Introduction Engagement of junior doctors in patient safety initiatives is high on the national agenda, but there is a lack of studies evaluating patient safety attitudes among junior doctors.
Methods The Junior Doctor–Patient Safety Attitudes and Climate Questionnaire is a multidimensional scale created using items from already-validated scales and inclusion of new items based on further review. It consists of three subscales: ‘knowledge and training’ (10 items), ‘attitudes to patient safety’ (15 items) and ‘perception of workplace safety climate’ (15 items). This was disseminated to foundation trainees, general practice trainees and hospital core and speciality trainees via the Deanery distribution lists and responses were collected anonymously.
Results A total of 527 complete responses were collected; although self-declared knowledge in patient safety concepts was high, there was less declared understanding of a ‘high reliability organisation’ (74% no/unsure) and the concept of active failures/latent conditions (60% no/unsure). The greatest agreement was demonstrated for the statement ‘Even the most experienced and competent doctors make errors’ (p<0.01). However, more senior trainees and surgical trainees (vs medical trainees) demonstrated greater agreement with ‘Medical error is a sign of incompetence’ (p<0.01). More junior trainees demonstrated greater agreement with ‘Management is more interested in meeting performance targets than focusing on patient safety issues’ (p<0.01).
Conclusions This study demonstrates subtle differences in attitudes to patient safety among junior doctors of different grades and specialities. These should be taken into account when designing interventions to improve patient safety education and culture among junior doctors.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
In the UK Department of Health's report, An Organisation with a Memory,1 the initiation of ‘clinical governance’ at a local organisational level was seen as the way to enhance patient safety performance, calling for unified reporting mechanisms, a systems approach in preventing, analysing and learning from error, but also a more open culture and active learning in which lessons were embedded into an organisation's culture and practice.
Reason2 elegantly highlighted that organisational accidents arise from the alignment of several contributing factors originating at many levels of the system because of the many barriers of protection, such catastrophic accidents, are rare events. In an ideal world, the defensive layers would be intact, but in reality they are full of holes like ‘Swiss cheese’; these gaps, weaknesses and failures occur for two reasons:
Active failures: unsafe acts (errors or violations) by those in direct contact with the system
Latent conditions: defensive gaps made by designers, builders, regulators and managers of the system; these exist because in a complex hazardous system all possible accident scenarios cannot be foreseen.
Although the latent conditions are the primary targets of a safety management system, most failures do begin with unsafe human actions, usually due to local triggers within the workplace (which are in turn a product of higher level latent conditions).2 Latent conditions can never be completely eliminated; therefore, the ‘last line of defence’ is always the junior doctor or nurse in direct patient contact and these are the target groups in which reliability and ‘error wisdom’ could be augmented through simple training in patient safety concepts.
The second advantage of engaging junior doctors in patient safety is related to tacit knowledge. Healthcare organisations often learn from error by the use of local and national reporting systems; however, such voluntary reporting systems do not provide reliable and systematic information for monitoring patient safety because many incidents go unreported.3 Management often focuses on information and data capture systems such as ‘incident reporting’; however, it is unable to capture the tacit knowledge that could enrich this from professionals on the frontline, with studies reporting a capture rate of only 1.5% of adverse medical events that occur at the sharp end.4 Junior doctors often make and witness errors,5 and therefore hold a significant amount of this tacit knowledge and patient safety intelligence from the frontline.
It is important to understand the attitudes of doctors to the patient safety agenda, in particular incident reporting, but also to patient safety initiatives as this provides a reflection of the safety culture which organisations have to understand to ensure engagement is meaningful. Data from industries where safety are of critical importance have demonstrated that safe performance is a ‘function of staff preoccupation with failure avoidance’. Therefore, assessing staff safety attitudes is an important approach to improving safety.6
Although studies have explored the perception of safety climate from the perspective of healthcare professionals using validated cultural assessment tools (mostly in the USA), none have been specifically targeted at junior doctors in the UK. The aim of this study was to develop an appropriate questionnaire and apply it to evaluate patient safety knowledge and awareness, attitudes to patient safety and perception of safety climate among junior doctors across a range of specialities and grades.
As there was no specific validated questionnaire identified in the literature focusing on junior doctors, we used items derived from three validated scales targeting medical students7 ,8 and healthcare professionals in America.9 The items were modified with a view to provide greater face validity from the perspective of UK-based junior doctors by two of the authors (PD and JD). The combined questionnaire is the Junior Doctor Patient Safety Attitudes and Safety Climate Questionnaire attached in online supplementary appendix 1. It consists of three domains, domain A: Knowledge and Training in Patient Safety (10 items), domain B: Attitudes to Patient Safety (15 items) and domain C: Safety Climate on My Ward/In Department (15 items). Items in domain A use categorical ‘yes/no/unsure’ response options and items in domains B and C use Likert-type 5-point agreement response options, from ‘strongly disagree’ to ‘strongly agree’.
This questionnaire was distributed via online ‘surveymonkey’ link and online questionnaire in PDF format via East Midlands Healthcare Workforce Deanery by Programme Directors who agreed to their trainees’ participation, including the Foundation School (n=800), GP VTS registrars (n=500), Medicine (n=320), Surgery School South (n=135), Anaesthetics School South (n=60) and Obstetrics & Gynaecology North (n=60).
Results are reported with descriptive frequencies. For the Knowledge and Training, domain A, χ2 test was used to evaluate differences in responses by grade and speciality for proportions endorsing ‘yes’ to each item. For the Attitudes to Patient Safety, domain B, and Safety Climate, domain C, the ‘strongly agree’ and ‘agree responses’ were collapsed to an ‘agree’ category and the ‘strongly disagree’ and ‘disagree’ responses were collapsed to a ‘disagree’ category, with ‘neutral’ left separate. Proportions endorsing each of these categories were calculated and for each statement, the ‘correct/desirable’ response was identified (either ‘agree’ or ‘disagree’); for example, in domain B, for the statement ‘Even the most experienced and competent doctors make errors’, the desired/correct response exhibiting positive safety attitude would be ‘agree’. The single-sample binomial test was used to test the null hypothesis that the proportions endorsing the ‘correct’ and ‘incorrect’ responses were equal at 50%. This was performed with the ‘incorrect’ responses including neutral responses (ie, for the above statement, % endorsing ‘disagree’ or ‘neutral’) and then also repeated with the ‘incorrect’ response excluding the neutral responses (ie, % endorsing ‘disagree’ only).
The proportions endorsing the ‘correct’ response for each statement in domains B and C were compared for statistical differences between grades (Foundation year trainees and Core trainees (CT) vs Registrar trainees) and between specialities (Medicine vs Surgery) using two-way χ2 test. The level of significance was set at p<0.01 (99%) and all data analysis was conducted with SPSS PASW V.19 (SPSS Inc., Chicago, USA) and Stata V.12 (StataCorp, College Station, Texas, USA).
A total of 527 junior doctors completed the questionnaires with a response rate of 28% (n=1875). For group comparisons, respondents were stratified into three main groups by grade from junior to senior (with 12 self-reporting no grade):
Registrar trainees (Reg at ST3+/SpR level), n=254.
Responses were also compared by speciality background of junior doctor and stratified into medical trainees (n=53) and surgical trainees (n=71).
Domain A: knowledge and training
Descriptives, frequencies and means
Table 1 shows the proportions for each response option in domain A; overall, 73% felt their training was providing adequate preparation to understand the cause of medical errors. Although self-declared knowledge in patient safety concepts was high, there was less declared understanding of a ‘high reliability organisation’ (74% no/unsure) and the concept of active failures and latent conditions (60% no/unsure).
Group comparisons by grade
Senior trainees endorsed ‘yes’ less frequently for the items, ‘I understand differences in errors and violations’ (46% Reg vs 63%/66% FY/CT; χ2=16.89, p<0.01) and ‘awareness of Swiss Cheese model of error causation’ (50% Reg vs 75%/71% FY/CT; χ2=31.66, p<0.01), while more junior trainees endorsed ‘yes’ less frequently for the item ‘I understand the role of healthcare organisations in error management’ (71% Reg vs 56%/68% FY/CT; χ2=9.94, p=0.01). There were also significant differences in proportions endorsing ‘yes’ to the statement ‘I understand the concept of ‘active failures’ and ‘latent conditions’ in the cause of an adverse event’ (34% Reg vs 40% FY vs 55% CT; χ2=12.16, p<0.01).
Group comparisons by speciality
A statistically significant difference between medical and surgical trainees was only demonstrated for proportions endorsing ‘yes’ to the statement ‘I understand the role of healthcare organisations in error management’ (84% medics vs 52% surgeons; χ2=13.313, p<0.01).
Domain B: attitudes to patient safety
Descriptives, frequencies and difference in proportions endorsing ‘correct’ response
Table 2 shows the proportions for each response option; proportions endorsing the ‘correct/desired’ response were significantly different to those endorsing the ‘incorrect’ response for all of the statements when excluding neutral responses from the ‘incorrect’ category (online supplementary appendix 2A lists p values in full). When neutral responses were included as ‘incorrect’, two statements, ‘If a junior doctor makes an error, the Consultant should take most of the responsibility’ and ‘I believe that filling in error reporting forms will help to improve patient safety’, did not achieve statistical significance (suggesting greater ambivalence/neutral attitudes to these).
The statement with the highest proportion of junior doctors endorsing the ‘correct’ response was ‘Even the most experienced and competent doctors make errors’ (99% ‘agree’/‘correct’ vs 1% ‘disagree or neutral’/‘incorrect’, p<0.001). The statement with the lowest proportion endorsing the ‘correct’ response was ‘It is only important to disclose errors to patients if they have resulted in harm’ (58% ‘disagree’/‘correct’ response vs 42% ‘agree or neutral’/‘incorrect’ response, p<0.001).
Group comparisons by grade
When compared with senior trainees, a significantly greater proportion of junior trainees endorsed the ‘correct’ response to the statements, ‘Medical error is a sign of incompetence’ (60% ‘disagree’ Reg vs 77% ‘disagree’ FY/CT; χ2=15.798, p<0.001) and ‘It is only important to disclose errors to patients if they have resulted in harm’ (50% ‘disagree’ Reg vs 63% ‘disagree’ FY/CT; χ2=7.019, p=0.008) (online supplementary appendix 2B).
Group comparisons by speciality
Compared with medical trainees, significantly lower proportions of surgical trainees endorsed the ‘correct’ responses for the following statements (online supplementary appendix 2C):
‘Medical error is a sign of incompetence’ (1.4% ‘disagree’, surgeons vs 80% ‘disagree’, medics; χ2=81.432, p<0.001)
‘Learning about patient safety is not as important as learning other more skill-based aspects of being a doctor (7% ‘disagree’, surgeons vs 80% ‘disagree’, medics; χ2=67.511, p<0.001)
‘It is only important to disclose errors to patients if they have resulted in harm’ (16% ‘disagree’, surgeons vs 69% ‘disagree’, medics; χ2=35.826, p<0.001).
For the statement, ‘The number of hours doctors work increases the likelihood of making errors’, a lower proportion of surgical trainees endorsed the ‘agree’ option but this only achieved statistical significance at the 95% CI level (56% surgeons vs 74% medics; χ2=4.444, p=0.035).
Domain C: safety climate of my ward/in department
Descriptives, frequencies and differences in proportions endorsing ‘correct’ response
Table 3 shows the proportions endorsing the response options for each statement. Online supplementary appendix 3A lists each statement, the difference between proportions endorsing the ‘correct/desired’ response and ‘incorrect’ response, including and excluding neutral responses from the ‘incorrect’ category and whether the differences were statistically significant. When neutral responses were excluded from the ‘incorrect’ category, all statements demonstrated statistically significant difference in proportions endorsing the ‘correct’ versus ‘incorrect’ option, except two statements, ‘The senior managers in my hospital listen to me and care about my patient safety concerns’ and ‘Nurses on my ward are likely to criticise me for making mistakes’, that is, there was neither significant agreement or disagreement with these statements and responses were more likely to be neutral or be distributed similarly among ‘disagree’ and ‘agree’ options.
The statement with the highest proportion of junior doctors endorsing the ‘correct’/desirable response was ‘Senior doctors in the department listen to me and care about my patient safety concerns’ (70% ‘agree’/‘correct’ vs 30% ‘disagree or neutral’/‘incorrect’, p<0.001). The statement with the lowest proportion endorsing the ‘correct’/desirable response was ‘Management is more interested in meeting performance targets than focusing on patient safety issues’ (20% ‘disagree’/‘correct’ response vs 80% ‘agree or neutral’/‘incorrect’ response, p<0.001).
Group comparisons by grade
Significant differences in proportions endorsing the ‘correct’/desirable option by grade were demonstrated across a larger number of items (online supplementary appendix 3B). When compared with junior trainees, a significantly greater proportion of senior trainees endorsed the ‘correct’/desirable response to the following statements:
‘Management is more interested in meeting performance targets than focusing on patient safety issues’ (30% ‘disagree’, Reg vs 11% ‘disagree’, FY/CT; χ2=28.485, p<0.001)
‘My suggestions about patient safety would be acted upon if I expressed them to management’ (45% ‘agree’, Reg vs 26% ‘agree’, FY/CT; χ2=19.662, p<0.001)
‘I know the proper channels to direct questions regarding patient safety’ (56% ‘agree’, Reg vs 37% ‘agree’, FY/CT; χ2=19.294, p<0.001)
‘The senior managers in my hospital listen to me and care about my patient safety concerns’ (39% ‘agree’, Reg vs 23% ‘agree, FY/CT; χ2=15.881, p<0.001)
‘The senior doctors in my department listen to me and care about my patient safety concerns’ (76% ‘agree’, Reg vs 63% ‘agree, FY/CT; χ2=9.205, p=0.002)
‘I would feel safe here being treated as an inpatient’ (66% ‘agree’, Reg vs 53% ‘agree’, FY/CT; χ2=8.357, p=0.004).
Group comparisons by speciality
There were no statistically significant differences in responses between the medical and surgical trainee groups for perception of workplace safety climate.
Our study is the first comprehensive evaluation of patient safety from a junior doctor perspective, linking the themes of knowledge and awareness, attitudes to patient safety and perception of the workplace safety climate. Linking these themes is important because patient safety initiatives that target junior doctors need to be based on an understanding of whether there are subtle differences across these different domains. In the few studies that have explored junior doctor attitudes, the sample sizes have been small and specific: the ‘junior doctor’ group is heterogeneous, comprising a wide range of speciality backgrounds, age groups and skill sets. Interventions to improve patient safety education and safety culture among junior doctors need to take into account any differences that may exist across grades and specialities; however, currently there are no studies that provide this level of data to inform the development of such interventions.
It has been noted that the steep professional hierarchy in medicine inhibits the free reporting of experiences of error, rule violation or poor performance by junior doctors to their seniors because of the assumption that it would inhibit career development.10 Rosenthal11 studied ‘problem doctors’ and found mistakes were generally accepted as part of the job, with an expectation that safety issues should be dealt with ‘inhouse’, showing the importance of medical collegiality: this is often central to the attitudes of the medical profession to incident reporting and more broadly to a management driven patient safety agenda.
In terms of attitudes, Sorokin et al12 surveyed 321 American trainees (house staff and medical students); most believed adverse events were preventable (61%) and thought that improved teamwork (88%) and better procedural training (74%) would reduce medical mishaps. Although 60% believed the fear of malpractice litigation inhibited discussion, 80% agreed that doctors must disclose adverse events to patients and grow more comfortable with disclosure as training progresses. The authors concluded that physicians-in-training support many of the systematic changes being proposed for patient safety.
Kroll et al5 investigated the experiences of, and responses to, medical error among junior doctors in a qualitative interview study of 39 preregistration house officers from 2001 to 2002. The study found that junior doctors commonly make and witness errors, some of which are serious, and that there is a prevailing norm of selective disclosure which is likely to limit the systematic reporting of error. Although professional bodies expect concerns for patient safety to take precedence,13 ,14 this is at variance with the medical hierarchy approach to error management. These interviews, however, took place nearly 10 years ago (since then significant changes have occurred in both medical education15 and the patient safety initiatives implemented nationally). However, reports suggest that despite such initiatives there is evidence that very little has changed.16 ,17
In contrast, a more recent survey18 of 109 anaesthetic trainees showed a high rate of respondents admitting their error to patients (68%) and also a high proportion reporting error via clinical incident reporting systems (87%). Flin et al19 used a questionnaire study to evaluate surgical team members’ attitudes to safety and teamwork in the operating theatre, including 93 trainee surgeons. Respondents generally demonstrated positive attitudes to behaviours associated with effective teamwork and safety; however, there were several areas where theatre staff did not appreciate the impact of psychological ‘human’ factors on technical performance.
Our cross-sectional snapshot study supports the more recent studies demonstrating positive attitudes to patient safety from trainees, both in the USA12 and the UK.18 ,19 Self-declared knowledge and awareness of patient safety concepts is high, although there was less understanding of more technical aspects such as the characteristics of ‘high-reliability organisations’ and the concept of ‘active failures and latent conditions’. Junior trainees were significantly less likely to declare an understanding of the role of organisations in error management, suggesting that the organisational aspects of patient safety are less well embedded in patient safety exposure or training for junior doctors. Interestingly, junior trainees were more aware of the ‘Swiss Cheese Model’ of error causation compared with senior ones, suggesting that the basic and key concepts of patient safety are probably starting to mesh into early years training either at medical school or the FYs; patient safety theory is a relatively new concept in relation to postgraduate medical training and it is unsurprising that the more senior trainees are less likely to be aware of such terminology.
The greatest disagreement was demonstrated with the statement ‘Most medical errors result from careless nurses’ and the greatest agreement with the statement ‘Even the most experienced and competent doctors make errors’, providing further evidence of positive patient safety attitudes among the new generation of doctors. Interestingly, when compared by speciality, medical and surgical trainees completely diverged regarding their views about ‘Medical error is a sign of incompetence’, with a significantly greater proportion of surgical trainees agreeing with this statement. Surgical trainees generally also felt that learning about patient safety was not as important as learning about more skill-based aspects of being a doctor, which is unsurprising given the emphasis on technical procedures in their speciality.
The results from the Safety Climate domain suggest a more neutral perception of their workplace environment; the greatest agreement with the statement that ‘senior doctors listen to me and care about my patient safety concerns’ supports a more positive perception of the safety culture. However, across grades, junior trainees were less positive about the role of management; in comparison with senior trainees; a significantly lower proportion of junior trainees agreed that they know the proper channels for directing questions about patient safety, that senior managers listen to them and care about their patient safety concerns and that their suggestions about patient safety would be acted upon if expressed to management. A significantly greater proportion of junior trainees agreed with the statement, ‘Management is more interested in meeting performance targets than focusing on patient safety issues’. This suggests there is room for improvement in the engagement of junior doctors in patient safety by organisations and the clinical governance systems.
The study has a number of a limitations; although the sample size is large, the response rate is borderline acceptable when considering the large cohort of junior doctors across the region that did not respond. As always with such non-responders, the issue of selection bias is present and the positive patient safety attitudes presented here could simply be as a result of the nature of the trainees who made the effort to respond to the questionnaire. The administration of the questionnaire itself could change the knowledge and attitudes of trainees while undertaking and reflecting on many of the items; the questionnaire therefore may be deemed as a positive patient safety intervention. There may also be satisficing bias; although anonymised, some may have been wary of being completely honest about their self-declared knowledge, attitudes or perception of safety climate (possibly explaining the high frequency of responses close to neutral in domain C). The questionnaire has face validity, but formal testing of its psychometric properties will be required to provide further evidence of reliability and validity, particularly if the aim is to produce summary scores indicative of each domain within the multidimensional scale. This will be addressed by a future study and the validation of the tool's psychometric properties will enable us to use this as a metric for robustly evaluating new interventions designed to improve junior doctor engagement in patient safety. Ultimately, a validated metric such as this could also be used to evaluate the cost-effectiveness of such interventions.
The work provides the first comprehensive insight into the current attitudes of junior doctors training across the East Midlands region and a key strength of the study is the range of specialities and care settings to which the respondents belong. Overall, this study demonstrates a positive attitude towards patient safety among junior doctors although any future interventions to improve patient safety education and culture among junior doctors will need to take into account the subtle differences across grades and specialities.
Contributors PD, JD and HS contributed to the conception and design of the study. PD and HS collected the data; PD and NT contributed to the analysis and interpretation of data. PD prepared the drafts and revisions of the article, JD, HS and NT provided further input on content and presentation of work and all approved the final version published.
Funding This research received no specific funding.
Competing interests We declare that we have no significant competing financial, professional or personal interests that might have influenced the performance or presentation of the work described in this manuscript.
Provenance and peer review Not commissioned; externally peer reviewed.