ORIGINAL ARTICLE
Internal medicine resident knowledge of transfusion medicine:
results from the BEST-TEST international education
needs assessment
Richard L. Haspel,1 Yulia Lin,2 Ranjeeta Mallick,3 Alan Tinmouth,4 Joan Cid,5 Hermann Eichler,6
Miguel Lozano,5 Leo van de Watering,7 Patrick B. Fisher,8 Asma Ali,8 and Eric Parks8
for the BEST-TEST Investigators
BACKGROUND: Blood transfusion is the most
common hospital procedure performed in the United
States. While inadequate physician transfusion medi-
cine knowledge may lead to inappropriate practice,
such an educational deficit has not been investigated
on an international scale using a validated assessment
tool. Identifying specific deficiencies is critical for devel-
oping curricula to improve patient care.
STUDY DESIGN AND METHODS: Rasch analysis, a
method used in high-stakes testing, was used to vali-
date an assessment tool consisting of a 23-question
survey and a 20-question examination. The assessment
tool was administered to internal medicine residents to
determine prior training, attitudes, perceived ability, and
actual knowledge related to transfusion medicine.
RESULTS: A total of 474 residents at 23 programs in
nine countries completed the examination. The overall
mean score of correct responses was 45.7% (site
range, 32%-56%). The mean score for Postgraduate
Year (PGY)1 (43.9%) was significantly lower than for
PGY3 (47.1%) and PGY4 (50.6%) residents. Although
89% of residents had participated in obtaining informed
consent from a patient for transfusion, residents scored
poorly (<25% correct) on questions related to transfu-
sion reactions. The majority of residents (65%) would
find additional transfusion medicine training “very” or
“extremely” helpful.
CONCLUSION: Internationally, internal medicine resi-
dents have poor transfusion medicine knowledge and
would welcome additional training. The especially
limited knowledge of transfusion reactions suggests an
initial area for focused training. This study not only rep-
resents the largest international assessment of transfu-
sion medicine knowledge, but also serves as a model
for rigorous, collaborative research in medical
education.
B
lood transfusion is the most common hospital
procedure performed in the United States.1
There is ample evidence, however, of inappro-
priate use of transfusions with a number of
studies documenting a lack of evidence-based practice.2-4
In fact, a Joint Commission and American Medical Asso-
ciation (AMA) report identified transfusion as one of the
five most overused medical procedures.5 Furthermore,
transfusions can pose risks, such as circulatory overload
ABBREVIATION: PGY = postgraduate year.
From the 1Department of Pathology, Beth Israel Deaconess
Medical Center and Harvard Medical School, Boston,
Massachusetts; the 2Department of Clinical Pathology,
Sunnybrook Health Sciences Centre, and the Department of
Laboratory Medicine and Pathobiology, University of Toronto,
Toronto, Ontario, Canada; the 3Clinical Epidemiology Program,
Ottawa Hospital Research Institute, and the 4Department of
Medicine, Ottawa Hospital and University of Ottawa, Ottawa,
Ontario, Canada; the 5Department of Hemotherapy and
Hemostasis, University Clinic Hospital, Barcelona, Spain; the
6
Institute of Clinical Hemostaseology and Transfusion Medicine,
Saarland University Hospital, Homburg/Saar, Germany; the
7
Center for Clinical Transfusion Research, Sanquin/LUMC,
Leiden, the Netherlands; and the 8American Society for Clinical
Pathology (ASCP), Chicago, Illinois.
Address reprint requests to: Richard L. Haspel, MD, PhD,
Beth Israel Deaconess Medical Center, 330 Brookline Avenue,
Yamins 309, Boston, MA 02215; e-mail:
rhaspel@bidmc.harvard.edu.
This study was funded by the Biomedical Excellence for
Safer Transfusions (BEST) Collaborative.
Received for publication July 29, 2014; revision received
October 3, 2014, and accepted November 3, 2014.
doi: 10.1111/trf.12968
© 2014 AABB
TRANSFUSION **;**:**-**.
Volume **, ** ** TRANSFUSION 1
HASPEL ET AL.
and transfusion-related acute lung injury (TRALI), which
can result in significant morbidity and mortality.6 Transfu-
sions have also been associated with worse outcomes in a
variety of clinical settings.7-10
To ensure appropriate and safe transfusions, physi-
cians must learn evidence-based practice. Unfortunately,
data suggest that there is significant variability and con-
siderable room for improvement in transfusion medicine
education.11,12 In a survey of 86 medical schools in the
United States, the majority spent less than 3 hours of cur-
ricular time on didactic teaching related to transfusion
medicine topics.11 The Joint Commission and AMA report
recognized “gaps in medical school and continuing pro-
fessional education” and noted that “there are very short
exposures to transfusion medicine in crowded medical
school and residency curricula, and most ordering physi-
cians do not receive additional education on transfusion
medicine.”5
A transfusion medicine knowledge needs assessment
is required to best understand and then address these edu-
cational deficits.13 While several studies have administered
transfusion medicine assessment tools to physicians, the
exams and surveys underwent minimal validation.14-18 In
addition, generalizability was limited as the study popula-
tions were confined to either a single center or a region.
We previously published a rigorously validated
transfusion medicine exam.19 Using Rasch analysis, a
method used in high-stakes testing, our exam showed
excellent reliability and ability to distinguish between
individuals with different levels of transfusion medicine
knowledge.19-22 In this study, we administered this exam,
as well as a validated survey regarding attitudes and per-
ceived abilities related to transfusion medicine, interna-
tionally to internal medicine residents. Encompassing
more than 23 sites in nine countries, this study represents
one of the largest medical education needs assessments
and can help direct future transfusion medicine educa-
tional initiatives. In addition, our work demonstrates the
utility of a collaborative approach, similar to oncology
study groups that enroll large numbers of patients to
answer important clinical questions, in medical education
research.23
MATERIALS AND METHODS
Participants
Participants included residents specializing in internal
medicine in all years of training or residents completing
internal medicine training before entering another spe-
cialty (e.g., anesthesiology). Individuals who had already
entered subspecialty training (e.g., hematology, cardiol-
ogy) were excluded. The majority of programs have up to
3 years of internal medicine training before entering a
subspecialty. In some countries, there are 1 or 2 years of
“foundation” training, completed by all physicians regard-
2 TRANSFUSION Volume **, ** **
less of specialty. These residents were eligible to take the
exam as were those individuals then specializing in inter-
nal medicine and completing the subsequent additional
required training (up to 3 years). As such, the majority of
data are derived from responses by Postgraduate Year
(PGY) 1 to 3 residents with a small percentage of PGY4
or 5.
Design of the exam
The exam design and validation process has been previ-
ously published.19 Briefly, content was generated by
members of the Biomedical Excellence for Safer Transfu-
sions (BEST) Collaborative (http://www.bestcollaborative
.org). Using a modified Delphi method, this international
group of transfusion medicine experts generated a priority
list of “knowledge or skills related to transfusion medicine
that are absolutely essential for physicians who are not
transfusion medicine specialists (e.g., internists, cardiolo-
gists), but whose practice includes the transfusion of
blood products.” Based on this list, questions were then
developed by BEST Collaborative members with the assis-
tance of the American Society of Clinical Pathology.
This organization oversees the US Pathology Resident
In-service Exam as well as laboratory technologist certifi-
cation exams.24 Questions were multiple choice with one
correct answer and four distractors. Validation was per-
formed by administering the exam to individuals with
expected a priori basic, intermediate, and expert knowl-
edge of transfusion medicine. Results were evaluated
using Rasch analysis (Winsteps; www.winsteps.com). This
psychometric approach, used in testing for medical licen-
sure and certification, compares exam results to those
predicted by a model based on question difficulty and
examinee ability.20,21 Exam quality is determined by calcu-
lating the “fit” of each question with the model with an
ideal score of 1.00. The final 20-question exam demon-
strated a reliability of 0.80 and an ideal average question fit
score of 1.00.
Design of the survey tool to assess resident
attitudes and perceived knowledge
The survey included questions on demographics, prior
transfusion medicine training during medical school and
residency, and resident attitudes and perceived knowl-
edge related to transfusion medicine. The survey was par-
tially based on a prior validated survey examining resident
knowledge of biostatistics.25 The current survey was vali-
dated using responses from anesthesiology residents and
faculty. This group was chosen as one having transfusion
experience but who would not be involved in the internal
medicine resident study. A total of 35 anesthesiology resi-
dents and faculty took the survey and were included in the
validation. Again, Rasch analysis was used to determine
reliability and question fit.22 The Rasch model compares
the individuals surveyed and the questions in the context
 MEDICINE RESIDENT TRANSFUSION KNOWLEDGE

of the construct being measured (in this case, perceived

transfusion medicine level of knowledge). Survey reliabil- TABLE 1. Characteristics of the 474 participants
ity was 0.94 and the average question fit score was 1.00. Sex* Number Percent
Female 202 50.8
The final survey tool consisted of 23 questions and is avail-
Male 196 49.2
able in Appendix S1 (available as supporting information PGY Number Percent
in the online version of this paper). 1 178 37.5
2 110 23.2
3 124 26.2
Assessment tool administration 4 36 7.6
>4 26 5.5
The assessment tool included both the survey questions Sites Number Number of residents
on demographics, attitudes and perceived knowledge, Australia 1 15
Canada 2 41
and the exam questions. It was available in English and England 3 83
was also translated into Spanish, German, and Dutch. Ireland 2 49
Responses to the questions were recorded by participants Italy 1 3
Germany 2 25
on Scantron forms (Scantron Corporation, Eagan, MN) The Netherlands 1 11
typically at a regularly scheduled teaching conference. Spain 1 17
Thirty minutes were allotted for completion of the assess- United States 10 230
Amount of training (%) Medical school Residency
ment tool. Responses were anonymous. After the forms None 12.1 27.6
were turned in, the local site investigator provided a 1 hr 28.0 32.9
30-minute review of the exam answers as an education 2 hr 24.8 21.9
3 hr 16.8 10.8
benefit for the participants. The assessment tool was 4+ hr 18.3 6.8
administered from March through July 2013. This study Quality of training (%) Medical school Residency
was approved for exempt status by the institutional review Not helpful 14.8 23.9
Slightly helpful 38.1 21.4
board at Beth Israel Deaconess Medical Center. Moderately helpful 36.0 37.7
Very helpful 9.6 14.8
Extremely helpful 1.5 2.1
Statistical analysis
* 76 individuals did not indicate male or female.
Only completed exams were included in the analysis.
Descriptive statistics such as mean, standard error (SE),
and percent were used to describe the demographics of
the participants, previous training in transfusion medi- the exam. (Table 1) Participants were evenly distributed
cine, attitudes toward transfusion medicine education, between male and female. While most respondents
and perceived knowledge of transfusion medicine. For the reported some transfusion medicine training in both
nine questions (Questions 14-22 on the survey) regarding medical school and residency (88 and 72%, respectively),
perceived ability to manage transfusion medicine–related the majority (53 and 55%) indicated that they participated
issues, a composite average was calculated for each indi- in only 1 or 2 hours of educational sessions. There was
vidual. Scores (% correct) were tabulated for the total evidence of greater training in medical school than in resi-
exam as well as by question. Statistical correlation of dency with 35% of participants noting 3 or more hours of
scores with covariates including PGY, quantity and quality training compared to 18% in residency. In regard to the
of transfusion medicine education in medical school and quality of training, 74% rated their medical school training
residency, perceived knowledge in transfusion medicine, as “slightly” or “moderately” helpful. For residency train-
and perceived ability in managing transfusion medicine– ing, these ratings were chosen by 59% of the participants
related patient issues was done using a mixed model in with a larger number indicating that the training was “not
which sites were considered to have random effects. To helpful” (24% vs. 15% for medical school).
correlate exam scores with perceived ability to manage
transfusion medicine–related issues, the residents were
divided into groups above and below the median compos- Survey responses
ite score. All analyses were performed using computer The overwhelming majority of residents knew that their
software (SAS 9.3, SAS Institute, Inc., Cary, NC). hospital had transfusion guidelines and how to contact
the blood bank. A majority but smaller percentage of resi-
RESULTS dents (72%) knew how to contact a transfusion medicine
physician. Eighty-nine percent of residents had obtained
Demographics of internal medicine consent for transfusion from a patient (Table 2).
resident participants Almost all the residents rated their knowledge of
Of 518 residents attempting the assessment tool, a total of transfusion medicine as “beginner” or “intermediate” (46
474 residents (92%) at 23 sites in nine countries completed and 48%, respectively). Residents were also asked to rate,

Volume **, ** ** TRANSFUSION 3

HASPEL ET AL.

TABLE 2. Background transfusion medicine TABLE 3. Attitudes and perceptions related to

knowledge and experience transfusion medicine
Question Percent Overall ability to handle transfusion medicine Mean (SE)
Does your hospital have transfusion guidelines?* issues that occur on-service*
Yes 85.0 PGY1 2.5 (0.04)
No 1.3 PGY2 2.8 (0.06)
Don’t know 13.7 PGY3 2.9 (0.06)
Do you know how to contact the blood bank? PGY4 2.7 (0.1)
Yes 96.6 PGY>4 3.0 (0.11)
No 3.4 I would rate my knowledge of transfusion Percent
Do you know how to contact a transfusion medicine MD? medicine as responding
Yes 72.2 No knowledge 1.7
No 27.8 Beginner 46.0
Have you obtained consent for transfusion from a patient? Intermediate 47.7
Yes 89.2 Advanced 4.2
No 10.8 Expert 0.4
To provide appropriate care to patients in
* All sites have transfusion guidelines. your practice, how important is
knowledge of transfusion medicine?
Not important 1.1
Slightly important 2.3
on a five-point Likert scale, their ability to manage a Moderately important 19.9
variety of transfusion medicine patient care issues (Ques- Very important 53.7
Extremely important 23.0
tions 14-22 on the survey). A composite average response How helpful would you find additional
was then calculated for each resident. The mean response training in transfusion medicine?
was between “fair” and “good” for PGY1, 2, 3, and 4 resi- Not helpful 0.8
Slightly helpful 6.3
dents (2.5, 2.8, 2.9, and 2.7, respectively) and good (3.0) for Moderately helpful 28.1
PGY>4 residents (Table 3). Very helpful 44.3
Residents believed that transfusion medicine knowl- Extremely helpful 20.5
edge was important in caring for patients with 77% rating * An average composite score of responses to Questions 14-22
on the survey. For each question, participants were asked to
this knowledge as “very” or “extremely” important. Almost rate their ability on a scale of 1 to 5 (1 = poor, 2 = fair,
all residents would welcome more transfusion medicine 3 = good, 4 = very good, 5 = excellent).
training with 93% rating such activity as at least moder-
ately helpful and 65% rating as either very or extremely
helpful (Table 3).
TABLE 4. Exam scores by PGY
PGY Number Mean (%)*† SE (%) Range
Exam results 1 178 43.9 1.5 10-80
The overall mean score was 45.7% with an individual 2 110 46.0 1.7 15-80
3 124 47.1 1.6 20-85
range 10% to 85% (Table 4). Mean scores adjusted for site 4 36 50.6 2.3 25-75
effect for each PGY ranged from 43.3% (PGY>4) to 50.6% >4 26 43.3 2.5 25-55
(PGY4). The overall difference by PGY reached significance Overall 474 45.7 0.56 10-85
(p = 0.007). The mean score for PGY1 (43.9%) was signifi- * The mean scores are adjusted for clustering by site.
† p = 0.007 for difference in score by PGY.
cantly lower than for PGY3 (47.1%; p = 0.02) and PGY4
(50.6%; p = 0.002) residents. The mean score of PGY4 was
also significantly greater than PGY>4 residents (p = 0.008).
Overall mean site scores ranged from 31.7% to 56.1% (p = 0.03) but not during residency (p = 0.12; Table 6). The
(Fig. 1). exam scores of those having more than 2 hours of training
In regard to question topic, the lowest scores were for during medical school were significantly higher than
questions regarding TRALI (Table 5). These scores ranged those with no training. Residents with perceived better
from 8.9% to 13.5%. Six of the eight questions with correct quality transfusion medicine training during medical
response rates of less than 25% were related to transfusion school or residency (i.e., rating training as very or
reactions. Questions with more than 75% correct response extremely helpful) also did significantly better on the
rates were related to resuscitation for acute bleeding, exam (both p = 0.03). In addition, residents with perceived
management of an acute hemolytic transfusion reaction, greater transfusion medicine knowledge also had higher
indications for irradiated blood products, and platelet exam scores (p = 0.002). There was not, however, a signifi-
(PLT) transfusions. cant association between exam scores and resident self-
There was a significant association between exam perceived ability to manage transfusion medicine issues
scores and amount of training during medical school (p = 0.13). For all of the above comparisons, the highest

4 TRANSFUSION Volume **, ** **


Fig. 1. Scores by site. Mean score (% correct) and SE for each site.
TABLE 5. Question scores by topic
Topic % Correct
TRALI
TRALI
TRALI
Allergic
RBC transfusion (symptomatic)
Septic reaction
Massive transfusion
TACO
Infectious disease risk
PLT transfusion (postsurgery)
Coumadin
Plasma transfusion (procedure prophylaxis)
Reaction reporting
PLT transfusion (prophylaxis)
RBC transfusion (prophylaxis)
Acute hemolytic transfusion reaction
Acute hemolytic transfusion reaction
PLT transfusion (procedure prophylaxis)
Irradiation
RBC transfusion (acute blood loss)
TACO = transfusion-associated circulatory overload.
average group score was 50.3% and the lowest was 43.0%
(Table 6).
DISCUSSION
We performed a large multi-center international transfu-
sion medicine education needs assessment using a rigor-
ously developed and validated survey and exam. In the
initial piloting of the exam, individuals with a priori
expected basic, intermediate, and expert transfusion
knowledge had average scores of 42%, 62%, and 82%,
respectively.19 In that context, for the current study,
residents overall did poorly with a mean score of 45.7%.
Residents showed particularly limited knowledge of trans-
fusion reactions. This finding raises concerns because of
MEDICINE RESIDENT TRANSFUSION KNOWLEDGE
the involvement of residents in several
aspects of transfusion practice. For
example, nearly 90% of residents had
obtained informed consent for transfu-
sion from patients. While PGY3 and
PGY4 had statistically higher scores
than PGY1 residents, the absolute differ-
ence in scores was <7%. In addition,
PGY > 4 residents had significantly
lower scores than PGY4 residents.
Our results are consistent with
previously published findings of
poor transfusion medicine exam
performance.14-17 As with any area of
medical research, however, studies
investigating attitudes and knowledge
must utilize accurate and precise assess-
ment tools. The studies cited above,
however, administered exams that had
undergone limited validation prior to
administration to the targeted resident population. In
addition, any exam quality analysis that was performed
post-administration relied on a classical test theory-based
8.9 approach. These results are dependent on the examinee
9.7
13.5 sample.20,21 That is, as opposed to an item response theory-
14.4 based approach (e.g., Rasch analysis), the same results may
16.5 not be obtained if the exam is administered to different
17.1
20.9 cohorts. As such, the validity and accuracy of these other
23.8 knowledge assessments is uncertain.
32.3 Compared with previous published work, our study
49.4
53.2 represents the largest assessment of transfusion medicine
59.5 knowledge of internal medicine residents throughout all
66.0 years of training and is the first to do so on an interna-
67.1
67.7 tional scale. In addition, as opposed to an assessment tool
67.9 designed by a single individual or small group of authors,
77.9 we utilized an international group of transfusion medicine
79.5
81.9 experts to develop content, the expertise of an organiza-
88.0 tion that develops national certification and in-service
exams to develop questions, and performed quality analy-
sis of both the exam and survey with a method used in
high-stakes testing such as for medical licensure.19-22
Our study also confirms the variability and relatively
minimal amount of training in medical school, regardless
of country, in transfusion medicine.11,12 Remarkably, 65% of
residents reported 2 hours or less of transfusion medicine
educational sessions during medical school. This result is
similar to findings reported from a survey of 86 American
undergraduate medical schools.11 Our study also adds
additional data that many residency programs offer less
than 3 hours of transfusion medicine education.
Several studies have shown a correlation between
greater transfusion medicine education and exam
scores.15,17 In our study, while residents who reported more
than 2 hours of transfusion medicine training during
medical school or rated training received during medical
Volume **, ** ** TRANSFUSION 5
HASPEL ET AL.

possible to obtain with smaller single-site or regional

TABLE 6. Correlation of training and perceived studies.
knowledge with exam scores There are several limitations to our study. While we
Question Number Mean* SE p value
enrolled a large number of residents from different coun-
Medical school hours
None 57 43.0 1.9
tries, our results cannot be said to represent every country
1-2 249 45.4 1.5 0.03 or region. Sites in developing countries were also not
2+ 165 47.3 1.6 included. Still, it appears that there is widespread room for
Residency hours
None 131 44.2 1.6
improvement in transfusion medicine knowledge among
1-2 260 46.2 1.5 0.12 medical residents. In addition, no single exam or survey
2+ 83 47.2 1.8 can assess all areas of importance and the results are not
Quality of medical school training
Not/slightly helpful 247 44.6 1.5
necessarily applicable to other specialties. We do hope our
Moderately helpful 168 46.8 1.6 0.03 exam (available by contacting the lead author) can be
Very/extremely helpful 52 48.3 2.0 used by educators in other specialties as a starting point to
Quality of residency training
Not/slightly helpful 214 44.7 1.5
test their own residents and as a backbone for adding rel-
Moderately helpful 178 45.8 1.5 0.03 evant questions. Our exam also does not demonstrate
Very/extremely helpful 80 48.6 1.8 ability in performance-based competencies and studies
Self-rated transfusion medicine knowledge
Beginner 226 44.1 1.5
using tools such as objective structured clinical examina-
Intermediate 226 47.0 1.5 0.002 tions may help in this regard.16
Advanced 22 50.3 2.6 In summary, using a validated exam and survey, we
Composite score of self-rated ability to manage transfusion
medicine issues
have demonstrated, across a wide variety of sites and coun-
Below median 237 45.0 1.4 0.13 tries, the poor state of internal medicine resident knowl-
Above median 237 46.5 1.4 edge of transfusion medicine. Given the large number of
* The mean scores are adjusted for clustering by site. transfusions performed worldwide, we hope that this study
can focus international attention on improving trainee
transfusion medicine education, offer a valuable knowl-
school or residency as very or extremely helpful did sig- edge needs assessment to help guide future effective edu-
nificantly better on the exam, the mean scores in these cational initiatives, and provide a model for collaborative
groups were still approximately 50% or lower. As such, research in other areas of medical education.
while teaching sessions do appear to have some benefit,
there is still a significant need to internationally improve ACKNOWLEDGMENTS
both the quantity and the quality of transfusion medicine
education. The authors thank John Mitchell, MD, Beth Israel Deaconess
The results of this study can be used to assist in the Medical Center (Boston, MA), for his assistance recruiting indi-
design and implementation of effective educational inter- viduals to take the pilot survey.
ventions. The low scores on questions related to transfu- The Biomedical Excellence for Safer Transfusions Collaborative
sion reactions help identify an important focused initial Transfusion Education Study (BEST-TEST) Site Investigators
topic area for development of transfusion medicine curri-
D.M. Arnold, MD McMaster University Hamilton, Ontario,
cula. Knowledge related to transfusion reactions and risks Canada
is critical for providing appropriate patient management J.P. Brooks, MD University of Florida Gainesville, FL
P.M. Carey, MD University of Cincinnati Cincinnati, OH
and obtaining informed consent. In the future, we also Medical Center
plan to query the sites with higher scores on the exam to T.M. Chalifoux, MD University of Pittsburgh Pittsburgh, PA
Medical Center
determine whether certain teaching methods or experi- C.S. Cohn, MD, PhD University of Minnesota Minneapolis, MN
ences may have led to better performance. The findings A.K. Davis, MBBS The Alfred Hospital Melbourne, Australia
N.M. Dunbar, MD Dartmouth-Hitchcock Hanover, NH
that residents recognized the importance of transfusion Medical Center
medicine knowledge and would find more training helpful W.H. Dzik, MD Massachusetts General Boston, MA
Hospital
is a hopeful sign that additional teaching sessions would M.K. Fung, MD, PhD University of Vermont Burlington, VT
be well received. A. Greinacher, MD Universitätsmedizin, Greifswald, Germany
K. Selleng, MD Greifswald
Our study is also one of the few to perform a knowl- A. Hayat, MBBS Galway University Galway, Ireland
edge needs assessment on an international scale. In Hospitals
M. Kulaga, MD Norwalk Hospital Norwalk, CT
areas such as oncology, cooperative groups are common- M.F. Murphy, MD Oxford University Oxford, UK
place and enable enrollment of large numbers of Hospitals
D. O’Donghaile, MB, BCh St James’s Hospital Dublin, Ireland
patients to answer important clinical questions. We have K. Pendry, MB, BCh Central Manchester Manchester, UK
demonstrated such a collaborative approach is feasible University Hospitals
F. Salerno, MD Policlinico IRCCS San Donato, Italy
in medical education research and can yield more gen- A. Ziman, MD University of California, Los Angeles, CA
eralizable and statistically relevant results that are not Los Angeles

6 TRANSFUSION Volume **, ** **


CONFLICT OF INTEREST
The authors have disclosed no conflicts of interest.
REFERENCES
1. Pfuntner A, Wier LM, Stocks C. Most frequent procedures
performed in U.S. hospitals, 2010. Healthcare Cost and
Utilization Project, Statistical Brief # 149. Rockville (MD):
Agency for Health Care Policy and Research; 2013.
2. Wallis JP, Dzik S. Is fresh frozen plasma overtransfused in
the United States? Transfusion 2004;44:1674-5.
3. Pantanowitz L, Kruskall MS, Uhl L. Cryoprecipitate. Pat-
terns of use. Am J Clin Pathol 2003;119:874-81.
4. Maskens C, Downie H, Wendt A, et al. Hospital-based
transfusion error tracking from 2005 to 2010: identifying
the key errors threatening patient transfusion safety.
Transfusion 2014;54:66-73.
5. The Joint Commission and the American Medical
Association-Convened Physician Consortium for Perfor-
mance Improvement (PCPI). Proceedings from the
National Summit on Overuse. 2012. [cited 2014 Dec 7].
Available from: http://www.jointcommission.org/assets/1/
6/National_Summit_Overuse.pdf
6. Vamvakas EC, Blajchman MA. Transfusion-related
mortality: the ongoing risks of allogeneic blood transfusion
and the available strategies for their prevention. Blood
2009;113:3406-17.
7. Bennett-Guerrero E, Zhao Y, O’Brien SM, et al. Variation in
use of blood transfusion in coronary artery bypass graft
surgery. JAMA 2010;304:1568-75.
8. Sherwood MW, Wang Y, Curtis JP, et al. Patterns and out-
comes of red blood cell transfusion in patients undergoing
percutaneous coronary intervention. JAMA 2014;311:836-
43.
9. Rohde JM, Dimcheff DE, Blumberg N, et al. Health care–
associated infection after red blood cell transfusion: a sys-
tematic review and meta-analysis. JAMA 2014;311:1317-26.
10. Salpeter SR, Buckley JS, Chatterjee S. Impact of more
restrictive blood transfusion strategies on clinical
outcomes: a meta-analysis and systematic review. Am J
Med 2014;127:124-31.
11. Karp JK, Weston CM, King KE. Transfusion medicine in
American undergraduate medical education. Transfusion
2011;51:2470-9.
12. Duguid J, Copplestone JA, CMO’s National Blood Transfu-
sion Committee. Teaching transfusion in UK medical
schools: a survey by the National Blood Transfusion Com-
mittee. Med Educ 2008;42:439.
13. Kern DE, Thomas PA, Hughes MT, editors. Curriculum
development for medical education, a six-step
MEDICINE RESIDENT TRANSFUSION KNOWLEDGE
approach. Baltimore (MD): Johns Hopkins University Press;
2009.
14. O’Brien KL, Champeaux AL, Sundell ZE, et al. Transfusion
medicine knowledge in Postgraduate Year 1 residents.
Transfusion 2010;50:1649-53.
15. Arinsburg SA, Skerrett DL, Friedman MT, et al. A survey to
assess transfusion medicine education needs for clinicians.
Transfus Med 2012;22:44-9.
16. Saidenberg E, Pugh D. The use of an objective structured
clinical examination to assess internal medicine residents’
transfusion knowledge. Transfusion 2014;54:1537-41.
17. Graham J, Grant-Casey J, Alston R, et al. Assessing transfu-
sion competency in junior doctors: a retrospective cohort
study. Transfusion 2014;54:128-36.
18. Strauss RG. Transfusion medicine education in medical
school: only the first of successive steps to improving
patient care. Transfusion 2010;50:1632-5.
19. Haspel RL, Lin Y, Fisher P, et al. Development of a vali-
dated exam to assess physician transfusion medicine
knowledge. Transfusion 2014;54:1225-30.
20. Tavakol M, Dennick R. Psychometric evaluation of a
knowledge based examination using Rasch analysis: an
illustrative guide: AMEE guide no. 72. Med Teach 2013;35:
e838-48.
21. De Champlain AF. A primer on classical test theory and
item response theory for assessments in medical educa-
tion. Med Educ 2010;44:109-17.
22. Boone WJ, Townsend JS, Staver J. Using Rasch theory to
guide the practice of survey development and survey data
analysis in science education and to inform science reform
efforts: an exemplar utilizing STEBI self-efficacy data. Sci
Educ 2011;95:258-80.
23. National Cancer Institute. US cooperative groups. 2014
[cited 2014 Sep 30]. Available from: http://www.cancer
.gov/clinicaltrials/international/collaborate/
cooperativegroups/uscooperativegroups
24. Rinder HM, Grimes MM, Wagner J, et al. Senior pathology
resident in-service examination scores correlate with out-
comes of the American Board of Pathology certifying
examinations. Am J Clin Pathol 2011;136:
499-506.
25. Windish DM, Huot SJ, Green ML. Medicine residents’
understanding of the biostatistics and results in the
medical literature. JAMA 2007;298:1010-22.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article at the publisher’s web-site:
Appendix S1. Survey portion of assessment tool.
Volume **, ** ** TRANSFUSION 7