The medical workforce is key to service quality. Organisations have a duty to develop their workforce—to ensure professional development, good governance and, from time to time, discipline staff. But what if part of the workforce is contracted from outside to fill gaps in the rota? That is the world of the ‘locum’—a peripatetic medical workforce that is in, but not of, the organisation.

Locum doctors are deployed in many countries of the world. There is a thriving international market across English-speaking countries,1 Western Europe2 3 and in the USA, where the Veterans Administration alone pays about $50 million per annum for temporary medical staff.4 Considering the size and importance of this human resources market, the subject has attracted surprisingly little academic attention.

Ferguson and colleagues peer into the world of the medical locum through an in-depth qualitative study based on interviews and focus groups.5 Participants included locums, non-locum clinicians and service managers as well as patients, mostly from hospitals and primary care centres in England. The emphasis is on ‘how locum doctor working arrangements affect quality and safety’. Based on their previous narrative systematic review,6 the authors claim that they have conducted the largest study on the topic of locums. The authors also claim that the existing literature is ‘largely contextual’ and does not cover the influence of the broader organisational system on the performance of locums.

Ferguson et al’s study found that the life of the locum is a difficult and lonely one, opening up many pathways to unsafe practice.5 First, it is very difficult for locums to adapt to different procedures and protocols as they move from one organisation to the next, although occasionally this unfamiliarity was beneficial in detecting quality and safety issues that existing staff took for granted. Second, locums are not routinely involved in quality development and quality improvement meetings, an issue that has greater impact on an organisation as the number of locums increases as a proportion of the medical workforce. Third, locums experience negative behaviours from permanent staff to the point of stigmatisation. Fourth, while permanent staff feel that locums often fail to resolve problems and that they practise defensively, this was seen by locums as a reaction to a hostile environment in which they may be scapegoated. Fifth, locums tend not to be included in induction and appraisal and may have moved on when feedback, including complaints, would otherwise be received.

The above evidence suggests that locum doctor arrangements are unkind and unfair, and potentially harmful to patient safety. This conclusion is buttressed by the observation that hospitals with low Care Quality Commission (the independent regulator of health and social care in England) scores make proportionally heavier use of locums than those with higher scores7 and by empirical observations and theoretical insights from the organisational science literature. Given Ferguson et al’s stated intention to emphasise context and take a ‘systems approach’,5 it is surprising that they do not seek to compare primary and secondary care contexts. Things may play out very differently in the more circumscribed context of primary care as opposed to the diffuse world of hospital practice where clinicians range across outpatient departments, operating theatres, hospital wards and so on. Ferguson et al provide a vivid and rather negative analysis of the world of the locum doctor. This is inconsistent with a German study which provides a much more optimistic picture finding that locums have greater job satisfaction than substantive staff, which is attributed to less bureaucratic control and greater perceived autonomy.2

So much for the diagnosis, what about the treatment? The authors provide few ideas for action, and when they do, the recommendations are rather general—‘create an environment which supports locum doctors…’. This is not a criticism since providing recommendations to improve locum services was not the purpose of the paper.

Nevertheless, findings from applied research must be actioned if they are to have value. Three issues arise:

1. Can we reduce reliance on locum cover?

Locum cover is expensive and, based on the above article, a threat to safety. However, the locum market meets a need. At the demand side, it fills gaps in the workforce arising from inevitable churn—doctors call in sick, resign at short notice or take other types of leave where temporary replacements are usually the solution. Not surprisingly, small hospitals employ proportionately more locums than large hospitals.7 Locums thus avoid efficiency loss and underutilisation of facilities and may improve safety by closing gaps in the service. At the supply side, locum work provides for a flexible lifestyle and can act as a ‘filler’ while developing a career, and locum hourly rates of pay are higher than for substantive posts at the same grade. Thus, locum services are here to stay. There nevertheless seems to be a good case for bearing down on the market and strongly encouraging all posts to be filled with non-locum staff—less money spent on locum doctors with more money for the substantive posts. Furthermore, the findings suggest that it is difficult to develop culture and standardise processes when a large portion of the workforce is made up of locums. Inspection processes could monitor the use of medical locums and nudge hospital managers to model their workforce requirements to find the optimal balance between substantive and temporary posts.

2. Is it possible to recommend specific actions based on the findings of this paper?

The findings by Ferguson et al highlight how locum working might affect patient safety as a result of the difficult environment in which locums have to work. Actions that organiations might take to mitigate these effects, as well as actions to benefit from the fresh pair of eyes that locums may provide, include:

Because locums find it hard to adapt to different procedures and protocols, organisations could be incentivised to standardise. A pilot can change effortlessly from flying one route with one team to another route with a different team that they have never worked with, because of high levels of standardisation. Given that locums are an ineluctable part of the workforce, standardisation is likely to improve safety. Widespread use of computerised notes and order entry systems should facilitate standardisation as would things like store cupboard layout and procedures to bleep staff.

Because locums operate largely outside of normal governance processes, they could have a bespoke induction when they join a clinical service to further improve safety, perhaps with an additional meeting to ensure key points were captured.

Because locums suffer discrimination, governance management for substantive staff could include the need to foster kindness generally and to raise consciousness of the barriers locums described in the paper.

Because locums often range across multiple organisations, their agencies could receive structured feedback from clinical services.

Since locums experience many services, they could be used as mini-management consultants and be asked for verbal or written feedback when they leave. This is also a point made by the German study referenced above who argue that locums contribute to exchange of knowledge and to the ‘uncovering of hazardous organisational practices’.2

3. What should be done to ensure that changes are implemented?

The policies and actions that might follow from Ferguson et al’s paper5 will not just happen; they require careful design. I recommend that, in England and beyond, design groups should be formed including policy makers, service managers, local agency managers and public contributors to develop a set of workable solutions for subsequent piloting, careful evaluation and later implementation. Thinking more broadly, the picture that emerges from both Ferguson et al and the German study is that locums are less beholden to bureaucratic control than permanent staff. The implication that locums need more bureaucratic control may be tempered by the thought that permanent staff are overly constrained by excessive bureaucratic requirements. Perhaps what we really need is more enquiry into the optimal level and types of bureaucratic control which should then be applied to substantive and locum posts alike.

In conclusion, the paper from Ferguson and colleagues provides useful findings regarding locums and their impact on patient safety. The paper should not be simply curated among the voluminous safety literature. It should be considered as a call to action by senior policy makers.

Not applicable.