Applications
| Title | Summary | Lead Contact |
|---|---|---|
|
UKMEDP001
Predicting Fitness to Practise issues from admission profiles in UK medical school entrants Approved on 15 October 2015 UKMED Advisory Board Published in BMC Medical Education in April 2018 as "Predictors of fitness to practise declarations in UK medical undergraduates" |
The selection of medical students is known to be a complex and ‘evidence light’ area with a wide variety of approaches exercised. Although it is established that cognitive and educational performance tends to predict academic performance, especially in the early years of medical school, less is known about the role of personal attributes and the predictors of later problems that may impair Fitness to Practise (FtP).
In recent years the GMC has required significant FtP concerns to be reported prior to provisional registration of UK graduates and are now over 300 per year are now considered, with rejection an occasional outcome (0-2/year). This is clearly problematic for the few individuals concerned but there is also a suspicion that those allowed to graduate may be ‘at risk’ of future professional behavioural issues arising. This is largely based on case-control studies conducted by Papadakis (2004) in the USA, who found medical school concerns were associated with future misconduct. A more recent paper by Norman(2015) has questioned the feasibility of this approach.
The holy grail of selection is avoiding recruiting ‘problem students’ in the first place but there is no evidence that could be achieved, let alone justified. The advent of the UK Medical Education Database (UKMED) and the possibility of matching a wide range of pre-admissions metrics with outcome markers such as progression through medical school and provisional registration data offers a first opportunity to explore the issue from a modelling perspective and inform selection practice.
Graduate cohorts from 2013 and 2014 offer over 12 000 individuals with matched data and approximately 800 reported FtP concerns recorded by the GMC. Around 8 000 of these should include a range of non-academic measures piloted within the UKCAT test in 2007 offering a unique cohort for study and a sufficient sample to merit exploratory analysis.
It is recognized that this data is highly sensitive and that great care would be required to ensure individuals could not be identified. This may require some blunting of the available data, such as categorising the most severe offences broadly enough to avoid highlighting the 0-5 excluded within these cohorts too clearly. However, it is suggested that, in outline, this can be achieved using the approach described and therefore UKMED has an opportunity and indeed responsibility to conduct pioneering research on such a critical area of selection practice. There is also the opportunity to follow these cohorts on into practice and to pilot methodologies for a prospective cohort study to that may include performance measures such as ARCP progression and FtP events.
References
|
Doctor Paul Tiffin
p.a.tiffin@dur.ac.uk |
|
UKMEDP002
What has been the impact of accelerated graduate-entry medicine courses in terms of educational and sociodemographic profile, success at medical school, completion of Foundation training, and specialty entry? Approved on 15 October 2015 UKMED Advisory Board Published in BMC Medical Education in November 2018 as "Impact of accelerated, graduate-entry medicine courses: a comparison of profile, success, and specialty destination between graduate entrants to accelerated or standard medicine courses in UK" |
A proportion of each annual intake to medical school already have a first degree in a different subject. Prior to 2000 these graduate entrants studied alongside school leavers in the existing UK five or six year medicine courses. Since then, around 800 graduates annually have entered the fifteen UK graduate-entry 4-year accelerated programmes as well as a smaller number who have continued to join the 5/6 year programmes. The profile of graduate-entrants has also been markedly different from undergraduate programmes in terms of age and subject background: nine of the graduate entry courses admitting students with degrees in non-science subjects.
Older research reported (e.g. James & Chilvers, 2001; Wilkinson et al, 2004) that graduates had a number of advantages in terms of attainment and progress at medical school compared to younger entrants with only secondary educational qualifications. However, it was unclear what exactly might be responsible for these advantages. More recent research (McManus et al, 2013) analysed the year 1 performance of entrants to twelve UK medical schools in terms of potential predictors of their attainment. In summary, that study confirmed the strong relationship between A-level performance and year 1 medical school assessments, but it also demonstrated weaker, but incremental predictive validity for a number of other pre-entry variables: these included GCSE performance, scores on UKCAT, and age>21 (who were most likely graduates). Demographic variables were also influential: in particular, men and students from non-white UK ethnic minority communities performed more poorly.
Several studies of attainment at individual UK medical schools have shown that graduate-entry students have performed comparably (Manning & Garrud, 2009) or better (Price & Wright, 2010) than undergraduate students in common assessments during the shared full-time clinical phase of those programmes. Some studies (e.g. Bodger et al, 2011) have attempted to identify predictors of attainment in graduate-entry programmes, with mixed conclusions, but commonly that prior academic record (e.g. secondary or tertiary educational qualifications) is a reliable predictor.
At present, therefore, there is no evidence about the relative success of graduates who have gone through the graduate-entry vs. the undergraduate medicine courses. There is also very little evidence at a national, pan-individual school level, about markers of success in these different types of course for those students with a prior degree. Two key questions concern the subject of that prior degree and its class or grade. Earlier, secondary educational record may also be an important factor in success. Age, gender, socioeconomic status, and ethnicity may also be relevant factors.
References
|
Professor Chris McManus
i.mcmanus@ucl.ac.uk |
|
UKMEDP003
Do the Educational Performance Measure decile score and SJT predict successful completion of the foundation programme? Approved on 15 October 2015 UKMED Advisory Board Published in BMJ Open in July 2018 as "Evaluating the validity of the selection measures used for the UK’s foundation medical training programme: a national cohort study" |
Summary: Please outline your proposed research.
Work Psychology Group published a report in February 2015 on the Validation of the F1 Selection Tools . The study was limited to 391 F1s and therefore did not have data on which trainees successfully completed the foundation programme. Their sample specifically targeted F1 doctors who had received particularly high or particularly low SJT scores. They make the following recommendation: “that further studies are undertaken to explore the relationship between performance at application and performance outcomes beyond F1 (for example at the end of F2 and into specialty training) and that application scores (particularly SJT scores) spanning the full range of scores are targeted. If the relationship between application scores and ARCP outcomes is to be examined further, a large population (ideally all schools) should be targeted, as incidences of unsatisfactory ARCP outcomes appear to be very rare (1.1% in the present sample).”
We will seek to measure the predictive validity of the Educational Performance Measure (EPM) decile score (note in 2012 the EPM score was in quartiles) and the Situation Judgement Test (SJT) , collected as part of medical students’ applications to the foundation programme captured on the Foundation Programme Application System (FPAS) using ARCPs from the two-year foundation programme as an outcome measure. The EPM deciles are a medical school performance score calculated by the applicant’s medical school based on performance in a number of assessments and divided into 10 equal groups (deciles) within the given medical school and not UK-wide. Each UK medical school has agreed with its students which assessments will be included in this measure.
We will also seek to understand the relationship between the EPM decile score and SJT scores and specialty recruitment (applications to specialty training made by F2 doctors) . Specifically we will investigate: whether the EPM decile score and SJT relate to which specialties F2 doctors apply to and whether offers of a place on the training programme are made.
The output of this research will allow us to fulfil our duties outlined in The Trainee Doctor
“10 Periodically, the GMC will analyse evidence from these sources to draw together a picture of the state of foundation and specialty training throughout the UK. This will show performance against standards by postgraduate deaneries, LEPs, medical Royal Colleges and Faculties and specialty associations and will seek to show which factors are most significant in predicting good and poor educational outcomes within training programmes and at the end of training “
It will also assist Medical Schools in meeting the following standard:
“172 Quality management will involve the collection and use of information about the progression of students. It will also involve the collection and use of information about the subsequent progression of graduates in relation to the Foundation Programme and postgraduate training, and in respect of any determinations by the GMC.” From Tomorrow’s Doctors (2009)
Note that from January 2016, these standards will be superseded by Promoting excellence: standards for medical education and training , the relevant requirement will be
“R2.5 Organisations must evaluate information about learners’ performance, progression and outcomes – such as the results of exams and assessments – by collecting, analysing and using data on quality and on equality and diversity”
The output of the analysis will help us begin to explore whether the EPM deciles and the SJT score are predictive of trainee performance as measured by ARCP and recruitment outcomes during the foundation programme and therefore whether it is appropriate to split medical school cohorts into groups based on EPM deciles and or SJT to report on their progress at foundation ARCP and at recruitment into specialty training from F2 in order to provide more granular reports on outcomes to medical school deans. We will use the output to enhance the existing medical school progression reports. This could, if the analysis shows that it would be appropriate, allow a medical dean to see the ARCP and recruitment outcomes of their top performing graduates versus their less well performing graduates; where the definition of these groups is determined by this proposed analysis.
We may find that ARCP outcomes for foundation are not suitably granular for establishing whether EPM and SJT are useful predictors of which trainees will struggle in foundation training. To establish whether the SJT and EPM predict which trainees required additional support from their foundation schools would require UKMED to have identifiable data on doctors in difficulty on the foundation programme. These may be doctors who received support in order that they were able to achieve an outcome 1 or an outcome 6 and therefore may not have been awarded an unsatisfactory outcome. The UKFPO Annual report for 2014 says there were 186 F1 and 163 F2s from UKMED medical school monitored via foundation schools’ doctors in difficulty (DiD) policies and processes. We do not know which ARCP outcomes these DiD s were finally awarded.
The output will also allow comparison of the ARCP and the recruitment outcomes across foundation schools adjusting for the prior attainment of the trainees (i.e. their performance on entry to the foundation programme). We know from the FPAS application handbook that higher ranking applicants get their first choice of school. So viewing outcomes by foundation school with no adjustment may reflect the given school’s intake more than the training it offers.
In addition the output will inform our work on differential attainment. We will be able to see whether ethnicity and gender adds to the prediction of ARCP and recruitment outcomes, after accounting for prior attainment as measured by the EPM decile score and SJT in the model. We will seek to understand the variation in EPM decile score and SJT by ethnicity, gender and socio-economic status.
Finally we will seek to understand whether the EPM decile score and SJT relate to trainee’s own rating of their level of preparedness as captured on the NTS 2015. This will inform our work on the construct validity of the NTS preparedness indicator.
The outputs of this work could be used to help the GMC decide how the SJT and EPM decile scores might contribute to a UK licensing assessment.
The analysis might help the UKPFO consider the best way to rank students for entry to the foundation.
There are limitations to this study; the UKMED phase 1 cohort does not include graduates from non-UK medical schools who do apply to Foundation training using the FPAS system. In future years UKMED may include those non-UK graduates from the point they apply to foundation training. Note these cases will never have data from the Higher Education Statistics Agency. Any conclusions will therefore be limited to UK graduates.
Note that SJT data are only available from 2013 onwards; in 2012 SJT was piloted for the first time and the 2012 parallel recruitment exercise data are not included within UKMED. The 2012 parallel recruitment exercise is described in the Medical School Council report.
Those who entered foundation in 2013 will now have a complete set of ARCP outcomes for the two years of their foundation programme so a complete cohort study is possible.
References
|
Mr Daniel Smith
dsmith@gmc-uk.org |
|
UKMEDP020
The role of academic attainment in understanding sex differences in specialty choice and fitness to practise. Approved on 15 June 2016 UKMED Advisory Board Published in BMJ Open in March 2019 as "Effect of sex on specialty training application outcomes: a longitudinal administrative data study of UK medical graduates" |
The proportions of men and women in different medical specialties varies greatly [1].
Understanding how and why is important for effective workforce planning and the provision of
future healthcare, and to reduce sex segregation in some specialties. It may also help us
understand other areas of stark sex differences, such as disciplinary action, where male
doctors have nearly 2.5 times the odds of facing medico-legal action [2], and doctors from
certain specialties are at higher risk of receiving sanctions [3].
Sex differences in specialty choice are partly explained by features such as how plannable,
technical, and intellectual a specialty is [1]; but success in obtaining a training place depends
on competition ratios, selection methods, and candidates’ previous academic attainment – also
potentially associated with sex. Academic performance is also important to help us understand
how sanctions relate to sex and specialty, because past academic performance predicts future
academic performance [4], and poor academic performance is associated with increased odds
of sanctions [5].
References
|
Doctor Emily Unwin
emily.unwin.12@ucl.ac.uk |
|
UKMEDP022
The allocation of doctors to specialty and general practice training posts by demographic and socio-economic characteristics Approved on 15 June 2016 UKMED Advisory Board Published in UKMED in July 2019 as "A sequential analysis of the specialty allocation process in the UK. Empirical evidence from the UKMED database." |
Becoming a medical practitioner is a competitive and complex process and its outcome determines the composition of the medical profession. There is a growing concern that the profession should reflect both the appropriate skills and a balance of social, economic, gender and ethnicity (GMC 2010)
The distribution of these characteristics is highly unequal both across medical specialties and between specialties and general practice (Goldacre, Laxton et al. 2010 and Rodríguez Santana,& Chalkley 2015). There is a persistent gender gap in certain specialties (men in surgery, women in general practice), and underrepresentation of those from deprived socioeconomic backgrounds (Arumpalam, Naylor et al. 2005) in highly competitive specialties.
We propose to analyse the outcome of the specialty allocation, recognising that it is a sequential process. Junior doctors’ preferences over the different training posts and their personal characteristics, qualifications and environment influence their applications; these are then assessed to determine the suitability of applicants to each training post and finally doctors’ decide which of the offers they have received to accept.
Our principle objective is to understand how demographic and socio-economic characteristics impact on each stage of this process; how do an individual’s characteristics correspond to their decision to apply, to their subsequent assessment and to their decisions to accept offers. Such an understanding is vital for the formulation effective strategies to ensure greater representativeness across specialities and general practice. Central to our approach is controlling for other factors, such as previous educational experience and attainment to establish the effect of demographic and socio-economic characteristics "other things equal".
References
|
Professor Martin Chalkley
martin.chalkley@york.ac.uk |
|
UKMEDP026
“Getting on” in medicine: a programme of study of careers trajectories and decisions of doctors Approved on 25 February 2016 UKMED Advisory Board Published in BMJ Open in September 2017 as "The relationship between school type and academic performance at medical school: a national, multi-cohort study" Published in BMJ Open in June 2018 as "Relationship between sociodemographic factors and selection into UK postgraduate medical training programmes: a national cohort study" Published in BMC Medical Education in December 2018 as "Geographical mobility of UK trainee doctors, from family home to first job: a national cohort study" Published in BMJ Open in March 2019 as "Relationship between sociodemographic factors and specialty destination of UK trainee doctors: a national cohort study" |
Millburn's "Fair Access to Professional Careers" highlighted that the increasing diversity of the UK medical student body in terms of age, ethnicity and gender is not reflected in terms of student socio-economic (SEC) background. Cleland et al's (2012) subsequent work identified that research is lacking on how widening access (WA) students and doctors progress in medicine. Only one recent UK study does so. Dowell et al. (2015), in a survey of 2050 Scottish GPs, found that those whose parents had semi-routine or routine occupations were more likely to be working in a deprived practice than those from professional families. While this suggests that SEC on entry to medical school may be associated with differences in career pathway, the study looked only at a sub-sample of doctors who were established in their career choice. In contrast, we are interested in "tomorrow's doctors", the generation who are currently deciding what (specialty) and where (location) they wish to work - or, indeed, when or if they wish to work as a doctor after the Foundation Programme. Are their careers decisions influenced by demographic factors such as SEC only or is there a complex relationship between these and other factors (e.g., medical school) given the medical training pathway is competitive (see MacKenzie et al. in press)? We must understand these relationships to identify how to address barriers to progression within medicine and to inform policy decisions about workforce planning. WA students are few and only UKMED provides sufficient data to address these questions appropriately.
References
|
Professor Jennifer Cleland
jen.cleland@abdn.ac.uk |
|
UKMEDP030
What demographic and educational factors predict doctors' decisions to apply for training programmes in particular medical specialties? Approved on 15 June 2016 UKMED Advisory Board Published in BMC Medicine in December 2017 as "Factors associated with junior doctors’decisions to apply for general practice training programmes in the UK: secondary analysis of data from the UKMED project" Published in BJPsych Bulletin in May 2019 as "Sociodemographic and educational characteristics of doctors applying for psychiatry training in the UK: Secondary analysis of data from the UK Medical Education Database project." |
Decisions to apply for postgraduate training programmes in particular medical specialties are shaped by a multitude of factors. These include the doctor's family background, schooling, undergraduate (medical school) training, early postgraduate professional experience, and the perceived likelihood of a successful application. Much survey-based research has been done into factors that are associated with the self-reported (intended) career preferences of both medical students and early-career doctors [1-3]. Intentions to follow a particular career pathway however are subject to change [4] and may not, at the ultimate decision point, materialise as actual applications to particular specialty training programmes.
The UKMED database provides a unique and valuable opportunity to add to the literature on career choice by permitting analysis, for a large cohort of doctors and a range of nationally-recruited UK specialty training programmes, of some of the important factors that are associated with the decision to apply for a place on those programmes.
Research in this area has particular importance for workforce planning. The most recent data on application ratios suggests that certain specialties are experiencing recruitment difficulties; from 2013 to 2015 General Practice, Paediatrics and Psychiatry consistently ranked within the five least competitive specialisms [5]. This proposed research may therefore influence changes to policy and practice in areas such as the provision of specialty training programmes, the design of undergraduate medical school courses, methods of recruitment and selection to medical school and the widening participation agenda, in order to encourage a better match between graduate career choice and service need.
References
|
Doctor Tom Gale
thomas.gale@plymouth.ac.uk |
|
UKMEDP032
What factors predict doctors' successful completion of core training in medicine and anaesthetics and their subsequent decisions to pursue higher specialty training Approved on 27 March 2017 UKMED Advisory Board |
Core training programmes for medicine and anaesthesia have high fill rates for CT1 entry compared to other specialties. However, these specialties suffer from below optimal conversion rates between core and higher specialty training posts.[1] As a result, there are many unfilled posts at entry to higher specialty training in medicine and anaesthetics. The CfWI has identified an urgent need to increase the number of ST3 posts, and to model the way in which the output from core training posts and ACCS flows into higher specialty training.[2,3] Research is required to understand the extent of attrition between core training and specialty training posts in these specialties, and factors that predict successful appointment to higher specialty training.
The main aim of our study is to identify factors, which predict doctors' successful completion of core training in medicine and anaesthesia, and their subsequent decisions to pursue higher specialty training. Previous work has investigated trainees’ perceptions regarding the weighting of individual and job-related factors influencing choice and selection to specialty training posts, but there is limited longitudinal research investigating factors which predict successful completion of training.[4] A large longitudinal prospective study, identified that previous academic attainment predicts undergraduate attainment in pre-clinical and clinical years of a medical degree, but socio-demographic factors are also important predictors of future clinical performance.[5]
The UKMED database provides a unique opportunity to investigate the contribution of a number of factors that predict successful completion of core training in medicine and anaesthesia, and successful progression to higher specialty training.
References
|
Doctor Tom Gale
thomas.gale@plymouth.ac.uk |
|
UKMEDP038
How do the professional outcomes of medical graduates from gateway courses compare to graduates from standard entry medicine courses? Approved on 28 March 2017 UKMED Advisory Board Published in BMC Medical Education in January 2020 as "A comparison of undergraduate outcomes for students from gateway courses and standard entry medicine courses" |
The postgraduate outcomes of those on 6-year medicine with a gateway courses will be compared to those on the 5 year standard entry courses after accounting for attainment and aptitude measured on entry and the students’ performance relative to their year within the school on exit, used to rank applicants to the foundation programme. UKMED Data are available for three schools that offered both types of course. Outcomes will be measured in terms of: progression through training as captured by ARCPs, performance in royal college medical exams and the specialty training programmes applied to and offered at CT1/ST1.
References
|
Mr Daniel Smith
Daniel.smith@gmc-uk.org |
|
UKMEDP039
What factors lead to success in obtaining an ophthalmology specialty training (OST) post and completing postgraduate ophthalmology examinations? Approved on 12 June 2017 UKMED Advisory Board |
Run-through surgical specialty training posts lead directly to a certificate of completion of training, making them attractive, but competitive career choices among doctors (1). Ophthalmology specialty training (OST) is a highly subscribed run-through post with competition ratios higher than the overall average for all specialty-training posts (2).
We aim to explore which factors lead to successful applications to ophthalmology training on the first attempt. We also aim to understand what influences success in ophthalmology postgraduate examinations on the first attempt.
The Royal College of Ophthalmologists’ (RCOphth) endorses this study. The RCOphth oversees a rigorous recruitment process but has little evidence to guide them on the predictive validity of their recruitment measures (personal communication). Understanding what makes their candidates successful in gaining a training post is a research priority. We aim to ascertain whether interview and shortlisting scores predict exam performance and recruitment into ophthalmology specialist training on the first attempt.
Our study explores a range of demographic, socioeconomic (4) and academic variables, which are hypothesised to influence success in attaining an Ophthalmology training post, as well as success in passing post-graduate exams on the first attempt. Previous studies have demonstrated the predictive validity of prior educational attainment (PEA) and UKCAT scores in predicting success in medicine (5).
Our project aims are twofold; 1) To enable RCOphth and other postgraduate colleges to operate an evidence-based recruitment process and 2) To guide applicants of the factors that may influence success in ophthalmology so that they are able to make realistic career choices.
References
|
Doctor Aditi Das
aditi.das@moorfields.nhs.uk |
|
UKMEDP041
Development of a UKMED multidimensional measure of widening access status. Approved on 25 September 2017 UKMED Advisory Board |
Contextual admissions data are increasingly employed in selection to the study of medicine in the UK, albeit with little knowledge about the quality of the indicators or the implications of their use on both widening access (WA) and student achievement. [1,2,3] Moreover, there is concern about the validity of contextualised admissions decision making because; contextual data have been shown to produce conflicting information on WA status, doubtful veracity of self-reported information and the extent of missing data values on contextual indicators generally [4] [5] Furthermore, little is known about the association between students’ contextual background characteristics and performance in medical school.[6, 7, 8] Triangulation of contextual indicators to identify those most likely to be disadvantaged is recommended to reduce numbers of false positive WA status identification but, may also serve to increase the number of false negative and introduce new injustices.
Contextual data comprise disparate measures of disadvantage which each capture an aspect of the underlying concept WA status. It is therefore important to know about the strengths and limitations of the most commonly used contextual indicators, singly and in combination. It is also desirable to efficiently combine the most reliable contextual indicators into a single multidimensional measure of WA status which UK medical schools can confidently use in their selection processes.
The UK Medical Education Database (UKMED) includes a range of contextual admissions indicators commonly used in selection to the study of medicine and provides a unique opportunity to achieve this study’s aims, the outcome of which has the potential to make WA to medicine fair, transparent and above all, evidence-based.
References
|
Doctor Paul Lambe
paul.lambe@plymouth.ac.uk |
|
UKMEDP042
Understanding variation in BME medical exam performance across the UK Approved on 26 September 2017 UKMED Advisory Board |
The landscape of exam result differential performance for Black, Asian and Minority Ethnic (BAME) individuals in the medical education system is complex. Performance data from undergraduate and postgraduate settings demonstrate lower success rates in doctors of BAME background. The recent “Fair Training for All” report by the General Medical Council used a qualitative approach to understand the barriers and facilitators to success in BAME doctors. One of the key findings of the report was the negative impact of poor performance in exams: poorer performance in exams adversely affected autonomy in job choice, increased likelihood of being separated from family and support networks, and increased chance of mental health problems. Failing exams can lower confidence, and resits can be felt to interfere with workplace learning.
There is relatively little published evidence on the BAME attainment gap, especially when considering local variation. Learning from the NHS clinical arena, identifying regional and local variation in performance is valuable as it (1) highlights exemplars of best practice and (2) helps allocate effort and resource to areas with poorer clinical outcomes. Measuring and publishing unit level NHS performance data is the central tenet of the Health Care Quality Commission.
Our planned analysis will use the NHS approach, and aim to identify medical school variation in BAME exam success, identifying where the attainment gap becomes more pronounced and may warrant further explanatory investigation. Central to this project will be to implement appropriate statistical methods to ensure accurate definition of medical school variation. In particular, the concept of ‘case-mix’ from clinical practice is relevant; understanding the characteristics of individuals within each medical school cohort prior to commencing study (e.g. prior academic attainment, socio-demographic background) is essential to interpreting any inter-school variance in attainment gap.
References
|
Doctor James Galloway
james.galloway@kcl.ac.uk |
|
UKMEDP043
Does medical school entry performance and medical school performance predict success on the Intercollegiate Membership of The Royal College of Surgeons (MRCS) exam ? Approved on 25 September 2017 UKMED Advisory Board |
In 2008, McManus and colleagues published an article in BMC Medicine highlighting the substantial difference in performance on the MRCP across medical schools and encouraged other groups to investigate whether similar patterns exist in other postgraduate UK examinations. This study reported data from those medical graduates taking PACES between 1989 and 2005. These cohorts entered medical school in the years before school examination grade inflation and before the widespread implementation of multi-method selection into medical school (e.g., the use of selection tests such as UKCAT, BMAT or GAMSAT). Moreover, the McManus study was also prior to changes in assessment during medical school, notably the introduction of ranking using the educational performance measurement (EPM), which is then used in selection into the next stage of medical training, the UK Foundation Programme. To the best of our knowledge, no previous studies have looked at the relationship between performance on medical school entry tests and performance on postgraduate examinations although UKCAT has been shown to be predictive of medical school final outcome (EPM). Nor has there been prior research looking at performance at medical school and performance on postgraduate examinations. The development of the UKMED database enables such studies. Our specific interest is surgical selection and training. The Intercollegiate Membership of The Royal College of Surgeons (MRCS) is a mandatory postgraduate exam for all aspiring UK surgeons wishing to apply for higher surgical training (ST3 selection). It is attempted by upwards of 6000 UK and overseas doctors annually. We wish to investigate whether medical school, medical school entry performance and medical school performance predict MRCS success.
References
|
Professor Jennifer Cleland
jen.cleland@abdn.ac.uk |
|
UKMEDP044
From family-home to education and from education to training: the spatial patterns of future doctors. Approved on 25 September 2017 UKMED Advisory Board |
This project aims to discover and analyse the spatial patterns in the movements of the future doctors: from their family-home to University and from University to training locations. This research is new in two aspects: we propose a robust and spatio-temporal quantitative statistical framework, while the majority of previous research is mainly qualitative; we take into account the starting point of each student pathway (for most of them their family-home postcode, although this may not be true for older entrants) and their geographic trajectories (movements to different post-codes representing education, training and work).
This project will focus on the following analysis:
1. Movements from ‘home’ (prior to entering medical studies) through to working as a consultant/GP. The end point will only be possible for those who have obtained their Certificate of Completion of Training (CCT). The majority of these will be GPs as GP training is shorter than the training time for other specialties. (It would be possible to extend the data analysis over time as more trainees within the data base complete a CCT)
Currently we have 6,832 doctors in the HESA data now working as GPs
If additional resources are available (e.g. more MSc students interested or by applying for funding a PhD), the following analyses (ordered for importance) will be carried out:
2. Home to medical school- full cohort. HESA data include all cases starting at a UK Medical school from 2002 to 2015.
3. Medical school to foundation – only those who have graduated by 2016 so those starting in 2011 and earlier.
4. Foundation to specialty training only those who entered foundation from 2012 (the first year UKFPO foundation application data are available).
The data available for each of these analyses will overlap but may not consist of precisely the same cases, therefore they may require different individual analysis.
This will represent a significant improvement compared to previous research based on qualitative methodologies (interviews), which focused on a restricted temporal window and did not account fully for the original residence of the student (often the family home).
References
|
Daniel Smith Analyst
daniel.smith@gmc-uk.org |
|
UKMEDP046
Are there differences between those doctors who apply for a training post in FY2 and those who take time out of training? Approved on 25 September 2017 UKMED Advisory Board Published in BMJ Open in November 2019 as "Are there differences between those doctors who apply for a training post in Foundation Year 2 and those who take time out of the training pathway? A UK multicohort study" |
Accurately predicting medical workforce supply is increasingly challenging as doctors no longer behave in time-recognised ways in terms of career decision making (Arthur et al., 2005; Cleland et al., 2016, 2017). For example, in the UK context, medical graduates are choosing not to progress through training as predicted. In 2016, nearly 50% of those graduates completing the Foundation Programme did not apply for specialty/GP training at the expected point in time (UKFPO Career Destination Reports 2015, 2016; Scanlan et al., 2017). Simply put, one in two of today’s medical graduates left the training pipeline at the first opportunity to do so in order while keeping their options open (i.e. with full registration and eligibility to apply for higher training). Instead, they opted to take a break from training. The percentage working overseas for a period of time has remained static, the number taking other types of NHS service posts including Development Fellow posts has increased as has those who decided to leave clinical practice. Given this “brain drain”, more understanding of the differences and similarities between those doctors who progress directly from FY2 into training, and those who take time out of training is crucial as this can inform policy and practice in relation to medical selection and attracting trainees to medical training across the breadth of specialties, and thus ensure sufficient doctors to deliver service now and in the future (Collins & Young, 2000; Gorman, 2017).
References
|
Mr Benard Kumwenda
r0bk15@abdn.ac.uk |
|
UKMEDP051
A comparison of the properties of BMAT, GAMSAT and UKCAT Approved on 25 September 2017 UKMED Advisory Board |
Currently the UK Clinical Aptitude Test (UKCAT) and the Biomedical Admissions Test (BMAT) are the two main aptitude tests used for selection into standard medical and dental undergraduate courses in the UK. Previously, both tests have been shown to significantly predict undergraduate performance in medical students (1, 2). However, to date, no direct comparison has been made between the two assessments, in terms of their ability to predict important outcomes, their sensitivity to sociodemographic variables and the degree to which they incrementally add value within the selection system, above and beyond that provided by conventional measures of academic attainment.
The Graduate Medical School Admissions Test (GAMSAT) is used for selection into some graduate entry medical courses. There is some weak evidence for the predictive validity of the GAMSAT in the early years of medical school (3, 4). However, no large studies of the predictive validity of GAMSAT in the UK have been conducted. It has also been noted (4) that the GAMSAT appears to possess more predictive performance than the UKCAT when predicting Educational Performance Measure (EPM), but as this was not the primary research aim, this issue was not explored in depth.
Comparisons of the properties of the three tests would allow the strengths and relative weaknesses of each of the tests to be evaluated, as well as their potential to widen (or narrow) participation in medicine. Thus, the results will provide selectors with an informed choice.
References
|
Doctor Paul Tiffin
pat512@york.ac.uk |
|
UKMEDP054
Declared disability in the UKMED dataset 2002-2016: an exploratory descriptive analysis Approved on 25 September 2017 UKMED Advisory Board |
Disability is an important consideration in the selection of medical students, in undergraduate and postgraduate medical training, and in the practice of medicine. First, adjustments may be required in order to accommodate disabled individuals. Second, disability may affect the fitness of individuals to practise medicine and/or their choice of postgraduate specialty. Third, disability may affect academic performance. Fourth, disability is the subject of legislation (1) which is a key imperative in decisions about disability and how to deal with it. Finally, and related to the previous point, equity with respect to disability and selection is important at both undergraduate and postgraduate level.
The General Medical Council (GMC) recognises the importance of disability, providing guidance to medical schools (2). Likewise the Medical Schools Council (MSC) provides advice on how to adjust for disability in the selection of medical students (3). A 2007 report by the British Medical Association (BMA) focused on inequality for disabled doctors and medical students (4). The focus of these documents appropriately reflects the key imperative of relevant legislation, and they provide useful, practical advice on what can be done to accommodate disabled colleagues.
Many important questions about disability nevertheless remain unanswered; some of these are outlined in the Research section below. The advent and linkage of the UKMED databases provides the framework for these and other questions to be explored in far more detail than ever before.
References
|
Doctor Michael Murphy
m.j.murphy@dundee.ac.uk |
|
UKMEDP057
Selection Tests as a predictor of acceptance rate and post-graduate success of widening participation students in undergraduate medicine Approved on 05 November 2017 UKMED Advisory Board |
Students from non-traditional backgrounds are underrepresented in UK medical schools. Those that successfully receive a place are also at the greatest risk of non-completion. The impact of specific admissions processes on widening participation (WP) and their ability to select students who will be successful are unclear.
We will compare the outcomes of WP medical students across different Medical Schools. Specifically, the study will focus on the use of the Biomedical Admissions Test (BMAT) and UK Clinical Aptitude Test (UKCAT) to determine:
1) The relative proportions of WP students at medical schools using the BMAT or UKCAT as part of their selection criteria.
2) Whether WP students attending BMAT-selecting Medical School perform significantly differently (including degree outcomes and drop-out rates) compared to their non-WP peers and compared to WP/Non-WP students in UKCAT-selecting Medical Schools.
3) The effect of WP status and admissions tests on the proportion of local WP students that they attract.
4) The effect of WP status and admissions tests on the employment of medical graduates; particularly the geographic location of trainee positions and whether they receive their first choice placement.
This will provide evidence regarding the utility of BMAT and UKCAT as determinants of success for students from WP backgrounds, and to identify possible relationships between WP status and admissions tests, propensity to attend a local medical school and propensity to return to or remain in a home region following graduation. Understanding these relationships will allow us to make recommendations regarding admissions processes to optimise WP student success.
References
|
Professor Kevin Murphy
k.g.murphy@imperial.ac.uk |
|
UKMEDP058
Modelling the consultant physician workforce Approved on 24 February 2018 UKMED Advisory Board |
Medical workforce issues are among the most crucial problems facing the NHS (Ref 1 and 2). Some factors contributing to this are:
- Fewer international doctors (possibly related to immigration controls and Brexit)
- Increased less than full time working within the medical workforce (possibly related to feminisation of the workforce) and
- New junior doctors' contract dispute leading to recent industrial action and the exposure of system wide non contractual issues within the trainee workforce linked to low morale and inflexibility of training programmes.
The government has pledged to increase medical school places by 1500 beginning September 2018 to attempt to address this complex issue.
We plan to construct a Markov mathematical model to model future progression from medical school entry to becoming a consultant physician using UKMED data to provide the numbers moving between different training stages up to higher specialty training (HST). JRCPTB data will be used for progression from HST level.
Sensitivity analysis will facilitate targeting of future interventions to increase the number of consultant physicians available in general and in frontline specialities such as acute medicine and geriatric medicine.
References
|
Doctor Johnny Boylan
ohn.boylan@rcplondon.ac.uk |
|
UKMEDP072
Factors associated with working as a locum Approved on 25 September 2017 UKMED Advisory Board |
Little is known about the numbers of doctors who leave training and become locums. This study seeks to establish the proportion of Drs who have worked in the UK as an F2 Dr between 2012 and 2016 and subsequently worked as a locum. We will consider those who worked as locum where there is evidence of failure to progress in training as captured by ARCP and recruitment data and those who worked as locum where there is no evidence of failure to progress in training. We will examine whether working as locum with no previous evidence of failure to progress in training is associated with demographic factors, geographic factors, specialty factors, previous academic performance, working or wanting to work less than full-time (LTFT) and the workload indicator from the final NTS submission prior to working as a locum.
References
|
Mr Daniel Smith
daniel.smith@gmc-uk.org |
|
UKMEDP073
Recruitment of trainees to obstetrics and gynaecology training programmes. Approved on 04 March 2018 UKMED Advisory Board |
Recruitment to Obstetrics and Gynaecology has shown a large fall in expression of interest in the specialty over time. This has become a significant issue, as more trainees are needed to cover a high-risk speciality such as O&G. Recruitment to O&G varies between medical schools, as does the amount of time spent doing O&G within the curriculum. Factors thought to influence career choice also includes experience as a student. In addition, by the third year after qualification, only 46% of those whose only choice was previously obstetrics and gynaecology were still committed to this career.
This research proposal will examine which factors predict recruitment to obstetrics and gynaecology training at undergraduate level and in the early years after qualification. We will consider the following factors:
1. medical school and course type and course level/school level data measuring timing and amount of teaching/exposure students had to obstetrics and gynaecology
2. socio-demographic variables
3. academic attainment on entry and exit from medical school
This research proposal will help understand which factors affect recruitment to the specialty and help approach likely challenges ahead. In particular, the aim is to enable exchange of information between medical schools and support areas of development of undergraduate O&G medical education in medical schools where recruitment is low and in addition to aid revision of the current RCOG undergraduate curriculum. Also the information will be used to allow resources to be targeted at areas in the early postgraduate years, which may increase recruitment.
References
|
Doctor Philippa Marsden
Philippa.marsden@newcastle.ac.uk |
|
UKMEDP077
The relationship between medical student Conscientiousness Index scores and later clinical performance: a pilot study Approved on 22 May 2018 UKMED Advisory Board |
Performance by medical and other healthcare students while in education may provide a guide to their later clinical practice. Demonstrating that a metric has Predictive Validity would therefore allow targeted training to be directed to individuals who may cause concern during their education or permit appropriate selection.
It is known that cognitive ability has significant Predictive Validity for later clinical practice in a number of settings. However, there are currently no demonstrated measures of personality related performance with such Validity.
Our project involves data captured as part of the Conscientiousness Index (CI) project(1). Data on students’ performance of routine tasks such as attendance and submission of assignments, was assembled to form a single score, known as the Conscientiousness Index. Positive, statistically significant correlations were observed with outcome performance such as examinations scores and independent staff ratings of professionalism. The process has since been repeated in other health care settings(2,3).
The CI Scores will represent the Predictor Variable and we wish to explore the relationship with this data with several Outcome Variables as contained in the UKMED database. In order to allow us to draw any conclusion on the predictive validity of the CI for future performance as a doctor, two principal methods of analysis will be used in this pilot: linear regression and data dichotomisation into high and low scoring individuals.
The long term benefit is to healthcare in the UK in particular, and potentially worldwide, in establishing the Predictive Validity of a personality trait measure, rather than to individuals.
References
|
Doctor Marina Sawdon
marina.sawdon@sunderland.ac.uk |
|
UKMEDP081
Junior Doctors’ Training Satisfaction and Progress: Longitudinal Examination of Medical Students Individual Differences, Academic Attainment and Work/Learning Environment Approved on 22 May 2018 UKMED Advisory Board |
Junior doctors in the NHS are under considerable work pressure (1) which may negatively impact their well-being and professional development. To retain a satisfied and competent work force, it is important to better understand which factors predict junior doctors’ work satisfaction, educational satisfaction, and successful progression through training.
The link between work conditions and job satisfaction are widely researched (2) but less is known about how work conditions and individual differences (e.g. personality and learning habits) affect junior doctors’ in the UK satisfaction and training progression.
Previous studies have shown doctors’ personality and early career experiences (e.g. stress) predict their perceptions of their workplace four to five years later (3). Surface learning habits and stress can lead to poorer learning and lower academic achievement in medical school (4) which in turn may negatively affect junior doctors’ postgraduate examination performance (5). Poor academic performance in medical school can also result in difficulties securing a training position within a competitive specialty and/or geographic location, with weaker trainees ending up working and learning in more challenging environments.
Work conditions and individual differences both predict learning and satisfaction, but we do not know how they affect junior doctors’ work and educational satisfaction and progression through training. Neither do we know whether satisfaction is related to educational outcomes.
This research project will investigate how individual differences among medical students and work/learning conditions are linked to junior doctors’ satisfaction and training progression.
References
|
Miss Asta Medisauskaite
a.medisauskaite@ucl.ac.uk |
|
UKMEDP082
Do measures of doctors’ academic ability and their Situational Judgement Test (SJT) scores moderate the relationship between sources of workplace stress and experiencing workplace burnout? Approved on 22 May 2018 UKMED Advisory Board |
For the first time, the 2018 National Training Survey (NTS) includes a measure of workplace burnout – the workplace-based burnout items from the Copenhagen Burnout Inventory. Based on previous research we hypothesise that higher levels of burnout will be associated with a higher workload, a less supportive environment and poorer clinical supervision. These potential sources of workplace stress are also measured by the NTS.
We will explore whether the relationships between the sources of workplace stress and burnout are moderated by the doctors’ academic achievement as measured by their Educational Performance Measure - the EPM decile and their SJT scores. Are more highly performing doctors and those with high SJT scores less likely to experience feeling of burnout when exposed to the same levels of workplace stress?
References
|
Mr Daniel Smith
daniel.smith@gmc-uk.org |
|
UKMEDP083
Factors associated with non-standard outcome at Annual Review of Competence Progression (ARCP) of higher trainees within surgical specialities in the United Kingdom. Approved on 22 May 2018 UKMED Advisory Board |
There are ten surgical specialities recognised within the United Kingdom (UK) National Health Service (NHS) (1). Entry into surgical specialties is competitive: in 2016 competition ratios at national recruitment for surgical specialities ranged from 1.31-6.57 to 1 training place (2.) However, despite competitive entry some trainees leave training before reaching the end of the curriculum. Hampton et al. suggested that between 2008-2012 1.7% of surgical trainees relinquished their training number, with drop-out rates as high as 4.2% in some deaneries (3). However, methodology was poor with data acquired by contacting each deanery directly, with variable and mostly poor response rates. There has also been suggestion from other sources that in surgery, as well as other medical specialties, non completion of training and non standard outcome at ARCP is higher in female trainees and in those from a minority ethnic background (4). For reasons of equality, fairness, workforce planning and finance it is important to investigate factors affecting retention in training and award of non standard outcome at ARCP (5).
Identification of these factors will allow further targeted research and support to be given, structures and attitudes modified and supported and remedial action to be taken if and where required, to ensure progression through training for all who are able to display the required competencies described by a curriculum.
References
|
Miss Hannah Boyd-Carson
hannahboydcarson@doctors.org.uk |
|
UKMEDP084
The Sociodemographic Characteristics of Oral and Maxillofacial Surgeons Approved on 22 May 2018 UKMED Advisory Board |
Oral and Maxillofacial Surgery (OMFS) is a unique surgical speciality because it requires two degrees, and this, as well as a number of other factors, increases training costs.[1] The amount that those engaged in the OMFS training pathway must pay out of their own pocket to achieve the mandatory requirements to complete the training has previously been estimated to cost £71,431 - £113,105.[1, 2] Although accelerated three year dentistry degrees provide a small financial cushion for qualified medics who intend to become OMFS surgeons, many years of university fees, debt and lost personal income can be a daunting financial barrier.[3] Training for OMFS is roughly three-to-four time more expensive than other surgical specialties, and surgery as a whole is significantly more expensive than medicine.[1, 4]
Owing to the profound financial implications of choosing to follow the training pathways towards becoming a OMF surgeon, it may be hypothesized that due to the expense of OMFS training, only students who have a more financially stable support network are able to undertake this career.
We aim to perform a retrospective examination of the pre-medical school socioeconomic demographics of those who enrol within OMFS and progress through training using data from the UKMED database. Using this data, we aim to describe the socioeconomic profile of medical graduates embarking on a career in OMFS. This will provide a great insight to determine whether efforts should be made to improve the financial accessibility of OMFS for potential future trainees.
References
|
Mr Declan Murphy
murphy.declan.1994@gmail.com |
|
UKMEDP085
Evaluating the outcomes and impact of less than full time training on the medical workforce Approved on 22 May 2018 UKMED Advisory Board |
The NHS is facing a serious shortage of doctors and hospitals are struggling to fill all their junior doctor posts. More junior doctors are taking career breaks because of concerns around work-life balance and a perceived rigidity within the structure of a medical career. Alongside this, the general interest in part-time working among junior doctors is rising, meaning that the demand for this style of training is likely to increase.
Aim of the project: To examine how junior doctors working part-time affects their progression through, and completion of, training to become specialists compared to those working full-time.
Method:
1. Demographics (ethnicity, age, sex), exam scores (e.g. A levels, Foundation SJTs) and socioeconomic data (including markers of widening participation) held for approximately 60,000 junior doctors will be compared between full-time and part-time doctors using statistical methods.
2. Time taken to complete specialist training and the rate of securing a consultant post will be compared between full-time and part-time junior doctors within a sample of approximately 2500 junior doctors. Questionnaires will be used to investigate working patterns after completion of training (i.e. full-time or part-time) and reasons for training part-time or full-time.
3. Thirty participants comprising junior doctors and trainers of doctors will be interviewed to understand their experiences of undertaking or providing LTFT training.
Findings will be used to make informed recommendations to the leading organisations in charge of training doctors on how to improve part-time working and use it to retain doctors within the profession for the benefit of patients.
References
|
Doctor Magdalen Baker
magdalen.baker.13@ucl.ac.uk |
|
UKMEDP087
Do ethnic differences in performance and selection across medical education persist when controlling for prior educational attainment? Approved on 07 December 2018 UKMED Advisory Board |
This study aims to investigate the causes of ethnic differences in performance and selection across every stage of medical education, by examining the attainment and selection outcomes of Black and Minority Ethnic (BME) doctors in comparison to their white peers, taking into account ethnic differences in prior attainment.
The lower achievement of BME medical students is widely acknowledged [1, 2]. While BME students enter medical school with slightly poorer A levels [3], the ethnic attainment gap widens at medical school [1], and persists into postgraduate medicine, with BME doctors performing relatively poorly in postgraduate examinations, in recruitment, and in ARCPs [1, 4], and being less likely to get their first choice of Foundation programme [5]. However, no comprehensive study to our knowledge has examined the ethnic attainment gap at each stage of medical education, whilst controlling for prior attainment.
McManus and colleagues found that it is more difficult for BME students to enter medical school even when they achieve the same A-level grades as their white peers [3]. This suggests the need to investigate the comparative selection outcomes for ethnic minorities within medical education. A large scale study investigating outcomes at each selection point of the medical education continuum, using more recent UKMED data and more granular ethnic groupings, would be extremely beneficial to our understanding in ethnic differences in selection outcomes.
This research would create the most comprehensive picture available of ethnic differences throughout the medical education and training pathway, supporting long-term aims of developing and targeting effective solutions.
References
|
Miss Halima Shah
halima.shah@ucl.ac.uk |
|
UKMEDP088
Investigating associations of post-graduate examination performance with socio-demographic characteristics, performance at medical school, medical school, SJT and foundation school: a focus on first stage examinations of MRCP, MRCGP and MRCPsych Approved on 07 December 2018 UKMED Advisory Board |
This study aims to examine the relationship between performance in the first stage of MRCP, MRCGP and MRCPsych with sociodemographic characteristics, the EPM (decile and additional educational achievements), Situational Judgement Test (SJT), medical school and foundation school. We particularly focus on specialties with a key community component, as well as the MRCP, because candidates, who pursue many other specialties with a community component, begin with the MRCP. Additionally, it is timely to reconsider the study of McManus et al. (2008) who demonstrated considerable variation in MRCP examination performance relative to candidates’ medical school. They concluded that “unexplained differences” at entry to medical school and specific medical school components might explain this variation. However, they were unable to look at whether candidate socio-economic background was a contributory factor. This is critical to investigate given: variation between medical schools in student socio-demographics; increasing policy focus on widening access to medicine; and emerging evidence of a relationship between socio-economic background and specialty choice (which will be reflected in who sits particular postgraduate examinations). Improvements in routine data management and the availability of standard performance measures mean it is now possible to do a more forensic examination of these associations. It is important to look at the educational performance measure (EPM) decile and additional educational achievements separately since the relationship between additional educational achievements and success in postgraduate professional exams remains unknown. This work will improve understanding, inform assessment and selection policy and help inform UK government policy regarding the future of healthcare delivery.
References
|
Isobel Cameron
i.m.cameron@abdn.ac.uk |
|
UKMEDP089
The UK Medical Applicant Cohort Study: Applications and Outcomes Study Approved on 07 December 2018 UKMED Advisory Board |
In March 2018, 1500 new English medical school places were created with an aim to “widen the social profile of medical students”.(1) Increasing the proportion of doctors from ‘non-traditional’ backgrounds (i.e. from under-represented social groups) is a priority in medical education; however the evidence for how to achieve it is still relatively poor.
The United Kingdom’s 38 medical schools differ in how they select and educate students, resulting in considerable variability in outcomes for graduates of different medical schools.(2, 3) In particular, some schools attract and/or accept considerably more non-traditional applicants.(4) It is uncertain why.
Most medical selection research examines the technical aspects of selection tests; however understanding how applicant choices affect selection outcomes is also needed since, as one admissions dean put it, “we can only select from those who apply”.(5) Long term follow-up of applicants is necessary to understand how applicant factors predict outcomes.
The proposed research is part of a National Institute for Health Research funded study, which aims to understand and improve medical applicant choices and outcomes.
In the current study we propose to analyse applications to study medicine in the UK from 2007 to 2017 to establish:
• Which medical school and applicant factors predict the combination of medical schools that applicants chose to apply to (the maximum being four);
• whether the choices of traditional and non-traditional applicants differ;
• whether choices predict the likelihood of receiving at least one offer;
• whether choices mediate the relationship between applicant social background and likelihood of receiving at least one offer.
References
|
Doctor Katherine Woolf
k.woolf@ucl.ac.uk |
|
UKMEDP091
Access to HE qualifications and widening participation in medicine Approved on 07 December 2018 UKMED Advisory Board |
Access to HE Diplomas developed to provide pre-university qualifications for mature learners without the usual secondary educational qualifications such as A-levels. Around 35,000 mature students now take Access to HE diplomas each year and circa 600 on diplomas allied to medicine (Mizon, personal communication) that are offered by eight colleges nationally. Overall, success in entering HE runs at 65-70% for those gaining the qualification (Farmer, 2017), but an initial trawl of the UKMED database has only identified 211 medical students entering with Access to HE qualifications between 2011-6, suggesting that the conversion rate to medicine degree programmes may be substantially lower.
At present the majority of UK medical schools do not recognise Access to HE qualifications as a sufficient requirement, and amongst those that do recognition may be restricted to Diplomas from only one or a few of the colleges offering them. One reason is probably a perception that these students may be at greater risk of not completing a medicine programme, though there is evidence that they have considerable persistence in their educational courses (Hinsliff-Smith et al, 2012). A more recent study (Wilkinson et al, 2015) reports positive experiences of Access to HE students entering the Bradford-Leeds medicine course, but no evidence about their likelihood of progressing and completing the programme.
This project proposes to improve the evidence base in two ways: (i) establishing more clearly the numbers and demographic profile of students entering medical school each year with Access Diplomas, and (ii) their relative success at medical school (progression & completion, FPAS educational performance measure, FPAS Situational Judgment Test).
References
|
Doctor Paul Garrud
paul.garrud@nottingham.ac.uk |
|
UKMEDP097
Investigating Core Medical Trainees’ experience of the training environment and associations with workplace-based assessments, progression, training and examination outcomes from an Equality & Diversity (E&D) perspective. Approved on 19 February 2019 UKMED Advisory Board |
Under the Public Sector Equality Duty conferred by the 2010 Equality Act, it is the statutory responsibility of organisations with a public function (such as Medical Royal Colleges) to collect, analyse and interpret data on the outcomes of their training and assessment programmes (1) and to consider how the nine protected groups defined by the Equality Act are differentially impacted by these programmes.
This research / data analysis project looks at these issues in more depth for UK core medical training (CMT) by investigating the progression of trainees as influenced by the quality of their training environment (measured by the JRCPTB’s CMT quality criteria), performance in workplace-based assessments and the bearing these factors have on their training outcomes (measured by non-completion of CMT and ARCP outcomes) and examination outcomes (measured by time taken to pass each part of the MRCP examination and number of attempts).
Previous studies (2-5) have confirmed the association of certain factors, such as demographics and recruitment scores, with training and examination outcomes however indicators of the quality of the training environment, and its interaction with the formative assessments that take place there, have not been included in previous CMT research or examined from an equality and diversity perspective.
We recognise the size and complexity of the analyses required, however we believe this study is essential to deepening understanding of differential attainment.
References
|
Doctor Miriam Armstrong
miriam.armstrong@jrcptb.org.uk |
|
UKMEDP098
Understanding progression in psychiatry training Approved on 22 February 2019 UKMED Advisory Board |
Attrition is a major concern within psychiatric training. Based on one study in 2012, just 65.8% of psychiatry trainees in the UK plan to stay in psychiatry.(1) In 2017 the fill rates for many higher specialty programmes in psychiatry (ST4) was lower than 60%.(2) The aim of this study, therefore, is to analyse the factors which could help to explain these high attrition rates.
The London School of Psychiatry recently carried out a survey with core medical trainees (CT3) which identified a number of factors related to job satisfaction which were central to the decision to step out of training. The systematic review published in 2017 on the crisis in psychiatry in the UK also identified a number of work and learning related barriers associated with choosing psychiatry as a career; for example, 25-50% of trainees were leaving because of lack of resources or lack of adequate supervision.(3) Moreover, studies on trainees in other specialties revealed that training progression depended on trainees’ sociodemographic characteristics. For example, females were doing better in exams,(4) but were more likely to dropout.(5) While this evidence suggests that work/learning related factors and sociodemographic characteristics are important to successful progression through training, these assumptions have not been tested with a large longitudinal sample investigating various work/learning and sociodemographic factors together and cannot be generalizable.
We will use mixed methods to explore the dropout rates in psychiatry training nationally and will investigate the factors which predict if a trainee progresses through training successfully or drops out.
References
|
Doctor Asta Medisauskaite
a.medisauskaite@ucl.ac.uk |
|
UKMEDP101
Investigating the potential factors that might influence Prescribing Safety Assessment (PSA) scores amongst UK final year medical students and their predictive validity for performance in early postgraduate training Approved on 17 May 2019 UKMED Advisory Board |
Prescribing is the main approach used by the NHS to treat illness, alleviate symptoms and prevent future disease – around 1.5 billion prescriptions are written annually costing £15 billion [1]. Sub-optimal prescribing is common in both primary and secondary care – approximately 10% of hospital and 5% of general practice prescriptions contain errors – these represent a major threat to public health and impose significant additional costs to the service [1]. While healthcare system factors contribute it is clear that all new prescribers should possess the necessary knowledge and skills to face this challenging environment. These concerns led to the development of the Prescribing Safety Assessment (PSA) by the British Pharmacological Society (BPS) and Medical Schools Council [2,3] to enable final year medical students to demonstrate that they were competent to assume independent prescribing responsibilities.
The PSA has been delivered to the graduates of all UK medical schools since 2014 (around 45,000 candidates). There are grounds for believing that the PSA is meeting its original objectives of raising the profile of prescribing skills amongst students and their medical schools, identifying candidates who are less well prepared in this key patient safety area and stimulating learning. There are also some reasons to believe that overall performance in the assessment environment is improving (e.g. improving scores on anchor items, reducing variation between schools).
The priority now is to investigate (i) the factors predict PSA performance, (ii) the relationship between PSA performance and other measures of attainment and (iii) the extent to which performance in the PSA predicts subsequent postgraduate performance (predictive validity). Previous studies have linked undergraduate performance to subsequent postgraduate attainment but these have usually involved very broad performance measures [4]. The unique feature of this study is that it will investigate these relationships for an assessment that focuses on a very narrow skillset (prescribing and supervising the use of medicines) that has been the focus of concern.
The proposed studies are large and will require complex analysis. However, we believe this study is essential to provide a clearer understanding of the factors that influence PSA scores and how predictive these are of subsequent performance in early medical training. Indeed, a recent external review of the PSA has called for this research to be undertaken [5]. We have brought together a research team with significant expertise, knowledge, resources and data analysis experience required to complete this work.
References
|
Professor Simon Maxwell
s.maxwell@ed.ac.uk |
|
UKMEDP103
Predictors of postgraduate exam performance in psychiatrists Approved on 17 May 2019 UKMED Advisory Board |
Previous work has explored the determinants of postgraduate performance in both medical specialities and general practice. In particular there has been considerable interest in the nature and causes of differential attainment in postgraduate educational performance between UK and International Medical Graduates (IMGs) (1-4). However, to date, a detailed analysis has not been carried out in relation to psychiatry. It is interesting to note that some degree of differential attainment in relation to the GP ‘Clinical Skills Assessment’ was noted between UK graduates who described their ethnicity as ‘White’ and those who identified as ‘Black and Minority Ethnic’ (BME). A previous investigation suggested that subtle communication and sociocultural differences may underlie differences in attainment between these two medical graduate groups (5). The effective practice of psychiatry demands excellent communication skills and cultural competence. It is noteworthy that one of the largest differential attainment gaps at Annual Review of Competence Panel (ARCP) between UK Graduates (UKGs) and IMGs was for psychiatry (3).
Thus, this study intends to elicit the predictors of psychiatric postgraduate exam performance in both UKGs and IMGs and describe any trends in differential attainment in relation to both graduate group (e.g. UKGs, IMGs) and ethnic self-identification.
References
|
Doctor Paul Tiffin
pat512@york.ac.uk |
|
UKMEDP104
How do students on gateway courses progress through medicine, compared to standard entry peers of similar backgrounds? Approved on 17 May 2019 UKMED Advisory Board |
Students from lower socioeconomic backgrounds are still traditionally underrepresented in medicine in the UK (1). Subsequently, there has been a recent increase in the number of medical schools offering placements on Gateway Year (GWY) programmes. Designed as a form of widening participation (WP), GWY programmes offer places based on contextual admissions, requiring students successfully complete a Year 0, or Gateway year, in order to matriculate onto standard entry (SE) medicine programmes. There has been limited research suggesting that students on GWY year programmes have slightly lower retention rates compared to SE cohorts, and that GWY year students typically perform academically lower than SE peers in preclinical years, but then as well in clinical years, compared to SE peers (2, 3, 4). However, such research has been limited in the number of GWY programmes (two, single-site studies) included. Further, the existing research compares progression of GWY to SE students as a whole, which may not account for mediating factors related to being from a WP background. This proposed project aims to better understand progression and retention rates of GWY students throughout medical school by not just comparing outcomes of GWY students to those on SE, but with particular focus on how GWY students compare to students of similar socioeconomic, “WP,” background who have met admissions criteria for SE. By comparing GWY students specifically to similar demographic subsets of SE students, this research might provide better context for understanding progression of GWY students, and WP students, overall.
References
|
Miss Angelique Duenas
hyad29@hyms.ac.uk |
|
UKMEDP105
Exploring the convergent validity of the Prescribing Safety Assessment (PSA) Approved on 17 May 2019 UKMED Advisory Board |
The University Clinical Aptitude Test (UCAT), first introduced in 2006 as the UKCAT, aims to discriminate between applicants with the same academic qualifications more fairly, to facilitate widening participation with recruitment from under represented social groups, and to accurately and robustly identify the attributes of a good clinician in potential students. The Prescribing Safety Assessment (PSA) examination was developed jointly by the Medical Schools Council and the British Pharmacological Society to address concerns about the prescribing abilities of newly-qualified doctors. The PSA assessment board has identified research work that can be used to further develop the PSA, one area being the relationship between the UCAT cognitive elements and prediction of PSA scores, especially the calculation items in the PSA (1). The UCAT includes quantitative reasoning questions which may predict the outcome in the calculation item (CAL) section of the PSA. Previously, it has been demonstrated that UKCAT scores demonstrate a statistically significant but modest degree of incremental predictive validity throughout undergraduate training(2). The proposed research will analyse the predictive value of the cognitive elements of the UCAT in relation to the scores achieved in the PSA, particularly the calculation skills items of the PSA.
The BioMedical Admissions Test (BMAT) is another method that some medical schools use to select appropriate students for a degree and career in Medicine. Whilst the UCAT is more widely used and the focus of one of the PSA assessment board's research objectives, the ability to assess the link between PSA scores and previous BMAT scores is important for the validity of both assessments.
Also, it is currently unknown whether performance on the PSA projects other aspects of postgraduate educational performance. For example, it may be that the PSA is associated with performance on tests of clinical knowledge, or even clinical skills, as evaluated by postgraduate Royal College membership exams.
References
|
Doctor David Hepburn
david.hepburn@hyms.ac.uk |
|
UKMEDP106
The fairness of the UKFPO post allocation process from the perspective of the patient Approved on 17 May 2019 UKMED Advisory Board |
Students graduating from UK medical schools spend two years in the UK Foundation Programme, based in NHS Trusts. Around 7,000 students are allocated to Foundation Programme training posts every year, with allocations based on students’ performance in key assessments. The allocation process prioritises student location preferences at both Foundation School (largely based on UK regions) and NHS Trust (local) level, so the students scoring highest on the assessments get their first choice location. While this seems fair to students, at its limit it would mean that all the best-performing students are allocated to hospital A in region X and all the worst-performing students to hospital B in region Y. This could be perceived as unfair from the perspective of the patient, because it could mean that the quality of care received depends – in part – on patients’ geographical location.
This study will therefore explore if there is a notable difference in students’ performance in three key assessments (the Situational Judgment Test, Educational Performance Measure and Prescribing Safety Assessment) by region of allocation and, for the West Midlands, within the region. We will use data for around 35,000 students graduating from UK medical schools between 2014 and 2018. The main analysis will use a one-way analysis of variance (ANOVA) approach to assess for differences in mean scores by Foundation School and we will produce a “heat map” for each outcome.
References
|
Doctor Celia Brown
celia.brown@warwick.ac.uk |
|
UKMEDP107
Can Situational Judgment Test (SJT), Educational Performance Measure (EPM) and/or Prescribing Safety Assessment (PSA) scores predict the likelihood of being sanctioned by the General Medical Council (GMC)? Approved on 17 May 2019 UKMED Advisory Board |
The aim of this project is to determine if performance in assessments taken while at medical school can predict whether a doctor will be sanctioned by the General Medical Council. Sanctions can take a number of forms, including of erasure, suspension, conditions, undertakings and warnings; all of which are serious outcomes for the doctor in question. Other evidence suggests that performance in “academic” assessments may be predictive of such outcomes, so we wish to add to the evidence base by considering three national assessments: the Situational Judgment Test, Educational Performance Measure and Prescribing Safety Assessment, for which scores are now available for almost all doctors graduating from UK medical schools and beginning work in the NHS.
We will include data for doctors starting the UK Foundation Programme in 2013-18 (from 2014 for the Prescribing Safety Assessment), and consider any sanctions imposed up to five years after graduation. We will therefore analyse data for around 42,000 doctors and 140,000 person-years of “exposure” (time working in the NHS when a sanction could be imposed). Our primary analysis will consider if there is a difference in the mean performance scores in the three assessments between those with and without a sanction. We will also consider if raising the minimum standard of performance required on any of the assessments could reduce the rate of sanctions, by investigating the “dose-response” relationship between scores and the probability of a sanction.
References
|
Doctor Celia Brown
celia.brown@warwick.ac.uk |
|
UKMEDP108
The effect of eLearning on acquisition of prescribing skills by medical students Approved on 17 May 2019 UKMED Advisory Board |
We will determine whether eLearning can enhance the skills required for safe prescribing. For this study, SCRIPT will be the eLearning resource being assessed and the Prescribing Safety Assessment (PSA) will be the test of the skills required for safe prescribing. As such, we will compare medical students’ performance in the PSA between the periods before and after access was granted to SCRIPT at their respective medical schools. Since the timing of the introduction of SCRIPT differed across medical schools, the study has a natural stepped-wedge design. As such, analysis will be performed using a Hussey and Hughes approach. If test scores are normally distributed, or can be transformed to normality (e.g. logarithmically), then the analysis will be performed using a general linear model. If normality cannot be achieved, then the scores will be dichotomised into pass/fail, and analysed using a binary logistic regression model. All analyses will be conducted using SPSS version 22 (IBM SPSS Inc., Chicago, IL, USA), with P<0.05 considered significant.
References
|
Doctor Sarah Pontefract
s.k.pontefract@bham.ac.uk |
|
UKMEDP110
Demographic predictors of performance on Multiple Mini Interviews for selection to undergraduate medical programmes in the UK. Approved on 20 November 2019 UKMED Advisory Board |
Multiple-Mini Interviews (MMIs) are an admission procedure designed to test personal attributes and overcome the difficulty of context specificity inherent in panel interviews (1). They have been demonstrated to have good feasibility, acceptability, validity, and reliability (2). Consequently, they have become increasingly popular in both undergraduate and postgraduate selection. Indeed, in the UK 57 of the 73 undergraduate medical programmes (including all entry routes) use MMIs within their selection procedures (3).
Our previous systematic review of MMIs identified that there was insufficient evidence regarding potential bias against groups that are underrepresented in medicine (2). Subsequent studies from North America have identified that there is potential differential attainment on MMIs for applicants from lower socioeconomic backgrounds or Black and minority ethnic groups. A study of five Californian medical schools demonstrated that applicants from disadvantaged backgrounds performed better on panel interviews but poorer on MMIs (4). Another study demonstrated that applicants from lower socioeconomic backgrounds, and applicants from Hispanic/ Latino backgrounds performed worse on MMIs (5).
Despite the considerable adoption of MMIs for entry to UK programmes, there has been no multi-institution research into potential bias against underrepresented groups. Given the imperatives to widen access to medicine in the UK, it is important and timely to explore this. This study will seek to identify if certain applicant characteristics are associated with differential performance on MMI. We aim to perform multivariate analyses to identify predictors of performance on MMI at six UK medical schools using data within the UK Medical Education Database (UKMED).
References
|
Doctor Eliot Rees
e.rees@keele.ac.uk |
|
UKMEDP112
How well in medical school applicants do UCAS predicted grades relate to attained grades in examinations, and how do predicted and attained grades differ in their predictive validity for undergraduate and postgraduate outcomes? Approved on 28 May 2020 UKMED Advisory Board Published in medRxiv in June 2020 as "Calculated grades, predicted grades, forecasted grades and actual A-level grades: Reliability, correlations and predictive validity in medical school applicants, undergraduates, and postgraduates in a time of COVID-19" |
The Covid-19 pandemic of March 2020 has resulted, for the first time ever, in Level 3 examinations in schools, such as A-levels and SQA assessments, being cancelled. These assessments normally form the core of medical school selection processes, in particular being a key part of the conditional offers received by many applicants. Ofqual in England has said that A-level results will be replaced by ‘calculated grades’, which will be derived from estimated grades and rankings provided by teachers at the end of May 2020 (and broadly similar arrangements are in place in Scotland, with ‘holistic assessments’ to be provided by teachers). In essence these calculated grades and holistic assessments are likely to correlate highly with predicted grades already provided by teachers on UCAS application forms in October 2019. However surprisingly little is published on the details of the relationship between predicted and actual grades, especially in high-achieving applicants such as those for medicine, and almost nothing on the key question of the relative predictive validity of predicted and actual grades.
This project is built upon two existing projects, UKMEDP089 and UKMEDP051. UKMEDP089 is a part of the UKMACS study of medical school applicants, which as one of its workstreams is looking at data from UCAS since 2007 on medical school applications, including attained and predicted grades. UKMEDP051 is a project which primarily is looking at the predictive validity of the aptitude tests BMAT, GAMSAT and U(K)CAT, in the context of Level 2 and Level 3 qualifications (e.g. GCSEs and A-levels), in relation to outcomes from undergraduate training (dropout, UKFPO EPM, UKFPO SJT and PSA) and from postgraduate assessments (such as MRCP(UK), MRCS, MRCGP, etc.). The data from P89 in the build of 21/2/2020 will provide detailed data on the relationships between predicted and attained grades, and the data from P51, in the build of 13/5/2019, will allow the analysis of predictive validity in applicant cohorts from 2010 to 2014, with undergraduate and postgraduate outcomes.
References
|
Professor Chris McManus
i.mcmanus@ucl.ac.uk |