participation in AMS training Need whole-system approach to AMS, including ‘educating patients’
|
Box 1. Summary of perceived influences on locums’ antibiotic prescribing, strategies, and suggestions