clinical reasoning
Growth of analytical thinking skills over time as
measured with the MATCH test
Jaap W Groothoff,1 Joost Frenkel,2 Godelieve A M Tytgat,3 Willem B Vreede,4 Diederik K Bosman5 &
Olle Th J ten Cate6
CONTEXT Ber’s Comprehensive Integrative Puzzle
aims to assess analytical clinical thinking in medical
students. We developed a paediatric version, the
MATCH test, in which we added two irrelevant
options to each question in order to reduce guessing
behaviour. We tested its construct validity and studied
the development of integrative skills over time.
METHODS We administered a test (MATCH 1) to
subjects from two universities, both with a 6-year
medical training course. Subjects included 30 stu-
dents from university 1 who had completed a paedi-
atric clerkship in Year 4, 23 students from university 2
who had completed a paediatric clerkship in Year 5,
13 students from both universities who had com-
pleted an advanced paediatric clerkship in Year 6, 28
paediatric residents and 17 paediatricians. We
repeated this procedure using a second test with
different domains in a new, comparable group of
subjects (MATCH 2).
RESULTS Mean MATCH 1 scores for the respective
groups were: Year 4 students: 61.2% (standard devi-
ation [SD] 1.3); Year 5 students: 71.3% (SD 1.6); Year
6 students: 76.2% (SD 1.5); paediatric residents:
88.5% (SD 0.7), and paediatricians: 92.2% (SD 1.1)
1
Department of Paediatrics, Emma Children’s Hospital, Academic
Medical Centre, Amsterdam, The Netherlands
2
Department of Paediatrics, Wilhelmina Children’s Hospital,
University Medical Centre Utrecht, Utrecht, The Netherlands
3
Department of Paediatrics, Zaans Medical Centre, Zaandam, The
Netherlands
4
Department of Paediatrics, Flevo Hospital, Almere, The
Netherlands
5
Department of Paediatrics, Onze Lieve Vrouwen Gasthuis (OVG),
Amsterdam, The Netherlands
6
Centre for Research and Development of Education, University
Medical Centre Utrecht, Utrecht, The Netherlands
Correspondence: Jaap W Groothoff MD, PhD, Emma Children’s
Hospital AMC, University of Amsterdam, PO Box 22700, 1100 DE
Amsterdam, The Netherlands. Tel: 00 31 20 566 7919;
Fax: 00 31 20 691 7735; E-mail: j.w.groothoff@amc.uva.nl
ª Blackwell Publishing Ltd 2008. MEDICAL EDUCATION 2008; 42: 1037–1043
(one-way ANOVA F = 104.00, P < 0.0001). Students of
both universities had comparable scores. MATCH 1
and 2 scores were comparable. Cronbach’s a-values in
MATCH 1 and 2 were 0.92 and 0.91, respectively, for
all subjects, and 0.82 and 0.87, respectively, for all
students.
CONCLUSIONS Analytical clinical thinking develops
over time, independently of the factual content of the
course. This implies that shortened medical training
programmes could produce less skilled graduates.
KEYWORDS multicentre study [publication type]; clinical
clerkship ⁄ *standards; paediatrics ⁄ *education; clinical compe-
tence ⁄ *standards; humans; students, medical ⁄ *psychology;
*thinking.
Medical Education 2008: 42: 1037–1043
doi:10.1111/j.1365-2923.2008.03152.x
INTRODUCTION
A number of test formats have been designed to
capture clinical thinking or reasoning skills. One
approach is the use of short-answer items with
reasonable psychometric properties.1 The drawback
is that these questions need marking and their
application in large groups of candidates is burden-
some. A compromise between multiple-choice and
constructed response questions is found in items with
a relatively long list of options.2,3 This solution
minimises the effects of cueing. However, construct-
ing long lists of options is not easy if it must be done
for many items. Another compromise involves
matching type items. A list of options may serve
several questions in a test by inviting candidates to
match options with item stems. A well-known format
is the extended matching items test,4,5 which has
been reported to show superior psychometric prop-
erties.6 The main disadvantage of this approach is
that it requires candidates to sift through a long list of
1037
J W Groothoff et al
Overview
What is already known on this subject
Few tools exist that are easily applied to assess
analytical clinical thinking in medical students.
In addition, information on the influence of the
length of exposure to medical education on the
development of integrative skills is lacking.
What this study adds
The MATCH, a variant of the Comprehensive
Integrative Puzzle, is an attractive, efficient tool
with adequate construct validity for the assess-
ment of analytical clinical thinking in students.
MATCH scores appear to develop over time,
independently of the factual content of the
medical course.
Suggestions for further research
Our study indicates that essential skills for daily
medical practice may be sufficiently developed
only after 5 years of medical training. This needs
further investigation.
alternatives, which makes it time-consuming and
hence less efficient in daily use.
The Comprehensive Integrative Puzzle (CIP), as
developed by Ber,7,8 may be seen as another exten-
sion of the matching test. This test asks students to
compare and contrast a number of associated diag-
noses in various domains. Students must fill an empty
grid of six diseases against a number of diagnostic or
therapeutic categories (e.g. history, physical exami-
nation, imaging, laboratory tests, treatment or follow-
up and prognosis) with 6 · 6 options in order to
complete six clinical scenarios. Each horizontal row
reflects a typical medical scenario. The test is
constructed as a forced-choice item test: for all six
diagnoses in the grid, six answer options are provided
for each of the categories.
This may induce guessing behaviour, such as when
the candidate knows four options, the last two can be
assumed to fill the remaining open spaces. We
designed a version of the CIP that provides more
options than blanks in the grid and thus disallows
strategic guessing behaviour. This is also advocated
with extended-matching items.9 We expected this test
to show less error variance and higher reliability.
1038 ª Blackwell Publishing Ltd 2008. MEDICAL EDUCATION 2008; 42: 1037–1043
As this test is now less of a puzzle, we called it the
MATCH (measuring analytical thinking in clinical
health care) test.
The aim of our study was to investigate the construct
validity of the MATCH test. Ber advocated that the CIP
should be used along the full continuum of medical
training.8 We therefore hypothesised that test scores
should correlate with stage of training. If we should
also find a favourable reliability, related to limited
testing time, the MATCH test might well be an
efficient tool with widespread application possibilities.
METHODS
Tests
We developed MATCH versions of the CIP, using
four blocks of five associated paediatric diagnoses
with two irrelevant answer options added per cate-
gory. The four blocks of the first MATCH test covered
four paediatric fields: urinary tract and renal diseases;
pulmonary diseases; paediatric diseases with skin
lesions, and gastrointestinal diseases (MATCH 1; see
Appendix). For each of the four blocks, candidates
were asked to combine the five diagnoses with five of
seven archetypical circumscriptions within each of
the different categories (i.e. medical history, physical
examination, additional investigative procedures,
therapy and prognosis). Each domain contained two
irrelevant options. In addition, in some cases, one
option might match with more than one disease.
A maximum total score of 4 · 5 · 5 = 100 could be
achieved. In order to investigate the reproducibility
of the tool, we developed a second test with the same
construction using different paediatric domains (i.e.
‘a coughing child’, ‘a child with stunted growth’,
‘a child with seizures’ and ‘a child with diarrhoea’
[MATCH 2]).
The test was constructed by three paediatricians (JWG,
GAMT and WBV). A first revision was performed by six
paediatric specialists from all the involved paediatric
fields (i.e. gastrointestinal diseases, renal diseases,
neurology, pulmonary diseases, general paediatrics).
All questions were revised a second time by the three
paediatricians. A third consecutive revision was made
by the three paediatricians after a pilot study of both
tests with 10 Year 5 students.
Structure of the medical programmes
The medical programmes of all Dutch medical
faculties are considered to be of equal quality but

they have different curricula. The medical pro-
gramme at the University of Utrecht (university 1)
introduces clinical training early in the curriculum.
Students undertake their first clerkships in Year 3 and
take a paediatric clerkship in Year 4, immediately
after the theoretical paediatric training. During Year
6, students may choose an advanced clerkship in
paediatrics.
By contrast with the Utrecht curriculum, that at the
University of Amsterdam (university 2) consists of a
4-year theoretical course, followed by 2 years of
clerkships at various locations. Each student under-
takes her first clerkship in paediatrics in Year 5, either
at the Emma Children’s Hospital/Academic Medical
Centre (AMC) in Amsterdam or at one of the
affiliated non-academic centres, among others at
Onze Lieve Vrouwe Gasthuis, Amsterdam, at the
Zaans Medical Centre and at the Flevo Hospital
Almere. In Year 6, students may elect to take a
second, advanced clerkship in paediatrics during
which they are supposed to act more independently.
This latter clerkship matches the senior clerkship of
the Utrecht curriculum in terms of structure and
educational aims.
Subjects
Between May 2006 and September 2007, we asked 72
medical students, 17 paediatricians and 30 residents
to complete MATCH 1, and a further 93 medical
students, 15 residents and four paediatricians to
complete MATCH 2. All students performed the tests
at the end of the particular clerkship. Locations of
participation were the Wilhelmina Children’s Hospi-
tal, University Medical Centre of Utrecht (UMCU),
AMC, and three Amsterdam-affiliated non-academic
centres, namely the Onze Lieve Vrouwe Gasthuis, the
Flevo Hospital Almere and the Zaans Medical Centre.
All students following their first clerkship were trained
at the AMC, UMCU or one of the three University of
Amsterdam-affiliated hospitals. All second clerkships
were scheduled in either the AMC or the UMCU.
Paediatricians and residents were recruited from all
six hospitals. In the study with the second test
(MATCH 2), we also asked all participants to score the
amount of time they needed to complete the test.
Scoring and statistics
Mean MATCH 1 and MATCH 2 scores for all groups
of students, paediatricians and residents were
obtained. Separate scores for all four blocks were
diverted to a 0–100-point scale by multiplying the
score four times. Students were compared with
ª Blackwell Publishing Ltd 2008. MEDICAL EDUCATION 2008; 42: 1037–1043
Analytical thinking assessment takes time
residents and paediatricians according to stage of
training. Cronbach’s a was calculated to establish the
test reliability for both tests. Differences between total
scores, as well as between the corrected separate
scores for all four blocks in both tests for all groups
were analysed by one-way ANOVA. SPSS Version 12
(SPSS, Inc., Chicago, IL, USA) was used for statistical
analysis.
Selection of subjects
Between May 2006 and September 2007, all consec-
utive students who started their paediatric clerkship
in one of the Amsterdam-affiliated hospitals (Year 5
students) (i.e. the AMC, the Flevo Hospital Almere,
the Onze Lieve Vrouwe Gasthuis, Amsterdam and the
Zaans Medical Centre) were asked to complete
MATCH 1 (May 2006–January 2007) or MATCH 2
(January 2007–September 2007). In order to analyse
the influence on clinical reasoning of the length of
time of exposure to education, we asked all Utrecht
students (Year 4 students) who started their clerk-
ships between September 2006 and November 2006
to complete MATCH 1 and all Utrecht students who
started their clerkships between July 2007 and
September 2007 to complete MATCH 2. Doctors,
including paediatricians as well as residents, were
randomly selected from all participating hospitals to
complete MATCH 1 and MATCH 2. Doctors who
completed MATCH 1 were excluded from the selec-
tion procedure for MATCH 2. All Year 6 students who
started their advanced paediatric course between
May 2006 and September 2007 were asked to
complete either MATCH 1 or MATCH 2. All
participants received one page of written instructions,
which included an example of one completed grid.
No mock questions were provided.
RESULTS
MATCH 1 was completed by 30, 23 and 19 students in
Years 4, 5 and 6, respectively. Equivalent figures for
MATCH 2 were 23, 60 and 10 students.
In both tests, the mean scores of students were
lower than the scores of doctors (i.e. residents and
paediatricians) (Table 1, Fig. S1). Cronbach’s a-
values were high in the total group and in students,
for both tests (Table 1). Scores for doctors analysed
separately showed somewhat lower a-values.
Standard errors of measurement, calculated as
standard deviation (SD) · square root (1 ) a), var-
ied from 0.28 to 0.40. Furthermore, for both tests,
mean total scores and separate scores for all four
1039
J W Groothoff et al

Table 1 Total scores and reliability analysis of all students and doctors

MATCH test 1 MATCH test 2

n Mean, range, SD Cronbach’s a n Mean, range, SD Cronbach’s a

All subjects 119 76.8, 48–99, 13.1 0.92 98 74.0, 40–99, 12.6 0.91
Students 72 68.4, 48–92, 9.5 0.82 81 70.8, 40–95, 11.1 0.87
Doctors 47 89.6, 79–99, 4.6 0.65 17 89.3, 71–99, 6.2 0.79

SD = standard deviation

blocks of questions for all groups of subjects were
significantly different, each time in favour of the
more experienced group (Table 2, Fig. S2).The
corrected mean differences in scores for blocks 1–4
between students and, by contrast, paediatricians
and residents, recalculated on a 100-point scale,
were 21.2 (SD 2.2), 25.0 (SD 1.8), 21.8 (SD 1.7)
and 21.8 (SD 2.3), for MATCH 1, respectively, and
17.5 (SD 3.0), 19.6 (SD 2.4), 18.0 (SD 2.2) and 20.9
(SD 2.6) for MATCH 2, each time all in favour of
the doctors (both P < 0.0001). In both tests, mean
scores of Year 4 students of University 1 were
significantly lower than those of Year 5 students of
University 2, whereas scores of Year 6 students of
both universities were comparable (Table 2). In
MATCH 2, mean testing time was 49.7 minutes
(range 22–90 minutes, SD 14.1) for students and
34.4 minutes (range 30–40 minutes, SD 3.1) for
residents and paediatricians; 83.3% of all students
completed the test within 60 minutes.

Table 2 Total scores and separate scores for all four blocks of questions in MATCH 1 and MATCH 2 for all groups of subjects

MATCH 1 MATCH 2

Total
Q1* Q2* Q3* Q4* score Q1* Q2* Q3* Q4* Total score

Year 4 53.9 64.4 65.7 (10.5) 60.7 (14.4) 61.2 (7.2) 66.4 (14.0) 55.3 (16.0) 59.5 (17.0) 67.5 (13.4) 62.2 (11.8)
students (10.4) (12.0)
Year 5 67.0 73.9 74.2 (10.8) 70.1 (13.3) 71.3 (7.7) 80.1 (9.4) 67.9 (11.0) 71.5 (14.6) 70.4 (13.4) 72.4 (8.0)
studentsà (11.1) (12.8)
Year 6 73.3 77.7 73.3 (10.3) 78.3 (15.7) 75.7 (5.3) 83.6 (11.7) 74.4 (13.0) 79.6 (9.9) 84.4 (12.0) 80.5 (9.6)
§
students (8.2) (9.7)
Residents– 82.4 94.7 90.3 (5.9) 86.6 (8.8) 88.5 (3.8) 93.6 (2.0) 83.7 (7.6) 86.9 (5.5) 91.5 (10.2) 88.9 (6.1)
(9.4) (5.6)
Paediatricians– 84.9 96.0 94.8 (5.1) 93.2 (6.7) 92.2 (4.7) 99.0 (2.0) 90.0 (9.5) 88.0 (9.8) 94.0 (4.0) 92.3 (5.2)
(9.1) (4.7)
F-value one-way 41.0 49.1 44.3 27.4 104.5 19.0 16.4 11.0 13.6 27.6
ANOVA**

* Scores on questions 1–4 multiplied by 4; Year 4)6 students, paediatric residents and paediatricians
Medical students at the end of the Year 4 paediatric clerkship at UMC Utrecht
à
Medical students at the end of the Year 5 paediatric clerkship at the University of Amsterdam
§
Medical students at the end of the Year 6 advanced paediatric clerkship at both universities
–
Paediatric residents and paediatricians at both universities
** All P < 0.0001

1040 ª Blackwell Publishing Ltd 2008. MEDICAL EDUCATION 2008; 42: 1037–1043
 Analytical thinking assessment takes time

This is much more like the cognitive strategy the

DISCUSSION
MATCH test evokes: diseases or ‘hypotheses’ are
given, and signs and symptoms are ‘checked’. The
The original CIP is an attractive and efficient tool, but
MATCH test stimulates the candidate to compose an
data on its reliability are as yet scarce. Furthermore,
illness script. We would expect the test to show
we expect that its closed formula might induce false
concurrent validity with other tests that aim to test the
high scores because it allows for the use of exclusion
availability of illness scripts in the candidate’s mind,
strategies as well as valid cognitive strategies during
such as the Script Concordance Test.13 Thus, if the
testing. For this reason, we developed a version that
MATCH test does not measure forward clinical
included irrelevant answer options. In this study, we
reasoning, at least it assesses analytical thinking in
aimed to establish the construct validity of the
terms of the identifying of components of clinical
MATCH test and to analyse the effect of the length of
scenarios that must be recognised and synthesised.
time a candidate had spent in medical training on
clinical reasoning skills.
It is interesting that students at Utrecht, who had
undergone 3 years of general medical training,
Both MATCH tests showed a high level of internal
scored on average about 10 points less than students
consistency and discriminative quality. Testing time
at Amsterdam, who had undergone 4 years of general
was < 60 minutes for > 80% of student candi-
medical training, although both sets of students
dates, which makes it a practical tool for use in the
had received virtually the same amount of paediatric
clinical setting. Reproducibility turned out to be
training. After 5 years, students at both universities
good. If reliabilities are related to testing time,
showed comparable results. This time-dependent
the MATCH test appears to be a very efficient
increase in MATCH scores seems to continue after
assessment tool.10
medical school. This may imply that the lengthy
exposure to clinical situations is essential for the
What exactly does the MATCH test apparently mea-
development of clinical skills and is independent of
sure so well? Candidates who receive high scores
the effects of theoretical courses. In other words, a
show the ability to compose logical clinical scenarios
student who takes a certain medical course over
from rich sources of choices. They must have a
3 years will not achieve the same levels of clinical
framework for these scenarios in mind to be able to
reasoning skills at the end of the course as a student
do so. The quality of this reference source relates to
who takes the same course over 4 years.
the number of years of clinical training and experi-
ence. Whether this process should be described as
Limitations of the study
‘clinical reasoning’ is a matter of debate. Clinical
reasoning, in its regular sense of forward reasoning,
This study was limited, firstly, by the voluntary
requires that the subject starts with clinical data which
nature of participation, which may have resulted in
serve as diagnostic cues to direct the thinking toward
relatively high scores. In addition, we were able to
pathologies and diseases. This resembles a regular
include only a relatively low number of paediatri-
clinical encounter. In the MATCH test, however,
cians. The reason for this was that we wanted to create
diseases are given and the candidate is asked to find
equal situations for all participants in MATCH 1 and
the diagnostic and therapeutic features that match
MATCH 2. Asking a doctor who had already partic-
best. In real life, patients do not present with
ipated in the MATCH 1 study to complete MATCH 2
identified diseases and thus the doctor’s quest relates
might have induced bias, as that doctor would have
to diagnosis, rather than symptoms. However, this
had an advantage over students for whom this kind of
argument may not be as strong as it sounds. In
testing was completely new. We have no control on
clinical reasoning, doctors quickly develop hypothe-
the validity of the test with respect to required
ses based on limited early cues from the patient.11
knowledge because no standardised test was available
Instead of a forward reasoning procedure, a backward
for comparison. However, this test was constructed
reasoning process develops that is much more efficient
and revised by a large panel of doctors who all agreed
in practice. The doctor seeks to confirm or reject
on the level of knowledge it was intended to test.
early hypotheses using a process of thought such as:
‘Given a hypothesis X, what symptoms from history,
physical examination, laboratory tests etc. will I
CONCLUSIONS
expect to find in this patient?’ In other words, a
particular ‘illness script’ in the mind of the doctor is
We believe that, despite the shortcomings of this
validated against findings in the patient in question.12
study, the MATCH is a reliable and very efficient tool

ª Blackwell Publishing Ltd 2008. MEDICAL EDUCATION 2008; 42: 1037–1043 1041
J W Groothoff et al
with which to assess clinical reasoning skills in
medical students. Scores on the MATCH test appear
to depend upon length of clinical experience, rather
than on theoretical examinations in the field. This
hypothesis should be investigated in greater depth in
the future.
Contributors: JWG was the primary investigator of the study,
initiated and participated in the construction of the tests,
and wrote the paper. JF, GAMT, WBV and DKB participated
in the construction of the tests and the conducting of the
study. OThJtC acted as senior educational supervisor of the
project. All authors contributed to the revision of the
manuscript and approved the final version for publication.
Acknowledgements: we thank all the individuals who
participated in the study.
Funding: none.
Conflicts of interest: none.
Ethical approval: the Ethical Committee of the Academic
Medical Centre has declared that no approval by their
committee was needed to conduct this study.
REFERENCES
1 Rademakers J, ten Cate TJ, Bär PR. Progress testing
with short-answer questions. Med Teach 2005;27
(7):578–82.
2 Schuwirth LWT, van der Vleuten CPM, Stoffers HEJH,
Peperkamp AGW. Computerised long-menu questions
as an alternative to open-ended questions in
computerised assessment. Med Educ 1996;30:50–5.
3 Rotthoff T, Baehring T, Dicken HD, Fahron U, Richter
B, Fischer MR, Scherbaum WA. Comparison between
long-menu and open-ended questions in computerised
medical assessments. A randomised controlled trial.
BMC Med Educ 2006;6:50.
4 Case SM, Swanson DB. Extended-matching items: a
practical alternative to free-response questions. Teach
Learn Med 1993;5 (2):107–15.
5 Case SM, Swanson DB. Constructing Written Test Ques-
tions for the Basic and Clinical Sciences . National Board of
Medical Examiners 2002. http://www.nbme.org.
[Accessed 20 October 2007.]
6 Fenderson BA, Damjanov I, Robeson MR, Veloski JJ,
Rubin E. The virtues of extended matching and
uncued tests as alternatives to multiple-choice
questions. Hum Pathol 1997;28:526–32.
7 Ber R. Design of an integrative course and assessment
method: the CIP (Comprehensive Integrative Puzzle).
1042 ª Blackwell Publishing Ltd 2008. MEDICAL EDUCATION 2008; 42: 1037–1043
In: Scherpbier AJJA, van der Vleuten CPM, Rethans JJ,
van der Steeg AFW, eds. Advances in Medical Education.
Dordrecht: Kluwer Academic Publishers 1997;84–6.
8 Ber R. The CIP (Comprehensive Integrative Puzzle)
assessment method. Med Teach 2003;25 (2):171–6.
9 Wilson RB, Case SM. Extended matching questions: an
alternative to multiple-choice or free-response
questions. J Vet Med Educ 1993;20 http://scholar.
lib.vt.edu/ejournals/JVME/V20-3/wilson.html.
[Accessed 10 November 2007.]
10 van der Vleuten CP, Schuwirth LW. Assessing profes-
sional competence: from methods to programmes.
Med Educ 2005;39 (3):309–17.
11 Elstein AS, Shulman LS, Sprafka SA. Medical Problem
Solving: an Analysis of Clinical Reasoning. Cambridge,
MA: Harvard University Press 1978;278–9.
12 Custers EJ, Boshuizen HP, Schmidt HG. The influence
of medical expertise, case typicality, and illness script
component on case processing and disease probability
estimates. Mem Cognit 1996;24 (3):384–99.
13 Charlin B, Roy L, Brailovsky C, Goulet F, van der Vle-
uten C. The Script Concordance test: a tool to assess
the reflective clinician. Teach Learn Med 2000;12
(4):189–95.
SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of this article.
Figure S1. Mean scores of students versus paediatricians
and residents of MATCH 1 (1a) and MATCH 2 (1b).
Figure S2. Scores of MATCH 1 (2a) and of MATCH 2 (2b)
of students according time exposure to medical training in
comparison with residents and pediatricians. 4th year ¼ -
clerkship students University 1 (UU); 5th year ¼ clerkship
students university 2 (UvA); 6th year ¼ advanced clerkship
students both universities; residents ¼ residents all centres;
pediatricians ¼ paediatricians all centres (see text).
Please note: Wiley-Blackwell are not responsible for the
content or functionality of any supporting materials
supplied by the authors. Any queries (other than missing
material) should be directed to the corresponding author
for the article.
Received 16 November 2007; editorial comments to authors
7 February 2008; accepted for publication 17 March 2008
 Analytical thinking assessment takes time

APPENDIX

Example of one of the four question blocks (Question 1, MATCH 1)

Disease History Physical examination Laboratory Imaging Treatment

Cystitis
Chronic renal failure
Nephrotic syndrome
Acute glomerulonephritis
Vesical ureteral reflux + pyelonephritis

1 History
a A 10-year-old boy has a headache for 1 day and brown-coloured urine; 2 weeks previously he had a sore throat
b A 15-year-old girl has severe colic pain and red-coloured urine
c A 15-year-old girl has pain in the belly and dysuria; she has never experienced these complaints before
d A 6-year-old girl has swollen eyes and a swollen belly; the previous week she had rhinitis and a sore throat
e A 5-year-old girl feels very ill, with back pain and dysuria; she also ‘feels warm’
f A 6-month-old boy with failure to thrive is vomiting frequently and is polyuric
g A 16-year-old girl has progressive polyuria and thirst of 3 weeks’ duration

2 Physical examination
a Strikingly pale child, slight tachycardia, normal blood pressure, height < ) 2 standard deviation
b No abnormalities
c Oedema in the eyelids, blood pressure 180 ⁄ 105, enlarged lymph nodes in the neck
d Dystrophic child, rapid pulse, thirsty
e Oedema in the legs, ascites and eyelids, blood pressure 90 ⁄ 60
f Temperature 38.0 C, lower belly pain
g Temperature 39.5 C, pain in the right flank

3 Laboratory tests
a Cloudy urine, leukocyturia, normal serum creatinine
b Urine protein stick strongly positive, normal serum creatinine
c Urine protein stick positive, sediment: a few leukocytes and erytrocytes, very low Hb, very high serum creatinine
d Very low serum potassium, low serum phosphate, high fractional Na-excretion
e Haematuria, normal Hb, high serum creatinine
f High fractional phosphate excretion, high fractional salt excretion, glucosuria
g Low fractional phosphate excretion, glucosuria

4 Imaging results
a Ultrasound belly and later voiding cystogram and Di Mercapto Succinic Acid (DMSA) scan
b Ultrasound: anatomy, kidneys?
c Computed tomography, kidneys
d X-thorax: overhydration?
e Not indicated; if necessary X-thorax to establish hydration situation
f Not indicated
g Magnetic resonance imaging, kidneys

5 Therapy
a Broad-spectrum antibiotics
b Salt and fluid restriction, anti-hypertensive drugs
c Calorie-enriched feeding, extra water and salt supply, 1-HO vitamin D
d Corticosteroids, salt restriction
e Co-trimoxazole, amoxicillin of nitrofurantoin
f Hydration, salt, phosphate, magnesium and other electrolyte supply
g Corticosteroids and cyclophosphamide i.v.

ª Blackwell Publishing Ltd 2008. MEDICAL EDUCATION 2008; 42: 1037–1043 1043