Abstract
Background
It is uncertain how transfusion knowledge translates to practice. The purpose of the study was to determine if higher scores on a validated Transfusion Camp knowledge assessment test were associated with transfusion order appropriateness.
Study Design and Methods
Eligible participants included postgraduate trainees and faculty physicians who had prescribed at least four transfusion orders in the preceding 6 months at two hospitals. Participant data and knowledge were collected using a web‐based questionnaire with a validated Transfusion Camp knowledge assessment tool. The most recent 4–10 consecutive transfusion orders per prescriber were independently dually adjudicated for appropriateness based on published criteria. The primary outcome was the correlation between the score on six questions on red blood cells (RBCs), platelets (PLTs), and plasma from the validated test and the percentage order appropriateness. Generalized linear regression was conducted to determine if factors (sex, specialty, participation in Transfusion Camp, previous transfusion education, self‐rated knowledge) were associated with appropriate orders.
Results
Seventy‐four participants (45 trainees, 29 faculty; 31 females, 43 males) completed the test. Median score was 66.7% (interquartile range [IQR]: 50.0, 83.3) for six questions on RBCs, PLTs, and plasma transfusions. Of 546 transfusion orders adjudicated, appropriateness was 90.7% (95% confidence interval [CI]: 87.9%–93.0%). The correlation between prescriber test scores and order appropriateness was very weak (r = −.08). In multivariable analysis, female prescribers (p = .02) and beginner (vs. intermediate) self‐rated knowledge (p = .01) were associated with higher transfusion appropriateness.
Conclusion
Transfusion knowledge test scores did not correlate with order appropriateness. Factors other than knowledge are key to understanding how to improve appropriate blood use.
Keywords: test, Transfusion Camp, validation
Abbreviations
- CI
confidence interval
- FP
frozen plasma
- IQR
interquartile range
- PLT
platelet
- RBC
red blood cell
- SD
standard deviation
1. BACKGROUND
Blood transfusions are a very common procedure, with at least 890,000 units of blood transfused per year in Canada, and over 16 million in the United States. Various specialties prescribe and transfuse blood products, 1 with evidence indicating that clinicians lack knowledge of key principles of transfusion medicine. 2 , 3 , 4 , 5 , 6 Recent audits of transfusion orders show that a large proportion of red blood cell (RBC), 7 platelet (PLT), 8 and plasma transfusions 9 are unnecessary. Given the inherent risks of transfusion such as allergic reactions, volume overload, and infectious complications, 10 it is important that transfusions are prescribed appropriately using evidence‐based guidelines.
A transfusion education program, Transfusion Camp, was established in Canada and implemented in 2012 to address trainee knowledge gaps in transfusion medicine. Coordinated through the University of Toronto and Canadian Blood Services, Transfusion Camp is delivered over the academic year to postgraduate medical trainees across various specialties. More than 400 medical trainees attend sessions annually. The impact of Transfusion Camp on trainee transfusion medicine knowledge and practice has been evaluated. 11 , 12 , 13 , 14 , 15 , 16 , 17 In the largest study, the pretest and posttest were administered to 390 trainees from 16 different specialties who attended at least 1 day of Transfusion Camp. On the posttest, participants reported overall improved knowledge and ability to manage various transfusion scenarios. An improved score on the posttest was seen in 79% of attendees, with higher attendance (attending 4 or 5 days vs. 1–3 days of the 5‐day Camp) being associated with higher test scores. 16 In a retrospective analysis, 83% of Transfusion Camp trainees reported that Transfusion Camp had made an impact on their transfusion practice. 17
Although Transfusion Camp has demonstrated improved trainee knowledge, attitudes, and confidence in transfusion medicine, an objective assessment has not yet been done to evaluate the impact of Transfusion Camp on prescribing behavior. This study would provide initial objective evidence that transfusion medicine knowledge has an impact on the appropriateness of trainee and faculty transfusion decisions. The primary objective of this study was to determine whether higher scores on the validated Transfusion Camp knowledge assessment test correlated with appropriateness of transfusion orders.
2. STUDY DESIGN AND METHODS
2.1. Participants and study design
A cross‐sectional observational study was performed at two centers to correlate scores on a validated transfusion knowledge assessment test 11 , 14 with appropriate transfusion prescribing for RBCs, PLTs, and frozen plasma (FP). The study was approved by institutional research ethics boards, and written informed consent was obtained from all participants. Using submitted electronic transfusion orders through hospital and blood bank systems at the participating hospitals, postgraduate trainees, hospitalists, or staff physicians on rotations in medicine, critical care, emergency and/or surgery who had prescribed at least four transfusion orders in the preceding 6 months were invited to participate in the study. Participant invitations were sent via email outlining the study aims, objectives, and methodology. Reminders were sent out every 2 weeks.
Participant data and knowledge assessment were collected using a web‐based survey tool (SurveyMonkey). The questionnaire collected baseline demographics including age, sex, specialty, postgraduate year level, previous transfusion medicine education, previous participation in transfusion camp, and self‐reported knowledge and ability in transfusion. The knowledge assessment test was based on the original BEST‐TEST, 11 modified for Transfusion Camp and validated using Rasch analysis showing the item in‐fit and out‐fit were both 1.0 14 ; it consisted of 20 multiple‐choice questions and required an estimated 30 minutes to complete. The knowledge assessment test is available upon request from the corresponding author.
Following questionnaire and knowledge assessment test completion, the prescriber's most recent transfusion orders (minimum 4 to maximum 10) prior to the questionnaire were adjudicated including those requesting either single or multiple components at a time (e.g., two RBC units ordered without a repeat hemoglobin in between). Orders were excluded if they were intended for a patient in an operating room and/or during a massive transfusion protocol, or if the ordering prescriber was unable to be accurately identified after review of the original order.
A medical chart audit was conducted by the study coordinators on each transfusion order based on modified adjudication criteria formulated from previous adjudication studies for RBCs, 7 PLTs, 8 , 18 and FP 19 (Supporting Information Appendices A–C). Transfusion medicine specialists (BT, KP, YL) then dually and independently adjudicated each transfusion order for appropriateness in REDCap. A discrepancy was noted as a difference in adjudication of appropriate versus inappropriate order. If there was a discrepancy in classification, both reviewers would perform an independent repeat review of the case. Any further disagreements were resolved by consensus. 20 , 21
2.2. Outcome measures
The primary outcome was correlation between participants' test scores for six questions on the knowledge assessment test specifically relating to indications for RBC, PLT, and FP transfusion and overall percentage of transfusion orders adjudicated as appropriate. The secondary outcome was correlation between participants' overall test score on 20 questions (which also covered other topics including transfusion reactions and blood group compatibility) and percentage of transfusion orders adjudicated as appropriate. The analyses also included the association of prescriber factors (e.g., sex [male vs. female], specialty [hematology vs. non‐hematology], faculty vs. trainee, participation in Transfusion Camp, previous transfusion education [3+ vs. 0–2 h], and self‐rated transfusion knowledge [beginner vs. intermediate vs. advanced/expert]) with the outcomes of knowledge assessment test scores and percentage of appropriate transfusion orders.
2.3. Statistical analysis
Descriptive analyses were conducted for the characteristics of total prescribers, and by group of faculty and trainees. Chi‐square test or Fisher's exact test were applied for comparing demographics between faculty and trainees. Wilcoxon rank‐sum test was used to compare transfusion order appropriateness between faculty and trainees. The Cohen's kappa coefficient was calculated for the dual adjudication of transfusion order appropriateness. Spearman's rank correlation coefficient r was calculated between participants' percentage test score and percentage of appropriate transfusion orders. Strength of the correlation was defined as: very strong: r = .80–1.00; strong: r = .60–.79; moderate: r = .40–.59; weak: r = .20–.39; very weak: r < .19. Univariate and multivariable generalized linear regression analyses were performed to search for significant predictive factors of transfusion appropriateness for all product orders, RBCs and PLT orders. Backward stepwise selection procedure was used in the multivariable analysis. All potential significant factors with p < .10 in the univariate analysis were included in the selection procedure. The final multivariable model only included factors with a significance of p < .05. A post hoc analysis was conducted to explore the effect of prescriber sex (female vs. male) on transfusion order appropriateness including comparison of demographics and transfusion order appropriateness between female and male prescribers, using the statistical analyses described above.
The sample size of 74 participants was calculated for the primary outcome to achieve 90% power to detect an increased correlation coefficient of .60 (strong correlation) from the null hypothesis correlation of .30 (weak correlation), using a 2‐sided test with a significance level of .05. All analyses were conducted using Statistical Analysis Software (SAS version 9.2, Cary, NC) and R package (version 4.3.1). A p‐value <.05 was considered statistically significant.
3. RESULTS
Between January 2022 and December 2022, 74 participants (45 trainees, 29 faculty) provided consent and completed the transfusion knowledge assessment test (Table 1). The most common specialties participating were internal medicine (32%), critical care (26%) and surgery (14%). Eight (11%) participants had previously participated in Transfusion Camp during their medical training. Participants rated their self‐assessed knowledge in transfusion medicine as beginner (female 45% vs. male 19%), intermediate (female 48% vs. male 60%), advanced (female 7% vs. male 19%), and expert (female 0% vs. male 2%) (female vs. male distribution, p = .047). The median score on the test was 66.7% (interquartile range (IQR): 50.0, 83.3) for the 6 questions that were specific to the topic of RBCs, PLTs, and plasma transfusion appropriateness and 65.0% (IQR: 55.0, 75.0) for the comprehensive 20‐question examination (Table 2). In multivariable analysis, average test scores were higher among faculty than in trainees (p < .0001), and in participants that had previously attended Transfusion Camp (p = .0001) (data not shown).
TABLE 1.
Prescriber demographics, faculty, and trainees.
| Total (N = 74) | Faculty (N = 29) | Trainees (N = 45) | p‐value | |
|---|---|---|---|---|
| Age categories, n (%) | <.0001 | |||
| 21–25 | 1 (1) | 0 (0) | 1 (2) | |
| 26–30 | 26 (35) | 1 (3) | 25 (56) | |
| 31–35 | 20 (27) | 5 (17) | 15 (33) | |
| 36–40 | 8 (11) | 5 (17) | 3 (7) | |
| >41 | 19 (26) | 18 (62) | 1 (2) | |
| Sex, n (%) | .58 | |||
| Female | 31 (42) | 11 (38) | 20 (44) | |
| Male | 43 (58) | 18 (62) | 25 (56) | |
| Current specialty training program, n (%) | .05 | |||
| Critical care | 19 (26) | 12 (41) | 7 (16) | |
| Emergency | 4 (5) | 2 (7) | 2 (4) | |
| Hematology | 4 (5) | 3 (10) | 1 (2) | |
| Internal medicine | 24 (32) | 6 (21) | 18 (40) | |
| Surgery | 10 (14) | 2 (7) | 8 (18) | |
| Other a | 13 (18) | 4 (14) | 9 (20) | |
| Current level of training (postgraduate year), n (%) | <.0001 | |||
| PGY1 | 5 (7) | 0 (0) | 5 (11) | |
| PGY2 | 10 (14) | 0 (0) | 10 (22) | |
| PGY3 | 8 (11) | 0 (0) | 8 (18) | |
| PGY4 | 5 (7) | 0 (0) | 5 (11) | |
| PGY5 | 4 (5) | 0 (0) | 4 (9) | |
| PGY6 | 4 (5) | 0 (0) | 4 (9) | |
| Clinical fellows b | 9 (12) | 0 (0) | 9 (20) | |
| Hospitalist | 5 (7) | 5 (17) | 0 (0) | |
| Attending staff physician | 24 (32) | 24 (83) | 0 (0) | |
| Hours spent in education sessions specific to transfusion medicine during the medical school training | .14 | |||
| None | 7 (9) | 5 (17) | 2 (4) | |
| 1 h | 18 (24) | 7 (24) | 11 (24) | |
| 2 h | 23 (31) | 6 (21) | 17 (38) | |
| 3 h | 9 (12) | 2 (7) | 7 (16) | |
| 4+ h | 17 (23) | 9 (31) | 8 (18) | |
| Hours spent in education sessions specific to transfusion medicine during the postgraduate residency training | .58 | |||
| None | 14 (19) | 4 (14) | 10 (22) | |
| 1 h | 13 (18) | 5 (17) | 8 (18) | |
| 2 h | 14 (19) | 5 (17) | 9 (20) | |
| 3 h | 8 (11) | 2 (7) | 6 (13) | |
| 4+ h | 25 (34) | 13 (45) | 12 (27) | |
| Participated in transfusion camp | .47 | |||
| Yes | 8 (11) | 2 (7) | 6 (13) | |
| No | 66 (89) | 27 (93) | 39 (87) | |
| Rate my knowledge of transfusion medicine as | .63 | |||
| Beginner | 22 (30) | 7 (24) | 15 (33) | |
| Intermediate | 41 (55) | 17 (59) | 24 (53) | |
| Advanced | 10 (14) | 4 (14) | 6 (13) | |
| Expert | 1 (1) | 1 (4) | 0 (0) | |
| To provide appropriate care to patients in your practice, how important is knowledge of transfusion medicine? | .87 | |||
| Not at all | 1 (1) | 0 (0) | 1 (2) | |
| Moderately | 19 (26) | 8 (28) | 11 (24) | |
| Very | 31 (42) | 12 (41) | 19 (42) | |
| Extremely | 23 (31) | 9 (31) | 14 (31) | |
Note: Significance of the bolded values are p < 0.05.
Other includes specialties not listed above: anesthesiology, cardiology, family medicine, hospital medicine, medical oncology, nephrology, neurology, plastic surgery, obstetrics and gynecology, psychiatry.
Clinical fellows were considered to be trainees for this study.
TABLE 2.
Transfusion knowledge assessment test scores.
| Total (N = 74) | Faculty (N = 29) | Trainees (N = 45) | p‐value | |
|---|---|---|---|---|
| Percentage test scores from 6 product‐specific exam questions | .0008 | |||
| Mean ± SD | 69.1 ± 19.8 | 78.7 ± 17.8 | 63.0 ± 18.8 | |
| Median (IQR) | 66.7 (IQR: 50.0, 83.3) | 83.3 (66.7, 100.0) | 66.7 (50.0, 83.3) | |
| Percentage test scores from all 20 exam questions | .005 | |||
| Mean ± SD | 64.1 ± 15.7 | 70.9 ± 12.8 | 59.8 ± 16.0 | |
| Median (IQR) | 65.0 (IQR: 55.0, 75.0) | 70.0 (65.0, 80.0) | 60.0 (50.0, 70.0) | |
Note: Significance of the bolded values are p < 0.05.
Abbreviation: IQR, interquartile range.
In total, 546 transfusion orders were adjudicated for 259 patients. Among these 416, (76%) were RBC orders, 114 (21%) PLT orders, and 15 (3%) plasma orders (Table 3). Eighty‐nine percent of RBC orders were for a single unit transfusion. All orders were adjudicated, even if they did not result in a transfusion; 97% of RBCs, 99% of PLTs, and 100% of plasma components ordered were transfused to patients. Appropriateness was 90.7% (95% confidence interval [CI]: 87.9%–93.0%) across all orders, with no significant difference between faculty and trainee results. Cohen's kappa coefficient for dual adjudication for appropriateness was 90.7% (95% CI: 84.7%–96.7%). The most common reason for both RBC and PLT orders being considered inappropriate was transfusion above the suggested thresholds in stable, non‐bleeding patients without active coronary artery disease (Hb ≥7 g/dL and PLT count >10 × 109/L, respectively). Transfusion order appropriateness by prescriber is shown in Table 4. There was no difference in number of orders evaluated per faculty versus trainee.
TABLE 3.
Transfusion orders.
| Type of product | |||
|---|---|---|---|
| Transfusion orders | RBC | Platelet | Plasma |
| N = 546 | N = 416 | N = 114 | N = 15 |
| Number of units ordered, n (%) | |||
| 1 | 372 (89) | 113 (99) | 0 (0) |
| 2 | 42 (10) | 1 (1) | 5 (33) |
| 3 | 2 (1) | 0 (0) | 8 (53) |
| 4 | 0 (0) | 0 (0) | 2 (13) |
| Number of units transfused, n (%) | |||
| 0 | 18 (4) | 2 (2) | 1 (7) |
| 1 | 358 (86) | 111 (97) | 1 (7) |
| 2 | 38 (9) | 1 (1) | 5 (33) |
| 3 | 1 (0.2) | 0 (0) | 6 (40) |
| 4 | 0 (0) | 0 (0) | 2 (13) |
Abbreviation: RBC, red blood cell.
TABLE 4.
Transfusion order appropriateness by prescriber.
| In 74 prescribers | Total prescribers (N = 74) | Faculty (N = 29) | Trainees (N = 45) | p‐value |
|---|---|---|---|---|
| Transfusion orders (of any product) | ||||
| Number of prescribers | 74 | 29 | 45 | |
| Median number of orders per prescriber (IQR) | 7 (5, 10) | 8 (5, 10) | 7 (5, 10) | .35 |
| Median percent appropriateness by prescriber (IQR) | 100.0 (85.7, 100.0) | 100.0 (85.7, 100.0) | 100.0 (85.7, 100.0) | .63 |
| RBC transfusion orders | ||||
| Number of prescribers | 74 | 29 | 45 | |
| Median number of orders per prescriber (IQR) | 6 (4, 7) | 6 (4, 7) | 5 (4, 7) | .99 |
| Median percent appropriateness by prescriber (IQR) | 100.0 (90.0, 100.0) | 100.0 (90.0, 100.0) | 100.0 (100.0, 100.0) | .83 |
| Platelet transfusion orders | ||||
| Number of prescribers | 47 | 23 | 24 | |
| Median number of orders per prescriber (IQR) | 2 (1, 3) | 2 (1, 4) | 2 (1, 3) | .76 |
| Median percent appropriateness by prescriber (IQR) | 100.0 (66.7, 100.0) | 100.0 (66.7, 100.0) | 100.0 (70.8, 100.0) | .63 |
| Plasma transfusion | ||||
| Number of Prescribers | 12 | 5 | 7 | |
| Median number of orders per prescriber (IQR) | 1 (1, 1.5) | 1 (1, 1) | 1 (1, 2) | .83 |
| Median percent appropriateness by prescriber (IQR) | 100.0 (25.0, 100.0) | 100.0 (100.0, 100.0) | 100.0 (0.00, 100.0) | .56 |
Abbreviations: IQR, interquartile range; RBC, red blood cell.
For the primary and secondary outcomes, there was no correlation between test scores and overall appropriateness of transfusion orders. The Spearman correlation coefficient (r) was −.08 (very weak) for the correlation between overall appropriateness of transfusion orders and the 6 blood component‐related questions, and −.09 for correlation with the total 20 questions.
In univariate analysis, factors found to be associated with higher transfusion order appropriateness for any product were female (vs. male) prescribers (96.0% vs. 88.6%, p = .001), and self‐rated transfusion knowledge as beginner versus advanced/expert (97.6% vs. 88.7%; p = .01) or intermediate (97.6% vs. 89.3%; p = .001) (Table 5). In the multivariable analysis, female prescribers (p = .02) and beginner (vs. intermediate) self‐rated transfusion knowledge (p = .01) were associated with higher transfusion appropriateness.
TABLE 5.
Univariate, multivariable analysis of transfusion order appropriateness by prescriber factors.
| Outcome: percentage of appropriate transfusion orders for any product (N = 546) | Estimate | SE | p‐value |
|---|---|---|---|
| Sex (male vs. female) | −0.08 | 0.02 | .001 |
| Specialty (hematology vs. nonhematology) | −0.03 | 0.06 | .55 |
| Faculty versus trainees | −0.01 | 0.02 | .71 |
| Participation in transfusion camp (yes vs. no) | −0.02 | 0.04 | .57 |
| Previous transfusion education during the medical school training (3+ vs. 0–2 h) | −0.01 | 0.03 | .67 |
| Previous transfusion education during the postgraduate residency training (3+ vs. 0–2 h) | −0.02 | 0.02 | .46 |
| Self‐rated transfusion knowledge (overall effect) | .003 | ||
| Beginner versus advanced/expert | 0.10 | 0.04 | .01 |
| Intermediate versus advanced/expert | 0.01 | 0.04 | .86 |
| Beginner versus intermediate | 0.09 | 0.03 | .001 |
| Multivariable model | Estimate | SE | p‐value |
|---|---|---|---|
| Sex (male vs. female) | −0.06 | 0.03 | .02 |
| Self‐rated transfusion knowledge (overall effect) | .03 | ||
| Beginner versus advanced/expert | 0.07 | 0.04 | .09 |
| Intermediate versus advanced/expert | 0.005 | 0.04 | .90 |
| Beginner versus intermediate | 0.07 | 0.03 | .01 |
Note: For the continuous outcome of percentage of appropriate transfusion orders for any product, the estimate is calculated as the slope of the independent variable. For example, in the multivariable analysis, female participants had 6% higher transfusion appropriateness than male participants. Significance of the bolded values are p < 0.05.
For RBC transfusion orders, the only factor found to be associated with higher appropriateness was female prescribers (vs. males; 98.0% vs. 92.5%, p = .02) in both univariate and multivariable analyses (Supporting Information Appendix Table 1). For PLTs, factors associated in the multivariable model with higher appropriateness included: female prescribers (vs. males; 84.1% vs. 77.1%; p = .01), hematology specialty (vs. non‐hematology; p < .0001), participation in Transfusion Camp (p = .0004), 0–2 h of transfusion education (vs. 3+ h; p < .0001), and beginner self‐rated transfusion knowledge (vs. intermediate; 96.2% vs. 77.2%; p < .0001 and vs. advanced/expert; 96.2% vs. 72.6%; p = .004) (Supporting Information Appendix Table 2). There were insufficient numbers of plasma orders to analyze for associated factors.
Given the finding that female prescribers were associated with higher overall transfusion order appropriateness, post hoc analyses were performed (Supporting Information Appendix Tables 3 and 4). Female prescribers were more likely to rate their knowledge as beginners. There was no difference in number of orders evaluated per female versus male prescriber. As noted above, female prescribers had higher transfusion appropriateness for any product (p = .006) and for RBC transfusion (p = .007).
4. DISCUSSION
We found that results on the validated knowledge assessment test did not significantly correlate with transfusion order appropriateness in either faculty or trainees. This was consistent whether analyzing the 6 questions on RBCs, PLTs, or plasma or all 20 questions in the validated knowledge assessment test. A lack of correlation could be explained by the knowledge assessment questions not being able to capture the clinical depth of knowledge associated with real‐life cases, where nuanced patient and laboratory‐based factors may affect transfusion thresholds and decision making. An alternative possibility is that the effects of individual transfusion knowledge were masked by the overall high rates of appropriate transfusion orders (90.7%), which may not be generalizable to other regions. The level of transfusion appropriateness observed in this study was high relative to other audits of RBC and PLT transfusion. 7 , 8 , 9 Both centers evaluated in this study incorporate prospective transfusion order screening, have dedicated transfusion specialists on site, and actively promote guideline‐based transfusion care, which may have influenced ordering practice beyond the individual's transfusion knowledge.
Key findings were the higher rate of overall transfusion order appropriateness seen with prescribers of female sex and those that had self‐rated as having a beginner level of transfusion knowledge (vs. intermediate or advanced). Interestingly, in this study, female prescribers rated their knowledge more often as beginner than their male counterparts. Having a self‐described lower level of knowledge may have led to stricter and more rigid adherence to standardized practices before ordering blood components, and less use of liberal transfusion as an intervention. Previous research has demonstrated that clinical care provided by female prescribers may lead to better overall outcomes. 22 This is an understudied aspect of clinical medicine that warrants further exploration, especially considering the presence of systemic inequity toward females in medicine. 23
Several subgroup populations also performed higher overall on the knowledge assessment test. Faculty performing higher than trainees regardless of their specialty can be explained by multiple potential factors, including clinical experience, mandatory continuing education maintenance courses, repeated feedback on transfusion orders, and having spent greater time in hospital sites with adherence to transfusion guidelines. Higher transfusion knowledge assessment scores also included individuals that had previously participated in Transfusion Camp. Only 10% of prescribers (six of eight were trainees) had previously participated in Transfusion Camp. The questionnaire did not collect the time since these eight individuals had participated in Transfusion Camp; however, since six of these were trainees, they likely attended Transfusion Camp within the past 5 years. This might have led to recall bias as they had previously taken a similar knowledge assessment test. Alternatively, this might indicate that involvement in Transfusion Camp provides an important foundation of knowledge not otherwise covered in undergraduate or advanced medical education.
Important strengths to this study include our ability to adjudicate a large volume of transfusion orders using up‐to‐date guideline‐based parameters to help determine appropriateness. The availability of an electronic‐based ordering system allowed us to determine the indication and compare with the most recent available lab test. Our study framework also allowed us to determine appropriateness based on both dosage (single vs. multiple component orders) as well as transfusion thresholds, which are important considerations in the modern era of blood conservation and best practices. We were also able to recruit a diverse group of practitioners including trainees and faculty from various medical subspecialties. This study can easily be expanded to include additional centers to further assess the role of the knowledge assessment test in affecting clinical decision making.
Several limitations within our study may impact its analysis and interpretation. The overall distribution of our adjudicated blood components was predominantly composed of RBCs (76%), and PLT (21%) orders. Only 3% of all orders represented plasma transfusions, which is known to be frequently inappropriately prescribed, 9 , 19 thus limiting our ability to detect overall differences in appropriateness for plasma transfusion. Greater signals of difference between the subgroups may have become apparent had plasma been more greatly represented. For PLTs, there were multiple factors which should be considered hypothesis generating given the lower number of PLT orders. Certain clinical settings were excluded such as intraoperative transfusion and massive hemorrhage protocols, in which transfusion may be dictated by clinical parameters and use of rotational thromboelastometry. Despite all orders being associated with an ordering clinician, it was also not possible to determine whether the decision for ordering transfusion of a blood component was made on behalf of more senior medical staff (e.g., attending staff in the case of a trainee) or in a team setting, which may have impacted the ability to detect differences in appropriateness between trainee and faculty prescribers. For example, trainees may be more often responsible for routine transfusion orders, whereas faculty may order transfusion in situations with high acuity or sudden patient decompensation, reflecting situations that are not often outlined by guidelines. Lastly, our results came from only two centers affiliated with the same university. Despite these centers both comprising a broad variety of clinical exposures including trauma, vascular and cardiovascular surgery, and hematologic malignancies, it is possible that the low number of centers limits generalizability of our results.
The limitations of this study serve to inform future studies aiming to correlate transfusion medicine knowledge assessment scores with clinical transfusion ordering practice. These include conducting the study in a center that does not have prospective transfusion order screening which may influence practice beyond the prescriber's knowledge, and extending the number of orders included per prescriber to increase the representation of the prescriber's transfusion ordering practice.
5. CONCLUSION
In our study, we were unable to demonstrate a correlation between transfusion knowledge assessment test scores and appropriateness of transfusion orders in both trainees and faculty prescribers. Factors that were associated with higher knowledge test scores included faculty in comparison to trainees, and prior participation in Transfusion Camp seminars. Female prescriber sex was associated with higher overall transfusion order appropriateness. Further research into the factors affecting clinical transfusion practice is warranted in order to better understand and facilitate better practice.
CONFLICT OF INTEREST STATEMENT
JC has research support from Canadian Blood Services and Octapharma. YL has research support from Canadian Blood Services and Octapharma and is a consultant with Choosing Wisely Canada. The remaining authors declare no conflicts to interest.
Supporting information
Data S1. Supporting Information.
ACKNOWLEDGMENTS
We would like to acknowledge Liying Zhang for statistical support. Financial support for this research is provided by the University of Toronto, Alexandra Yeo Chair Grant in Benign Hematology, and Canadian Blood Services (Transfusion Medicine Research Program Support Award), funded by the federal government (Health Canada) and the provincial and territorial ministries of health. The views herein do not necessarily reflect the views of Canadian Blood Services or the federal, provincial, or territorial governments of Canada.
Tordon B, Meirovich H, Malkin A, Pavenski K, Moorehead A, Ginsborg L, et al. Correlation of the Transfusion Camp knowledge assessment test with clinical transfusion practice. Transfusion. 2024;64(12):2371–2379. 10.1111/trf.18035
REFERENCES
- 1. Shehata N, Forster A, Lawrence N, Rothwell DM, Fergusson D, Tinmouth A, et al. Changing trends in blood transfusion: an analysis of 244,013 hospitalizations. Transfusion. 2014;54(10 Pt 2):2631–2639. [DOI] [PubMed] [Google Scholar]
- 2. Saidenberg E, Pugh D. The use of an objective structured clinical examination to assess internal medicine residents' transfusion knowledge. Transfusion. 2014;54(6):1537–1541. [DOI] [PubMed] [Google Scholar]
- 3. Arinsburg SA, Skerrett DL, Friedman MT, Cushing MM. A survey to assess transfusion medicine education needs for clinicians. Transfus Med. 2012;22(1):44–51. [DOI] [PubMed] [Google Scholar]
- 4. Rahav Koren R, Suriu C, Yakir O, Akria L, Barhoum M, Braester A. Physicians' lack of knowledge ‐ a possible reason for red blood cell transfusion overuse? Isr J Health Policy Res. 2017;6(1):49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kasraian L, Tavassoli A. A survey of resident physicians' knowledge concerning transfusion medicine in Shiraz, Iran. Asian J Transfus Sci. 2014;8(2):118–120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Graham J, Grant‐Casey J, Alston R, Baker P, Pendry K. Assessing transfusion competency in junior doctors: a retrospective cohort study. Transfusion. 2014;54(1):128–136. [DOI] [PubMed] [Google Scholar]
- 7. Spradbrow J, Cohen R, Lin Y, Armali C, Collins A, Cserti‐Gazdewich C, et al. Evaluating appropriate red blood cell transfusions: a quality audit at 10 Ontario hospitals to determine the optimal measure for assessing appropriateness. Transfusion. 2016;56(10):2466–2476. [DOI] [PubMed] [Google Scholar]
- 8. Hill‐Strathy M, Pinkerton PH, Thompson TA, Wendt A, Collins A, Cohen R, et al. Evaluating the appropriateness of platelet transfusions compared with evidence‐based platelet guidelines: an audit of platelet transfusions at 57 hospitals. Transfusion. 2021;61(1):57–71. [DOI] [PubMed] [Google Scholar]
- 9. Khandelwal A, Minuk L, Liu Y, Arnold DM, Heddle NM, Barty R, et al. Plasma transfusion practices: A multicentre electronic audit. Vox Sang. 2022;117(10):1211–1219. [DOI] [PubMed] [Google Scholar]
- 10. Kracalik I, Mowla S, Basavaraju SV, Sapiano MRP. Transfusion‐related adverse reactions: data from the National Healthcare Safety Network Hemovigilance Module—United States, 2013–2018. Transfusion. 2021;61(5):1424–1434. [DOI] [PubMed] [Google Scholar]
- 11. Haspel RL, Lin Y, Fisher P, Ali A, Parks E, Biomedical Excellence for Safer Transfusion (BEST) Collaborative . Development of a validated exam to assess physician transfusion medicine knowledge. Transfusion. 2014;54(5):1225–1230. [DOI] [PubMed] [Google Scholar]
- 12. Haspel RL, Lin Y, Mallick R, Tinmouth A, Cid J, Eichler H, et al. Internal medicine resident knowledge of transfusion medicine: results from the BEST‐TEST international education needs assessment. Transfusion. 2015;55(6):1355–1361. [DOI] [PubMed] [Google Scholar]
- 13. Wheeler AP, Delaney M, Fung M, Gorlin J, Kutner JM, Lam JCM, et al. Pediatric resident knowledge of transfusion medicine: results from the BEST‐TEST3 international education needs assessment. Transfusion. 2021;61(8):2487–2495. [DOI] [PubMed] [Google Scholar]
- 14. Lin Y, Khandelwal A, Kapitany C, Chargé S. Transfusion Camp: successes and challenges in scaling and expanding a transfusion medicine education program. Transfus Apher Sci. 2022;103629:103629. [DOI] [PubMed] [Google Scholar]
- 15. Lin Y, Cserti‐Gazdewich C, Callum J. University of Toronto Transfusion Camp Organizing C. Evaluation of “Transfusion Camp,” a postgraduate transfusion medicine education program using the BEST‐TEST knowledge assessment tool. Transfusion. 2015;55(8):2049–2051. [DOI] [PubMed] [Google Scholar]
- 16. Lin Y, Tilokee E, Charge S, Alam A, Cserti‐Gazdewich C, Lau W, et al. Transfusion Camp: a prospective evaluation of a transfusion education program for multispecialty postgraduate trainees. Transfusion. 2019;59(6):2141–2149. [DOI] [PubMed] [Google Scholar]
- 17. Yeung KCY, Kapitany C, Chargé S, Callum J, Cserti‐Gazdewich C, D'Empaire PP, et al. Transfusion camp: a retrospective study of self‐reported impact on postgraduate trainee transfusion practice. Transfusion. 2023;63(4):839–848. [DOI] [PubMed] [Google Scholar]
- 18. Etchells M, Spradbrow J, Cohen R, Lin Y, Armali C, Lieberman L, et al. Audit of appropriate use of platelet transfusions: validation of adjudication criteria. Vox Sang. 2018;113(1):40–50. [DOI] [PubMed] [Google Scholar]
- 19. Tinmouth A, Thompson T, Arnold DM, Callum JL, Gagliardi K, Lauzon D, et al. Utilization of frozen plasma in Ontario: a provincewide audit reveals a high rate of inappropriate transfusions. Transfusion. 2013;53(10):2222–2229. [DOI] [PubMed] [Google Scholar]
- 20. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O'Neal L, et al. The REDCap consortium: building an international community of software platform partners. J Biomed Inform. 2019;95:103208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata‐driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Tsugawa Y, Jena AB, Figueroa JF, Orav EJ, Blumenthal DM, Jha AK. Comparison of hospital mortality and readmission rates for Medicare patients treated by male vs female physicians. JAMA Intern Med. 2017;177(2):206–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Steffler M, Chami N, Hill S, Beck G, Cooper SC, Dinniwell R, et al. Disparities in physician compensation by gender in Ontario, Canada. JAMA Netw Open. 2021;4(9):e2126107. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1. Supporting Information.