undergraduate teaching

Randomised, controlled study investigating the

optimal instructor: student ratios for teaching
suturing skills
Adam Dubrowski1 & Helen MacRae2

INTRODUCTION Recently, there has been a shift in learning, and lower ratios of instructors to students
away from practising procedures on patients for the resulted in significantly less learning. These findings
first time and towards bench model teaching of are in keeping with current motor learning theories.
clinical skills to undergraduate medical students.
However, guidelines for the most effective KEYWORDS randomised controlled trial (publica-
instructor : student ratio for technical skills training tion type); humans; *suture techniques; clinical
are unclear. This has important implications for competence ⁄ *standards; teaching ⁄ *methods; educa-
staffing laboratory based teaching sessions. The pur- tion, medical, undergraduate ⁄ *methods; analysis of
pose of this study was to assess the optimal ratio of variance.
teachers to learners during the teaching of a simu-
lated wound closure. Medical Education 2006; 40: 59–63
doi:10.1111/j.1365-2929.2005.02347.x
METHODS A total of 108 undergraduate medical
students participated in a 1-hour course on wound
closure. They were randomised to 3 groups, each with INTRODUCTION
a different instructor : student ratio (Group A: 6–12;
Group B: 3–12; Group C: 1–12). Students were Both practical and theoretical benefits accrue from
evaluated on a pre-test, an immediate post-test and a teaching surgical skills on bench models rather than in
delayed retention test using an objective, computer- traditional clinical environments. This approach
based technical skills assessment method. Collectively offers important practical benefits in saving money,
termed the Ôeconomy of movementsÕ, the total time time and faculty resources, as well as addressing
taken to complete the task and the number of ethical and patient safety concerns.1,2 Also, a more
movements executed were the primary outcome theoretical advantage of using this new teaching
measures. environment is that it allows researchers to apply
motor learning principles which may serve as an edu-
RESULTS Improvements in the economy of move- cational research vehicle to facilitate these savings.3
ments were the same for Groups A and B and were
better than in Group C (P < 0.005). Although the bench model approach has been
pioneered primarily in the realms of surgical educa-
DISCUSSION The optimal instructor : student ratio tion,4,5 teaching clinical skills to medical students is
was 1 instructor for 4 students. Higher ratios of becoming a common and desirable practice.6,7
instructors to students resulted in no improvements However, teaching clinical and technical skills to
undergraduate medical students is quite different to
teaching such skills to junior residents for several
1
University of Toronto, Department of Surgery, Toronto, Canada reasons. First, medical students have a minimal
2
Department of Surgery, University of Toronto, Surgical Skills Centre,
Toronto, Canada arsenal of technical skills, whereas junior residents
have been exposed previously to a number of clinical
Correspondence: Adam Dubrowski PhD, University of Toronto,
Department of Surgery, The Wilson Centre, 200 Elizabeth Street, Eaton and technical skills. Secondly, the level of interest in
South 1E 583, Toronto, ONM5G 2C5 Canada. Tel: (416) 370 4194; learning clinical and technical skills is inconsistent
Fax: (416) 340 3792; E-mail: [email protected]

Blackwell Publishing Ltd 2005. MEDICAL EDUCATION 2006; 40: 59–63 59

60 undergraduate teaching

to this technical skill, we intended to estimate the

optimal instructor : student ratio for teaching them
Overview wound closure using interrupted sutures. We chose
this particular skill because it is one of the funda-
What is already known on this subject mental and most common skills in every doctor’s
arsenal and is one of the first skills taught to all
Bench model training of technical, clinical medical students. Despite its importance, however,
skills is effective. this skill ) according to informal evaluation data
from our laboratory ) is not acquired by a large
Guidelines for optimal instructor : student percentage of medical students during their surgical
ratio have been developed for teaching cog- clerkship; as a result, they are unlikely to provide
nitive clinical skills. even minor surgical services to their patients once
they are in independent practice.
Clinical technical skills are a subset of psy-
chomotor skills.
METHOD
Optimal instructor : student ratio for teaching
psychomotor skills is skill and learner Participants and procedures
dependent.
A group of 108 undergraduate medical students
What this study adds (years 1 and 2) volunteered to participate in a
training course in instrument suturing and knot-tying
A total of 3 instructors for 12 students is the (simulated wound closure); we repeated this course
most optimal ratio when teaching technical, on 3 separate occasions in order to accommodate all
clinical skill of wound closure to junior med- participants. After giving an explanation of the
ical students. purpose of the experiment, we asked the first 36
participants to sign an informed consent approved by
Suggestions for further research the local Institution Research Board. Following this,
an expert in the field of general surgery gave the
Optimal instructor : student ratios should be participants verbal instructions about the proper
investigated for other clinical skills. suturing technique and provided a demonstration of
this skill. Then, all participants performed a pre-test
Optimal instructor : student ratios should be consisting of wound closure using 10 interrupted
investigated for trainees in other levels of sutures with instruments, thus establishing the base-
training. line performance for further analyses. One week
later, all pre-tested participants took part in a
suturing course, during which they were assigned
among medical students, whereas surgical residents randomly into 1 of the 3 experimental groups
are inherently interested in and excited about (n ¼ 12 in each group) and practised simulated
learning these skills. Thirdly, the typical undergra- wound closure tasks for 1 hour. Eight postgraduate-
duate medical class is larger than a typical first-year year 4 (PGY 4) general surgery residents were asked
surgical resident class. Taken together, these factors to serve as instructors. In order to standardise the
make the teaching of clinical skills to undergraduate learning experience all instructors were asked, dur-
medical students more difficult and, consequently, ing a pre-course briefing session, to teach the skill
impose great demands on teaching faculty resources. using the same method. The 3 experimental groups
The development of clear guidelines outlining the were then formed, each with a different instruc-
optimal instructor : student ratio presents one way of tor : student ratio: Group A, practised with 6
optimizing the learning experience while preserving instructors attending to 12 students (6–12, or a high
teaching resources. Such guidelines have been instructor : student ratio); Group B, with 3 instruc-
developed for teaching cognitive skills;8,9 however, it tors attending to 12 students (3–12, or a medium
is not yet clear whether the teaching of technical instructor : student ratio); and Group C, with 1
motor skills would follow the same protocols. instructor attending to 12 students (1–12, or a low
instructor : student ratio). Upon completion of this
In a simulated environment involving a group of course, all participants were required to perform an
undergraduate medical students who were all novices immediate post-test. They were then asked to return

Blackwell Publishing Ltd 2005. MEDICAL EDUCATION 2006; 40: 59–63

 61

in 1 week to perform a retention test. The immediate

post- and retention tests were identical to the pre-test. RESULTS
This course was repeated in 3 consecutive months.
The Shapiro–Wilks tests showed that the experimen-
Assessment tal data were normally distributed for all groups. The
analysis of base performance assessed during the pre-
During the pre-test and both post- and retention tests, test revealed no group differences on either the total
each participant’s performance was evaluated using time (P ¼ 0.07) or on the number of movements
the Imperial College Technical Skills Assessment (P ¼ 0.2.12).
Device (ICSAD). The construct validity of this device
has been studied extensively, and it is now widely The results of the analyses of difference scores will be
recognised as a suitable form of evaluation of technical discussed separately, first for the total time and then
competence for beginners (see Aggarwal et al.10 for a for the number of movements variables. All results
current review). This evaluation method relies on are presented in Figs 1 and 2.
quantifying hand motion characteristics by using a
commercially available magnetic tracking system Total time
(IsotrackII, Polhemus, Vermont, USA) to track the
positions of magnetic markers placed on participants’ The change in the total time showed main effects for
hands. The positions are tracked in space with group (P ¼ 0.017) and test (P ¼ 0.018). The 2
approximately 20 samples per second sampling fre- groups with highest instructor : student ratio (i.e.
quency (Hz), and the data are integrated over time in Group A: 6–12 and Group B: 3–12) shortened their
order to calculate the instantaneous velocity of each performance time equally. In addition, the partici-
marker. Based on these position and velocity profiles, 4 pants in these 2 groups shortened their performance
measures are obtained: number of hand movements, time significantly more than did the participants in
time per movement, total distance covered and aver- the lowest ratio group (i.e. Group C: 1–12) (Fig. 1).
age speed of each movement. Although all 4 param- Lastly, for all 3 groups, the difference scores
eters can be used to describe hand motions associated decreased from the post-test to the retention test,
with laparoscopic performance,11 only the total time
and the number of hand movements are used to (a)
describe performance related to open procedures.12,13 600
Therefore, for the purpose of this study, the Ôeconomy 500
Total Time (s)

of movementsÕ will be defined as the time taken and the 400

number of movements. We chose this method of 300
assessment because it is currently the most objective >
200 =
and most cost-efficient method of assessing novice
100
performance.10
0
Group A Group B Group C
Analysis
(b)
600
Initially, we tested the normality of data using the 500
Total Time (s)

Shapiro–Wilks test; we then performed a 1-way analysis

400
of variance (ANOVA) with a group (A, B and C) as a
300
factor to test for performance differences between the >
3 experimental groups on the pre-test. Next, for each 200
participant, we calculated the difference scores by 100
subtracting the pre-test score from each of the 2 post- 0
test scores. The difference scores for the total time and Post-test Retention-test
the number of movements were analysed in 2 separate
Figure 1 This plot illustrates difference scores for total
mixed-design ANOVAs. Each ANOVA model consisted of a
time taken to perform the suturing skill (a) as a function of
between-subject factor: 3 groups (A, B and C); and a instructor : students ratio, where Group A: 6–12, Group B:
within-subject factor: 2 test phases (immediate post- 3–12 and Group C: 1–12; and (b) as a function of the
test, retention test). All ANOVA differences significant at retention of the skill from immediate post-test to retention-
P < 0.05 were analysed further using the Tukey test. Groups A and B improved their times significantly
honestly significant difference (HSD) method for post more than did Group C. Also, all groups showed significant
hoc comparison of means. decay in their performance after a week’s retention period.

Blackwell Publishing Ltd 2005. MEDICAL EDUCATION 2006; 40: 59–63

62 undergraduate teaching

150 teaching resources is to set clear guidelines concerning

Number of Movements

optimal instructor : student ratios. Currently, educa-

100 tional research that targets the teaching of technical,
rather than cognitive skills at both the graduate and
= > the undergraduate medical levels sets no effective
50
guidelines for the number of students per teacher.

0 The instructor : student ratio was viewed here as a

Group A Group B Group C
feedback on performance problem with the implica-
Figure 2 This plot illustrates difference scores for the tion that, as a single instructor must divide his or her
number of movements during the performance on the time among several students when providing feed-
suturing skill as a function of instructor : students ratio. back during practice, a higher instructor : student
Again, Groups A and B improved their times significantly ratio increases the amount of direct feedback provi-
more than did Group C. ded to each trainee and a lower ratio decreases the
amount of feedback. Basic research in the field of
suggesting that some forgetting occurred due to the motor skill acquisition suggests that, next to practice,
1-week retention period. appropriate augmented feedback is the most
important variable for facilitating motor learning.14,15
Number of movements Augmented feedback is typically provided externally
and offers information either about the success in
The change in the number of movements showed meeting a movement goal or about how a movement
only a main effect for group (P ¼ 0.005). The 2 is carried out. Although, intuitively, we assume that
groups with the higher ratios of instructor : student feedback facilitates learning, the exact amount of
(i.e. Group A: 6–12 and Group B: 3–12) reduced feedback, its nature and timing influence learning in
their number of movements during the performance different ways.14–17 For example, it has been shown
of the skills equally. However, the participants prac- that, during practice, performance improves as more
tising with the lowest ratio of instructor : students feedback is given;15 however, when learning was
(i.e. Group C: 1–12; Fig. 2) showed significantly less evaluated after a rest period with no augmented
improvement on this measure. There were no inter- feedback ) which is akin to working independently
actions between the experimental groups and test in a clinical setting and simulated in the present study
phase in any of the 2 acquired measures. by the retention test ) the retention of performance
level on the skill was found to be superior when it was
originally practised with less feedback.17 Conversely,
DISCUSSION inadequate feedback during the learning phase,
because of a low instructor : student ratio, can also
This study was designed to test the usefulness of a lead to less long-term learning.
bench model training session to teach clinical skills to
a large group of undergraduate medical students, Collectively, our results suggest that a single 1-hour
and specifically to identify the optimal instruc- intensive training course leads to significant
tor : student ratio for teaching a surgical suturing improvements in the technical aspects of the per-
skill within a single 1-hour training session. formance on simulated wound closure. These
improvements were persistent after 1 week without
The impetus for this study was provided by recent practice, suggesting that true long-term learning has
attempts to utilise bench model training, developed taken place.14
originally for the training of junior surgical residents,
to enhance the teaching of clinical and technical skills The present data also indicate that, for the particular
to medical students. This pedagogical shift brings skills of wound closure, 3 instructors supervising 12
many practical challenges which need further educa- students resulted in the optimal amount of learning.
tional research. These challenges come from the Decreasing this ratio so that more students were
medical students’ limited knowledge base and diverse taught by 1 instructor resulted in relatively less
levels and areas of clinical interest, as well as from the learning, although improvements were still present;
sizes of classes being taught. Together, these factors but increasing this ratio to 6 instructors for every 12
place increased demands on teaching resources, and students did not yield significantly better learning.
ultimately on educational funds. One possible way of These results are in agreement with the existing
optimising the learning experience while preserving motor learning literature, in which there is growing

Blackwell Publishing Ltd 2005. MEDICAL EDUCATION 2006; 40: 59–63

 63

evidence that the optimal amount of augmented 2 Spencer FC. Deductive reasoning in the lifelong con-
feedback for the learning of a motor skill interacts tinuing education of a cardiovascular surgeon. Arch
with the current knowledge base of the learner.18 Surg 1976;111(11):1177–83.
Typically, more feedback is necessary in the initial 3 Dubrowski A, Carnahan H. Principles of motor learn-
ing: can they be applied to surgical education? FOCUS
stages of learning and less as the learner’s knowledge
Surg Ed 2003;20(3):22–2.
base expands. Accordingly, undergraduate students
4 Reznick R, Regehr G, MacRae H, Martin J, McCulloch
require at least 3 instructors for every 12 students in W. Testing technical skill via an innovative Ôbench sta-
order to optimise their learning experience. What is tionÕ examination. Am J Surg 1997;173(3):226–30.
not yet clear, and what needs further empirical 5 Anastakis DJ, Wanzel KR, Brown MH, McIlroy JH,
attention, is how this ratio changes with the growing Hamstra SJ, Ali J, Hutchison CR, Murnaghan J, Reznick
technical knowledge base. RK, Regehr G. Evaluating the effectiveness of a 2-year
curriculum in a surgical skills center. Am J Surg
Although stimulating, these findings are limited to 2003;185(4):378–85.
the clinical skill of interest and to the students’ level 6 Kneebone RL, Scott W, Darzi A, Horrocks M. Simula-
of mastery of this skill. Further research is needed to tion and clinical practice: strengthening the relation-
ship. Med Educ 2004;38(10):1095–102.
study how this instructor : student ratio changes as a
7 Silverman J, Wood DF. New approaches to learning
function of skill difficulty and training level. For
clinical skills. Med Educ 2004;38(10):1021–3.
example, other clinical skills taught frequently on a 8 Lam JW, Farina JP. Personalized competency-based
bench model are intubations and intervenous line instruction applied to medical technology education.
insertions, skills which pose inherently different Am J Med Technol 1980;46(3):173–8.
technical complexities and which may therefore 9 Foley R, Smilansky J, Yonke A. Problem solvers or
require different instructor : student ratios. Also, medical dictionaries: what are we teaching for? Annu
when skills are taught to more advanced students, the Conf Res Med Educ 1977;16:233–8.
predicted ratios could change. Lastly, because the 10 Aggarwal R, Hance J, Darzi A. Surgical education and
gap between the most beneficial estimated instruc- training in the new millennium. Surg Endosc
tor : student ratio and the least beneficial one was 2004;18(10):1409–10.
11 Torkington J, Smith SG, Rees B, Darzi A. The role of
relatively large in this study, it is possible that other,
the basic surgical skills course in the acquisition and
more beneficial ratios are possible. Nevertheless,
retention of laparoscopic skill. Surg Endosc
mindful of these limitations, we have identified and 2001;15(10):1071–5.
explored an area that could have an important 12 Datta V, Chang A, Mackay S, Darzi A. The relationship
bearing on the teaching of clinical technical skills to between motion analysis and surgical technical assess-
undergraduate medical students. ments. Am J Surg 2002;184(1):70–3.
13 Bann SD, Khan MS, Darzi AW. Measurement of surgi-
cal dexterity using motion analysis of simple bench
Contributors: AD was responsible for the initial tasks. World J Surg 2003;27(4):390–4.
conceptualization, design, statistical analysis and 14 Schmidt RA, Lee TD. Motor Control and Learning: a
manuscript preparation. HM was also responsible for Behavioral Emphasis. Champaign, IL: Human Kinetics
conceptualization and manuscript preparation. 1999.
Acknowledgements: the authors would like to thank Dr 15 Magill RA. Motor Learning: Concepts and Applications.
Zane Cohen, Chief of Surgery, Mt Sinai Hospital for partial New York, NY: McGraw-Hill 2000.
funding support of this study, Lisa Satterthwaite, manager 16 Bilodeau EA, Bilodeau IM. Some effects of introducing
of the Surgical Skills Centre at Mount Sinai Hospital and and withdrawing knowledge of results early and late in
Jennifer Chou, research assistant. practice. J Exp Psych 1958;58:142–4.
Funding: this research was supported by the Faculty of 17 Salmoni AW, Schmidt RA, Walter CB. Knowledge of
Medicine Dean’s Excellence Fund, University of Toronto. results and motor learning: a review and critical reap-
Conflicts of interest: none. praisal. Psychol Bull 1984;95(3):355–86.
18 Guadagnoli MA, Lee TD. Challenge point: a frame-
Ethical approval: ethical protocol has been approved by
work for conceptualizing the effects of various practice
the University of Toronto Health Sciences Research Ethics
conditions in motor learning. J Mot Behav
Board.
2004;36(2):212–24.

Received 23 February 2005; editorial comments to authors

REFERENCES 29 April 2005; accepted for publication 27 May 2005
1 Wanzel KR, Ward M, Reznick RK. Teaching the surgi-
cal craft: from selection to certification. Cur Prob Surg
2002;39(6):573–660.

Blackwell Publishing Ltd 2005. MEDICAL EDUCATION 2006; 40: 59–63