Title
Authors
Corresponding author
Co-authors
Driven to distraction: a prospective controlled study of
a simulated ward round experience to improve patient safety
teaching for medical students
Ian Thomas, Laura Nicol, Luke Regan, Jen Cleland, Drieka
Maliepaard, Lindsay Clark, Kenneth Walker & John Duncan
Ian Thomas
Highland Medical Education Centre, University of Aberdeen,
Centre for Health Sciences, Old Perth Road, Inverness, Scotland,
UK. IV2 3UJ.
Email: [email protected]
Telephone: +44 1463 707081
Laura Nicol
Surgical Simulation Fellow
Surgical Academic Unit, University of Stirling, Inverness, UK.
Luke Regan
Clinical Teaching Fellow
Highland Medical Education Centre, University of Aberdeen, Inverness, UK
Jennifer Cleland
John Simpson Chair of Medical Education Research
University of Aberdeen, UK
Drieka Maliepaard
Clinical Skills Facilitator
Highland Medical Education Centre, University of Aberdeen, Inverness, UK
Lindsay Clark
Clinical Skills Facilitator
Highland Medical Education Centre, University of Aberdeen, Inverness, UK
Kenneth Walker
Consultant Surgeon
Department of General Surgery, NHS Highland, Raigmore Hospital, Inverness,
UK
John Duncan
Consultant Surgeon
Department of General Surgery, NHS Highland, Raigmore Hospital, Inverness,
UK
Key words
Word count
Patient safety
Simulation
Ward rounds
Medical error
Distraction
2,921words
(Excluding title page, abstract, reference, figures and tables)
ABSTRACT
Background
Poor situational awareness is an important dynamic in medical error. This is of consideration
for junior doctors, for whom distraction is endemic in clinical work.
Trapping and
eliminating work-based interruption is essential in mitigating error.
Objective
To assess whether simulation-based training can improve undergraduate distractor
management and reduce medical error-making.
Design
A prospective non-randomised controlled study.
Setting & Participants
Twenty-eight final year medical students from one UK medical school.
Methods
Twenty-eight students undertook a simulated baseline ward round. 14 students formed an
intervention group and received immediate feedback on distractor management and error. 14
students in a control group received no feedback. After four weeks, students participated in a
post-intervention ward round of comparable rigour. Changes in medical error and distractor
management between simulations were assessed with Mann-Whitney U-tests using SPSS®
Version 21.
Results
At baseline error rates were high. The 14 students in the intervention group committed a total
of 72 baseline errors (median of 5 errors per student; range 3-7). The 14 students in the
control group exhibited a comparable number of errors – with 76 total errors observed
(median of 6 errors per student; range 4-7). Many errors were life-threatening and included
prescribing patient-allergic antibiotics, inappropriate thrombolysis and medication overdoses.
At baseline distractions and interruptions were poorly managed in both groups. In the
intervention group a total of 29 distractions/interruptions were mishandled (median of 2
mishandled
per
student;
range
1-4).
In
the
control
group
a
total
of
38
distractions/interruptions were poorly dealt with (median of 2.5 mishandled per student;
range 1-5).
All forms of simulation training reduced error post-intervention. In the intervention group
the total number of errors post-intervention fell to 17 (median of 1 error per student; range 03), representing a 76.39% fall (p-value <0.0001). In the control group the total number of
errors also fell - to 44 (median of 3 errors per student; range 1-5), representing a 42.11%
reduction (p-value = 0.0003). Although error rates fell in both groups, intervention reduced
error rates by an additional 34.28% over control (p-value 0.035).
Management of distractions improved in the intervention group alone, with a total of 4
distractions/interruptions mishandled at post-intervention (median of 0 mishandled per
student; range 0-1), representing an 86.21% improvement (p-value 0.0001).
Conclusion
Medical students are not inherently equipped to manage common ward-based distractions to
mitigate error. These skills can be taught through simulation. Curricular integration of
simulated ward round experiences is recommended.
INTRODUCTION
Although error is ubiquitous in the medical profession,
1
it is most common amongst junior
doctors.2-4 Central to the error-making cascade are non-technical skills. Non-technical skills
is an umbrella term including situation awareness, task management, communication skills,
teamwork, leadership and decision making. 5 Failure of these skills is considered responsible
for 56-82% of healthcare-related error. 6&7
Endsley 8 has defined situation awareness as: ‘the perception of elements in the environment,
within a volume of time and space, the comprehension of their meaning and their projection
of their status in the near future.’
Situation awareness is of particular importance to healthcare professionals, as in the hectic
clinical setting, there are a number of environmental factors to contend with – including
interruptions and distractions. An interruption can be defined as any event that causes: ‘the
cessation of productive activity before current task completion for any externally imposed,
observable or audible reason.’
9
An example of an interruption in the clinical workplace
would be a doctor’s pager – where the doctor may have to interrupt primary task performance
to answer the page. Meanwhile, a distraction can be broadly defined as: ‘a stimulus from an
external source that is not followed by cessation of activity, but produces observable
behavioural change in the individual as they continue productive efforts.’
9
An example of
distraction in the clinical environment would be background noise, such as multiple
concurrent conversations in the ward, which may adversely affect concentration.
Distraction and interruption are endemic in the clinical environment. Doctors spend 16-24%
of their working lives engaged in simultaneous activities
workflow interruptions per hour.
13
10-12
and receive an average of 5.3
Through overwhelming cognitive load
14
the adverse
consequences of distraction and interruption are varied and include drug administration
errors,
15, 16
protocol deviance,
17
emergency skills performance.
poor clinical reasoning,
19-21
18
through to impaired surgical and
In healthcare it is estimated that individual distractions
and interruptions result in error rates in the region of 6-12%. 9 & 15 However, with an interplay
of multiple distractions and interruptions, error rates could be as high as 38.9%. 15
The psychological literature suggests that situation awareness can be taught to dampen the
adverse consequences of distraction and interruption. In a systematic review, Li et al.
explored its role on human-computer interaction errors.
22
With repeated exposure to
distraction and interruption, individuals committed fewer primary task errors. Deemed the
‘practice-effect,’ Li extended the argument into healthcare, calling for doctors to be trained in
distraction and interruption management to improve patient safety.
Repeated practice with distraction and interruption allows individuals to more readily detect
its presence. Heightened appreciation of these environmental factors allows the individual to
take appropriate action to minimise its adverse consequences.
In healthcare situation
awareness training has led to the advent of ‘sterile cockpit’ procedures for nurses conducting
drug administration rounds.
The wearing of red ‘do not disturb’ tunics to minimise
distraction and interruption has seen drug administration errors fall by over 42%. 23
With newly qualified doctors committing the highest number of errors, there is a compelling
argument for evaluating situation awareness training pre-graduation. As such, non-technical
skills training in medical undergraduate curriculums are now coming under increasing focus.
The World Health Organisation
24
and the General Medical Council
25
are calling for greater
use of simulated environments in the delivery of non-technical skills training. Reflecting
this, novel simulated ward experiences 26-29 are being reported in the literature. For example,
Nikendi et al.
26
studied 45 final year medical students undertaking a three-patient simulated
ward round experience. In this simulation, students achieved only 64.3% of the technical and
non-technical skills learning goals required for graduation. The authors argued that ‘ward
round training which eases the transition from observing ward rounds to conducting them on
one’s own is urgently required.’
In response, simulated ward experiences which bridge this gap have been reported.
McGregor et al. 27 designed a simulated ward round experience focusing on diagnostic skills,
multi-tasking, prioritisation and help-seeking behaviour.
Meanwhile, McGlynn et al.
28
devised a simulated ‘evening on-call’ for final year medical students, who were challenged to
deal with common on-call tasks such as prescribing night sedation, interpreting blood results
and ECG’s and dealing with acutely unwell patients. Periodically, students were interrupted
with pager requests to add fidelity.
Both studies documented high levels of student
satisfaction and reported participants to be subjectively better prepared for Foundation doctor
practice.
Newer generation simulations are now starting to focus on medical error and the impact of
ward-based distraction and interruption. Smith et al.
29
constructed a simulated ward round
experience for 20 final year medical students, rich in error-prone tasks and realistic
distractions and interruptions such as telephone calls and the ward radio. Post-simulation
students received feedback on their performance from the faculty. The study did not quantify
undergraduate error-making rates, but again reported on high levels of student acceptability.
Elsewhere in the postgraduate literature, Pucher et al.
4
evaluated a simulated ward
experience for surgical trainees, who were assessed on the number of adverse events
committed. Error rates were quantified, with junior trainees committing a significantly
greater number of mistakes than more senior colleagues. However, the simulations were not
repeated to evaluate for a practice effect.
Acceptability of simulated ward rounds for non-technical skills training is now well
documented.
To move simulated ward round initiatives forward, the question must no
longer pertain to participant satisfaction and documentation of medical error, but rather the
power to change patient safety behaviours.
Hence, the objective of this study is:
To assess whether a simulated ward round experience with targeted feedback can improve
undergraduate management of distractions and interruptions to reduce error-making.
METHODS
Development of the simulated ward round experience
In March 2013, a simulated ward round experience was created at the University of
Aberdeen’s Inverness campus, for final year medical students with a specific focus on
distraction, interruption and error. Undergraduates played the part of a Foundation doctor
leading a 30-minute ward round, consisting of three volunteer patients:

An elderly female with pneumonia

A post-operative male with chest pain

An elderly diabetic male with confusion
For each patient, supporting documentation was prepared, which included medical notes and
relevant results. Ward-based protocols were provided to assist the doctor with management
plans and to facilitate prescribing. At each bed space the doctor was expected to complete a
number of clinical duties, with associated potential medical errors as shown in Table 1.
Table 1 – Task completion and associated potential errors during the ward round
Expected task completion
Potential associated errors
At the start of the simulation
Correctly prioritises patients
(sees patient with chest pain first)
Does not correctly prioritise patients
Bed 1 – Clinical Problem: Pneumonia
Correctly diagnoses pneumonia
Uses blood results in notes to calculate
patient’s CURB-65 disease severity score
Calculates CURB-65 score
Prescribes appropriate antibiotic therapy for
patient based on ward protocol
Checks antibiotic vial with nurse ahead of
medication administration
Fails to reach or gives incorrect diagnosis
Fails to recognise that blood results in notes
do not correspond to correct patient
Incorrectly calculates score
Fails to recognise patient is allergic to firstline therapy
Authorises administration of date-expired
medication vial
Bed 2 – Clinical Problem: Post-operative chest pain
Diagnoses non-ST elevation myocardial
infarction (NSTEMI)
Prescribes appropriate therapy for NSTEMI
based on ward protocol
Doctor asked to prescribe Paracetamol for
separate patient
Fails to reach or gives incorrect diagnosis
Fails to appreciate patient is post-operative
and anti-coagulation is contraindicated
Fails to recognise patient receiving Cocodamol and Paracetamol contraindicated
Bed 3 – Clinical Problem: Diabetic with cognitive impairment
Amends Insulin dose as per recommendation
in notes from diabetic specialist nurse
Deals with upset relative who comes into the
ward wanting to speak to the doctor
Misreads handwritten entry in notes as 25
units: and not 2.5 units: resulting in overdose
Does not check relative’s identity and
discloses information on incorrect patient
The simulated ward had a one-way mirror. From behind the glass, faculty facilitated the
simulation and deployed a set of six realistic, time-critical distractions and interruptions,
reflecting the commonality of environmental factors encountered on the ward:

Distraction 1:
Ward radio switched on (at 70 decibels)

Distraction 2:
Domestic hoovers (uses vacuum cleaner) at the patient bed space

Interruption 1:
Doctor’s pager set off

Interruption 2:
Additional prescription task

Interruption 3:
Telephone call into ward

Interruption 4:
Dealing with an upset relative
The student’s performance was assessed by a member of faculty, who documented medical
errors and ‘distractor management’ on a standardised checklist. For ease of terminology,
‘distractor management’ is utilised throughout the article and refers to student handling of
both distraction and interruption.
In April 2013, five senior medical students took part in a pilot study. The average student
committed 5.20 medical errors per simulation and displayed limited skills in distractor
management. This implied that the simulation was of suitable difficulty for future research.
Study design
The main research was conducted as a prospective, non-randomised controlled trial.
Participant recruitment
The following participant selection criteria were used:
Inclusion criteria: Medical students of final year undergraduate training.
Exclusion criteria: Students unable to complete two simulated ward round experiences.
In September 2013 - an intervention group was recruited to the study. The sample population
was drawn from 26 final year medical students undertaking their first clinical placement of
the academic year at Raigmore Hospital, Inverness.
20 students (76.9%) expressed an
interest in taking part with 14 students selected by simple randomisation, using a table of
random numbers.
The aim of this randomisation process was to minimise any further
recruitment bias from the volunteer cohort, for example, in terms of academic ability.
In November 2013 - a control group was recruited. The sample population was drawn from a
total of 25 final year medical students undertaking their second clinical placement of the
academic year at Raigmore Hospital, Inverness. 22 students (88%) expressed an interest in
taking part, with 14 students again selected by simple randomisation.
The intervention
All students undertook a baseline ward round. Following a lag time of 4 weeks students
completed a post-intervention ward round of comparable rigour, achieved by modification of
the baseline ward round’s surface characteristics. This is shown in Appendix 1.
Test-retest theory 30 suggests that observed differences across short time intervals (from hours
to days) are likely to represent a ‘carryover effect’ due to memory, practice or mood. True
behavioural change is more likely to have occurred if differences persist across larger time
intervals - hence the rationale for selecting a lag time of one month between simulations.
Following the baseline ward round, participants in the intervention group received immediate
individualised feedback, targeted on distractor management and its relation to medical error.
As shown in Figure 1, feedback was standardised using SET-GO criteria
31
and delivered by
an experienced clinician who observed the simulation. The feedback aimed to heighten
situational awareness – by improving the recognition and management of both distractions
and interruptions. Although the relationship between distractor management and error was
highlighted, the details of specific errors made by the student in simulation were not
discussed. Participants in the control group did not receive this targeted feedback between
simulations.
Figure 1 – Delivering feedback using SET-GO criteria
G1
S
What I Saw
I saw that as you were prescribing anticoagulation for the patient with a
heart attack, that you looked distracted – were you?
E
What Else I saw
I also saw whilst you were prescribing, that there was a ward radio playing
loudly in the background. Did you notice that?
T
What do you Think?
I found that distracting.
What do you think?
G
What Goal are we trying to achieve?
In future, it would be helpful to eliminate these distractions – especially
when you are carrying out a high-risk task such as anticoagulation, which
can be prone to error.
O
Offers of help
Could I suggest next time, that you ask for the music to be turned down or
switched off?
Outcome measures
A standardised checklist was used to document student performance at both baseline and
post-intervention ward rounds.
The primary outcome measure was:

The number of medical errors committed at each simulation.
The secondary outcome measure was:

The number of distractions and interruptions mismanaged at each simulation.
The aim of the dichotomous standardised checklist was to minimise subjectivity in the
assessment process.
The checklist was completed contemporaneously by a trained member
of the simulated ward round faculty (a clinical teaching fellow), who had the opportunity to
practice this assessment during the simulated ward round piloting phase.
An example
standardised checklist is provided in Appendix 2.
Research validity
The validity of the research was achieved by strict participant inclusion criteria – all
participants were final year medical students within the University of Aberdeen. Entry into
the study from the volunteer pool was undertaken by simple randomisation. All participants
were briefed about the simulated ward round experience, but blinded to the study aims of
assessing medical error and distractor management.
Baseline and post-intervention
simulations were standardised by strict adherence to a faculty manual and a standardised
checklist aided objectivity in assessment of student performance.
standardised using SET-GO criteria.
Feedback was also
Sample size and statistical power
Data from the simulated ward round pilot was used prospectively to inform a sample size
calculation. At pilot, the average student committed 5.20 medical errors per simulation
(standard deviation 2.17). Although there is no previous research with undergraduates, work
from the postgraduate domain suggests that with elimination of distractions medical error
could fall by as much as 42.78%.
23
Assuming a significance level of 0.05 and a power of
80%, eight students would be required in each intervention and control arm.
Statistical analysis
Quantitative analysis of error and distractor management was undertaken using SPSS ®
Version 21 statistical software. Both error and distractor management were non-parametric,
with median and range reported for both data sets. Change in error rates and distractor
management between baseline and post-intervention simulations was evaluated utilising
Mann-Whitney U-tests. To assess correlation between error rates and the mismanagement of
distractions, a Spearman’s rank co-efficient was used.
A flow diagram of the research methodology is shown in Appendix 3.
RESULTS
Study demographics
All participants completed baseline and post-intervention simulations. The intervention and
control groups were comparable across baseline demographics as shown in Table 2:
Table 2: Baseline demographics between intervention and control groups
Demographic
Intervention
Control
P-value
Participants
Males
Females
Average age
14
5
9
23.5
14
5
9
23.71
1.00
Standard deviation 2.79
Standard deviation 2.70
1.00
0.8382
Median number of errors per
student at baseline
5.0
6.0
(Range 3-7)
(Range 4-7)
0.50286
(U-value 83)
Median number of
distractions/interruptions
mismanaged per student at
baseline
2.0
2.5
(Range 1-4)
(Range 1-5)
0.25848
(U-value = 73)
Baseline medical errors and distractor management
There was no statistically significant difference in errors or distractor management between
intervention and control groups at baseline.
In the intervention group a total of 72 errors were committed at baseline (median of 5 errors
per student; range 3-7). In the control group baseline errors were similar – with 76 errors
witnessed (median 6 errors per student; range 4-7).
The pattern of error between intervention and control groups mirrored one another closely as
shown in Figure 2. Errors were witnessed across a spectrum of patient safety domains. The
most commonly committed errors seen at baseline were:

Utilising ‘wrong-patient’ blood results to make management decisions.

Prescribing antibiotics to which the patient was allergic.

Incorrectly checking antibiotic vials as safe to administer.

Inappropriately anticoagulating a patient with a myocardial infarction in the immediate
post-operative period.
At baseline distractions and interruptions were poorly managed in both groups. In the
intervention group a total of 29 distractions/interruptions were mishandled (median of 2
inappropriately dealt with per student; range 1-4).
In the control group a total of 38
distractions/interruptions were poorly dealt with (median of 2.5 inappropriately dealt with per
student; range 1-5).
The relationship between error and distractor management was assessed. A Spearman’s rank
co-efficient was performed on the baseline data set from all 28 students, comparing
individual error-making rates with the number of distractions/interruptions mismanaged.
There was a strong positive correlation between error and the mismanagement of
distractions/interruptions with a Spearman’s rank co-efficient of 0.663 (p-value = 0.01).
Figure 2 – Overview of medical errors made at baseline
Change in patient safety behaviours between simulations
At post-intervention, error-making fell in both groups as shown in Figure 3.
In the intervention arm the total number of errors post-intervention fell to 17 (median of 1
errors per student; range 0-3), representing a 76.39% fall in overall error (p-value <0.0001,
U-value 0.5).
In the control group the total number of errors also fell post-intervention to 44 (median of 3
errors per student: range 1-5), representing a 42.11% fall in overall error (p-value = 0.0003,
U-value 19).
Although medical error rates fell in both groups, the main study finding is that intervention
reduced error rates by an additional 34.28% over control (p-value 0.0035, U-value 34).
Using this effect size of 34.28% a final study power was calculated. With a sample size of
14, a significance level of 0.05 and a standard deviation of 1.03 of error across groups at
baseline, the study power was calculated to be over 80%.
A sub-group of life-threatening error was also analysed. Life-threatening errors were defined
as:

Prescription of antibiotics to which the patient was allergic

Anticoagulating or thrombolysing a patient with a myocardial infarction in the immediate
post-operative period

Prescribing an significant overdose of medication (either Insulin or Thyroxine)
In the intervention group a total of 25 life-threatening errors were committed at baseline
(median of 2 life-threatening errors per student; range 0-3). In the control group baseline lifethreatening errors were similar – with a total of 27 witnessed (median of 2 life-threatening
errors per student; range 1-3). Post-intervention total life-threatening errors fell to 8 in the
intervention group (median of 0 life-threatening errors per student; range 0-2).
This
represented a 68% reduction (p-value 0.00142, U-value 28). In the control group, the total of
life-threatening errors committed post-intervention fell to 18 (median of 1 life-threatening
error per student; range 0-3), representing a 33.33% reduction (p-value 0.0536, U-value
55.5).
Figure 3 – Change in error making rates between baseline and post-intervention simulations
Post-intervention distractor management improved in the intervention group, with a total of 4
distractions/interruptions mishandled (median of 0 mishandled per student; range 0-1),
representing an 86.21% improvement (p-value 0.0001, U-value 13). In the control group
there was a non-significant 2.63% improvement (p-value 0.96012, U-value 96.5), with 37
distractions/interruptions mishandled at post-intervention (median of 2.5 mishandled per
student; range 0-5).
DISCUSSION
What this research adds
Final year medical students are not inherently equipped with the skills to manage ward-based
distraction to mitigate error. However, the results of our novel study indicate that simulated
ward round training can teach undergraduates these skills.
The magnitude by which
simulation can reduce error rates is striking. In the control arm, a group who received no
targeted feedback, error fell by 42%, supporting a practice effect. However, it is simulation
with targeted feedback that is critical to achieve highest educational utility.
In the
intervention arm with a specific critique around distractor management, to enhance situation
awareness, medical error fell by 76%.
Interestingly, the students who took part in the main research were at the very outset of final
year. Students in the pilot study however, were at the end of final year and a matter of weeks
pre-graduation. Despite six months of additional undergraduate experience, the median
errors committed at baseline were similar (Intervention 5.0 errors; Control 6.0 errors; Pilot
5.0 errors). This suggests that undergraduates do not gain competency in these skills through
routine curricular exposure alone.
Simulated ward rounds therefore provide a safe
environment in which to teach these essential skills ahead of graduation.
This is the first study of its kind in an undergraduate population. Although striking, the
results are in keeping with related studies from the postgraduate domain where various
situational awareness training courses have shown medical error-making rates to fall from 2680%.
23 & 32-34
Strengths and weaknesses of the study
This study has several strengths. It is the first study of its kind to document a role for
undergraduate simulated ward round training beyond mere student acceptability.
The
research evidences potential for simulation-based training to positively influence patient
safety behaviours. With meticulous standardisation of simulations and feedback, coupled
with participant blinding – the data set is robust and reliable. Indeed, owing to the magnitude
of error-reduction, despite the small sample size, the study is powered above 80%.
However, this study has the following limitations:

This study was not designed as a randomised controlled trial.
Initially, it seemed
appealing to recruit participants from a single academic block with randomisation to
either intervention or control arms. However, all medical students based at the Inverness
campus share on-site accommodation - and as such, there was concern about crosscontamination between study arms. A conscious decision to run a prospective controlled
study was made, with intervention and control groups formed across two separate
academic blocks.

The potential for selection bias exists. Participants took part on a voluntary basis – and
arguably an engaged volunteer cohort, may be most amenable to the effects of simulation
and feedback.

At recruitment, all students were blinded to the study’s primary outcomes on measuring
error, distraction and interruption.
However, following the baseline simulated ward
round, students in the intervention group received a specific-critique centred on these
themes and little else. As such, it could be argued that the blinding is compromised –
with students in the intervention group aware of the pitfalls to avoid in future simulations.
With this limitation in mind – a lag time of four weeks was specifically chosen between
simulations. This was an attempt to ensure that any improvements in error-making and
distractor management were more likely to represent true behavioural change as opposed
to short-term regurgitative learning. To address this area further, future studies may wish
to include additional intervention arms, where feedback includes extra components above
and beyond distractor management alone.
Implications for future teaching and research
Although study limitations are accepted, the importance of the research findings mandates
feasibility assessment for curricular integration.
In considering feasibility, it is paramount to consider cost and faculty resource. The cost of
running a single day of simulation per student is £61, which includes consumables and
faculty cost. Given its patient safety implications, this is arguably a cost-effective teaching
tool. However, measures to minimise cost and reduce faculty burden exist and are currently
being evaluated at our institution to potentiate the feasibility of curricular roll-out.
We are currently researching a two-patient simulated ward round experience, which delivers
group as opposed to individualised feedback.
The frequency of interruption has been
reduced, to the use of the pager and telephone alone – the interruptions subjectively reported
by students to be most detrimental to performance. In this way, we have been able to reduce
faculty input from 59 to 30 hours/day of simulation, with a resultant 61% cost-saving.
It is important that this study does not stand alone, but instead drives future research. The
following areas of future research are suggested:

Research is needed to evidence whether a lower-cost, curriculum-feasible version of the
simulation, suitable for institutional roll-out can deliver the same level of behavioural
change witnessed in this study. Ideally, such research should be on a multi-centre basis
with incorporation of greater participant numbers.

To assess the extent by which performance in a simulated ward round can predict future
performance.
For example, do simulated ward round behaviours correlate with
achievement at final undergraduate examinations? If so, the simulated ward round may
provide a unique opportunity for detecting students in difficulty to allow timelier
remediation.

In the research study, the benefits of simulation were witnessed over a lag period of 4
weeks. While this may represent a trend towards behavioural change, further research is
required to assess whether such benefits are truly sustained in the long-term. Indeed, it is
important, albeit challenging, to assess whether positive patient safety behaviours are
transferrable into clinical practice in the longer-term – where such behaviours crucially
have the potential to influence patient outcome.
CONCLUSION
Medical undergraduates are not inherently equipped with the skills to manage distractions in
order to mitigate error. It is essential these skills are taught pre-graduation. Simulated ward
round training with targeted feedback has the capability to significantly improve these patient
safety behaviours. Arguably, simulation is the only safe and practical environment in which
students can acquire these skills. With Greenaway’s Shape of Training review
32
calling for
full registration at the point of graduation, medical schools should give due consideration to
curricular integration of simulated ward experiences - to help better prepare students for safe
clinical practice.
Appendix 1 – Changing surface characteristics between pre- and post-intervention simulations
Pre-intervention ward round
Ward round characteristic
Mrs Iris Jones: 80 year old female
Mr John Smith: 65 year old male
Mr Stanley Evans: 70 year old male
Post-intervention ward round
Mrs Isabel Swan: 70 year old female
Mr William Tate: 55 year old male
Mr Arthur Jenkins: 72 year old male
GENERIC FEATURES
Patient demographics
Patient conditions
Mrs Jones: Pneumonia
Mr Smith: NSTEMI post-laparotomy
Mr Evans: Confusion secondary to cognitive impairment
Sepsis
Post-operative myocardial event
Confusion
Mrs Jones: Sepsis: Bed 1
Mr Smith: Post-operative myocardial event: Bed 2
Mr Evans: Confusion: Bed 3
Order of priority:
Mr Smith: Bed 2: SEWS score = 3
Mrs Jones: Bed 1: SEWS score = 1
Mr Evans: Bed 3: SEWS score = 0
Mrs Jones: Diagnoses pneumonia
Utilises history and examination findings, notes, blood
results, chest X-ray and sputum pot
The blood results in the notes do not belong to Mrs Jones
Mrs Swan: Urinary Tract infection
Mr Tate: STEMI post-laparoscopic cholecystectomy
Mr Jenkins: Confusion secondary to hypothyroidism
Mrs Swan: Sepsis: Bed 2
Mr Tate: Post-operative myocardial event: Bed 3
Mr Jenkins: Confusion: Bed 1
Patient bed space allocations
Prioritises patients correctly by SEWS
score
1
1
PATIENT WITH SEPSIS
Demonstrates appropriate diagnostic skills
Checks identity of all test results
Order of priority:
Mr Tate: Bed 3: SEWS score = 4
Mrs Swan: Bed 2: SEWS score = 2
Mr Jenkins: Bed 1: SEWS score = 0
Mrs Swan: Diagnoses urosepsis
Utilises history and examination findings, notes, blood
results, urinalysis and urine specimen pot
The blood results in the notes do not belong to Mrs Swan
Calculates a CURB-65 score
Calculates sepsis score as marker of disease
severity
Calculates a urosepsis score
Patient allergic to Penicillin
Should be given Erythromycin and not Amoxicillin
Prescribes appropriate antibiotics based on
ward-protocol
Patient allergic to Amoxicillin
Should be given Ciprofloxacin and not Tazocin
2
The antibiotic vial is date-expired
The antibiotic vial is of incorrect dosage
Checks antibiotic vial appropriately with
staff nurse prior to drug administration
Mr Smith: Diagnoses Non-ST elevation MI
Utilises history and examination findings, notes, blood
results and ECG’s
Demonstrates appropriate diagnostic skills
Patient should not be anticoagulated with therapeutic
heparin as recently post-operative
Prescribes appropriate therapy based on
chest pain ward-protocol
Asked for regular prescription of Paracetemol: patient on asrequired Co-codamol - therefore contraindicated
Prescribes appropriately when asked to do
so by staff nurse for unrelated patient
3
4
Mr Evans: confusion secondary to dementia
2
5
Mr Tate: Diagnoses ST elevation MI
Utilises history and examination findings, notes, blood
results and ECG’s
POST-OPERATIVE MYOCARDIAL EVENT
PATIENT WITH CONFUSION
Patient should not be anticoagulated with thrombolysis as
recently post-operative
3
Asked to write up routine dose of Warfarin: patient’s INR is
high – therefore contraindicated
4
5
Mr Jenkins: confusion secondary to hypothyroidism
Should amend dose of Lantus from 1.5 units to 2.5 units:
medical note entry could be mistaken for 25 units
Amends routine dose of patient medication
appropriately – despite poorly written
instruction in medical notes
Should amend dose of Thyroxine from 50µg to 100µg:
medical note entry could be mistaken for 100mg
The relative belongs to Mr Smith and not Mr Evans
Checks identity of relative wishing to speak
with doctor at confused patient’s bedside
The relative belongs to Mr Tate and not Mr Jenkins
6
1. Ward radio
2. Domestic hoovering
6
Distractions and Interruptions (denoted in numbered boxes):
3: Pager
4. Additional prescribing task
5. Telephone call
6. Dealing with upset relative
19
Appendix 2 – Standardised checklist for assessing error and distractor management
Candidate number (coded):
Medical error
At the start of the simulation
Did not correctly prioritise patients and order in
which to be seen.
(i.e. Mr Smith, Mrs Jones then Mr Evans).
Bed 1: Mrs Jones
Fails to diagnose chest infection or has to be
prompted to diagnosis.
Uses incorrect patient’s blood result to calculate
CURB score.
Incorrectly calculates CURB score (regardless of
blood results used should = 2)
Prescribes antibiotic to which patient is allergic.
(i.e. Augmentin instead of Erythromycin).
Checks antibiotic vial as satisfactory when it has
actually passed its expiry date.
Bed 2: Mr Smith
Fails to diagnose post-operative non- ST elevation
myocardial infarction, or requires prompting.
Inappropriately administers therapeutic Clexane to
the patient who is post-op.
When asked to prescribe Paracetemol for different
patient does so: not appreciating that they are
already taking Co-codamol.
Bed 3: Mr Evans
Prescribes incorrect dose of Lantus
(Should prescribe 2.5 units and not 25 units)
Does not checks the identity of the relative wishing
to speak to them gives information pertaining to a
different patient.
if error
occurred
Supporting comments
Please use this space to document any errors you felt that the student made, that were not
covered by the checklist above:
Total number of errors made during simulation
20
Please also comment on how the following distractions/interruptions were dealt with by the
student:
Distractions/Interruptions
How this was dealt with by the student?
 Appropriately dealt with
 Inappropriately dealt with
Radio playing in the background
Comments:
 Appropriately dealt with
 Inappropriately dealt with
Cleaner hoovering at patient bedside
Comments:
 Appropriately dealt with
 Inappropriately dealt with
FY1 pager going off
Comments:
 Appropriately dealt with
 Inappropriately dealt with
Asked to prescribe Paracetemol for a separate
unrelated patient
Comments:
 Appropriately dealt with
 Inappropriately dealt with
Being asked to speak to a patient relative
Comments:
Dealing with telephone call into ward
 Appropriately dealt with
 Inappropriately dealt with
Comments:
Please use this space to document any other comments you may have about how the student
dealt with distractions and interruptions:
21
Appendix 3 – Study design flow diagram
Eligibility: 26 final year medical
students at University of Aberdeen
Volunteers: 20 students
Eligibility: 25 final year medical
students at University of Aberdeen
Volunteers: 22 students
Recruitment: 14 students
Recruitment: 14 students
September
2013
November
2013
Baseline
Ward round
Baseline
Ward round
Ward round
Targeted
feedback
No targeted
feedback
October
2013
December
2013
Post-intervention
ward round
Post-intervention
ward round
Dropout rate:
Intervention group N=0
Control group N=0
Data analysis: 14 students
Data analysis: 14 students
Exclusion rate:
Intervention group N=0
Control group N=0
22
DECLARATIONS
Acknowledgments
The author would like to thank The Clinical Skills Managed Educational Network for its
financial support – without which this research would not have taken place.
Copyright statement
The Corresponding Author has the right to grant on behalf of all authors and does grant on
behalf of all authors, an exclusive licence (or non-exclusive licence for UK Crown and US
Federal Government employees) on a worldwide basis to the BMJ Publishing Group Ltd, and
its Licensees to permit this article (if accepted) to be published in British Medical Journal of
Quality and Safety and any other BMJPGL products and to exploit all subsidiary rights, as set
out in our licence.
Competing Interest Statement
All authors have completed the ICMJE uniform disclosure form at
www.icmje.org/coi_disclosure.pdf and declare that: the corresponding author alone had
financial support from The Clinical Skills Managed Educational Network for the submitted
work. This took the form of a £2,000 grant to finance a Master’s in Medical Education
course fee – for which this research project was conducted (Grant reference number CS
MEN044/N). Otherwise, there are no financial relationships with any other organizations
that might have an interest in the submitted work.
Ethics approval status
Ethics approval for this research study was sought and granted from the University of
Aberdeen’s College and Ethics Research Board (Application No. CERB/2013/1/837). All
participants in the study gave informed written consent prior to taking part in the research.
Transparency declaration
The lead author (Ian Thomas) affirms that the manuscript is an honest, accurate, and
transparent account of the study being reported; that no important aspects of the study have
been omitted; and that any discrepancies from the study as planned (and, if relevant,
registered) have been explained.
23
REFERENCES
1. Kohn L, Corrigan J & Donaldson M. To err is Human – Building a safer health system.
Washington DC: National Academy Press; 2000
2. Ryan C, Ross S, Davey P et al.
Prevalence and causes of prescribing errors: the Prescribing Outcomes for Trainee
Doctors Engaged in Clinical Training (PROTECT) study. Public Library of Science
2014; 3;9(1):e79802
3. Arora S, Sevdalis N, Nestel D, et al.
The impact of stress on surgical performance: a systematic review of the literature.
Surgery 2010; 147(3):318-330
4. Pucher P, Aggarwal R, Srisatkunam T, et al. Validation of the Simulated Ward
Environment for Assessment of Ward-Based Surgical Care. Annals of Surgery 2014;
(2):215-21
5. Flin R, O’Connor P & Crichton M.
Safety at the Sharp End: A guide to non-technical skills. Aldershot: Ashgate Publishing;
2008
6. Scharein P & Trendelenberg M.
Critical incidents in a tertiary care clinic for internal medicine. Biomed Central Research
Notes 2013; 16(6):276
7. Cooper J, Newbower R, Long C, et al.
Preventable anaesthesia mishaps: a study of human factors. Anaesthesiology
1978;49(6):399-406
8. Endsley, M.
Toward a theory of situation awareness in dynamic systems. Human Factors 1995; 37(1),
32–64
9. Flynn E, Barker K & Tyrone Gibson J.
Impact of interruptions and distractions on dispensing errors in an ambulatory care
pharmacy. American Journal of Health-System Pharmacy 1999; 56; 1319-25
10. Spencer R, Coiera E & Logan P.
Variation in communication loads on clinical staff in the emergency department. Annals
of Emergency Medicine 2004; 44(3):268-273
24
11. Tipping M, Forth V, O’Leary K, et al.
Where did the day go? A time-motion study of hospitalists. Journal of Hospital Medicine
2010; 5(6):323-328
12. Weigl M, Muller A, Sevdalis N, et al.
Relationship of multitasking, physicians’ strain and performance: an observation study in
ward physicians. Journal of Patient Safety 2013; 9(1):18-23
13. Weigl M, Muller A, Zupance A, et al.
Hospital doctors’ workflow interruptions and activities: an observation study. British
Medical Journal of Quality and Safety 2011; 20(6):491-497
14. Sweller J.
Cognitive load during problem solving: Effects on learning. Cognitive Science 1988;
12(2):257–285
15. Westbrook J, Woods A, Rob M, et al.
Association of interruptions with an increased risk and severity of medication
administration errors. Archives of Internal Medicine 2010; 170(8):683-690
16. Leblanc V, Macdonald R, McArthur B, et al.
Paramedic performance in calculating drug doses following stressful scenarios in a
human patient simulator. Pre-hospital Emergency Care 2005; 9(4):439-444
17. Liu D, Grundgeiger T, Sanderson P, et al.
Interruptions and blood transfusion checks: lessons from the simulated operating room.
Anaesthesia and Analgesia 2009; 108(1):219-222
18. Pottier P, Dejoie T, Hardouin J, et al.
Effect of stress on clinical reasoning during simulated ambulatory consultations.
Medical Teacher 2013; 35(6):472-480
19. Moorthy K, Munz Y, Dosis A, et al.
The effect of stress-inducing conditions on the performance of a laparoscopic task.
Surgical Endoscopy 2003; 17(9):1481-1484
20. Persoon M, Broos H, Witjes J, et al.
The effect of distraction in the operating room during endourological procedures.
Surgical Endoscopy 2011; 25(2):437-443
21. Harvey A, Bandiera G, Nathens A, et al.
Impact of stress on resident performance in simulated trauma scenarios.
Journal of Trauma and Acute Care Surgery 2012; 72(2):497-503
25
22. Li S, Magrabi F & Coiera E.
A systematic review of the psychological literature on interruption and its patient safety
implications. Journal of the American Medical Informatics Association 2012; 19(1):6-12
23. Fore A, Sculli G, Albee D, et al.
Improving patient safety using the sterile cockpit principle during medication
administration: a collaborative, unit-based project.
Journal of Nursing Management 2013; 21(1):106-111
24. World Health Organisation.
WHO Patient Safety Curriculum Guide. 2009.
URL: http://www.who.int/patientsafety/education/curriculum/en/
(Accessed September 2013)
25. General Medical Council. Tomorrow’s Doctors. 2009.
URL: http://www.gmcuk.org/TomorrowsDoctors_2009.pdf_39260971.pdf
(Accessed October 2013)
26. Nikendi C, Kraus B, Briem S, et al.
Ward rounds: how prepared are future doctors?
Medical Teacher 2008; 30(1):88-91
27. McGregor C, Paton C, Thomson C, et al.
Preparing medical students for clinical decision making: a pilot study exploring how
students make decisions and the perceived impact of a clinical decision making teaching
intervention.
Medical Teacher 2012; 34(7):508-517
28. McGlynn M, Scott H, Thomson C, et al.
How we equip undergraduates with prioritisation skills using simulated teaching
scenarios.
Medical Teacher 2012; 34(7):526-529
29. Smith S, Henn P, Gaffney R, et al.
A study of innovative patient safety education.
Clinical Teacher 2012; 9(1):37-40
30. Allen M & Yen W
Introduction to measurement theory
Monterey (CA): Brook/Cole; 1979
26
31. Silverman J, Kurtz S & Draper J.
Teaching and learning communication skills in medicine.
Radcliff Medical Press: Oxford; 1998.
32. Ford D, Seybert A, Smithburger P, et al.
Impact of simulation-based learning on medication error rates in critically ill patients.
Intensive Care Medicine 2010; 36(9):1526-1531
33. Pluyter J, Buzink S, Rutkowski A, et al.
Do absorption and realistic distraction influence performance of component task surgical
procedure?
Surgical Endoscopy 2010; 24(4):902-907
34. Park J, Wagar S, Kersey T, et al.
Effect of distraction on simulated anterior segment surgical performance.
Journal Cataract and Refractory Surgery 2011; 37(8):1517-1522
35. Greenaway, D.
Shape of Training - Securing the future of excellent patient care. October 2013. URL:
http://www.shapeoftraining.co.uk/static/documents/content/Shape_of_training_FINAL_R
eport.pdf_53977887.pdf (Accessed May 2014)