Faculty Bibliography

BACKGROUND: Testing of platelets (PLTs) for bacterial contamination is required by the AABB Standards but is not fully standardized. On January 31, 2011, a new AABB Standard, 5.1.5.1.1, specified that bacterial detection methods for PLT components shall use assays either approved by the Food and Drug Administration (FDA) or validated to provide sensitivity equivalent to these FDA-approved methods. METHODS: An Internet-based survey of AABB member institutions was conducted from May to June 2012, to document current practices used in 2011 for bacterial detection in different PLT products and to assess the impact of the new standard. RESULTS: Of 1053 AABB member institutions surveyed, 40 of 99 blood centers (40.4%) and 184 of 954 hospital blood banks or transfusion services (19.3%) responded. Sixty-four respondents manufactured PLTs. Apheresis PLTs (APs) were predominantly screened with the BacT/ALERT system (89.5%); the majority (95.2%) were cultured with at least 8 mL of product. There was substantial variation in the minimum incubation time of cultures before release of PLTs (range, 0 to >24 hr). Recalls of released AP for possible bacterial contamination were largely successful (67.3%); successful interdiction before transfusion was associated with incubation for more than 12 hours before release (p < 0.01). After Standard 5.1.5.1.1 took effect, there was a decrease in production of whole blood-derived PLT concentrates (WBPCs). Point-of-issue ("rapid") immunoassays were used to screen a substantial proportion of WBPC PLTs, but were rarely used as secondary tests for previously cultured APs. CONCLUSION: The survey identified variability in culture methods and release times with AP, while use of WBPC decreased after AABB Standard 5.1.5.1.1 became effective.

Background/Case Studies: >25% of the 15 million RBC transfusions (TXN) in the US are probably medically unnecessary. Many studies have demonstrated the safety of using lower hemoglobin (Hgb) TXN thresholds. In 2010 the 11 acute care hospitals in our system adopted a common set of RBC TXN triggers to standardize practice and improve patient care. The adult triggers were A) Hgb <7 g/dl and symptomatic anemia, B) Hgb <9 g/dl with significant cardiac, neurological, or respiratory disease, C) Blood loss >20% refractory to fluid resuscitation, and D) Pre-op for major surgery with anticipated blood loss >1000 ml and Hgb <7 g/dl. Education was part of the roll-out of the new triggers. In 2011 we sought to evaluate the adherence to the triggers. Study Design/Methods: The pre-TXN Hgb result just prior to each RBC order was pulled from the EMR into a data warehouse where it could be manipulated. The number of RBC orders placed at pre-TXN Hgb values were sorted and grouped. Because of the way the messages come from the blood bank computer system into the EMR, it is not possible to electronically extract the Hgb just prior to RBC issuance. Thus we were forced to look at the Hgb just prior to RBC order as a surrogate for adherence to the RBC TXN triggers Results/Findings: From 1/1/11 to 10/31/11, 16,375 RBC orders were placed of which 13077 had a preceding Hgb level. 34,891 RBCs were transfused. 4474 orders were placed with a pre-TXN Hgb >/= 9 g/dl. 3481 orders were placed with a pre-TXN Hgb <7 g/dl. 5695 orders were placed with a pre-TXN Hgb 7- <9 g/dl. 9.6% of RBCs were ordered at a Hgb >10 g/dl. A portion ordered at higher Hgb levels were for pre-operative patients. 10% of RBCs issued were returned to the blood bank. 3% of RBCs were issued using emergency release, presumably for trauma and other rapidly bleeding patients, and thus may explain a portion of the RBC units ordered at Hgb >10 g/dl Conclusion: 27% of RBCs are being ordered and transfused to patients with Hgb >9 g/dl. The more restrictive RBC TXN triggers adopted have not eliminated unnecessary transfusions. While computers are able to sort blood orders based on the use of lab data and the physician selected TXN triggers, smarter computers which are able to assess lab values and supporting clinical documentation are needed. The blood bank needs to play a proactive role in ensuring that RBCs are not issued to patients who do not satisfy the specific TXN trigger selected by the ordering physician. Until such clinical support exists from the EMR, auditing must continue to ensure adherence to the RBC TXN triggers and prevent clinically unnecessary TXN. (Table presented)

Background/Case Studies: >25% of the 15 million RBC transfusions (TXN) in the US probably are medically unnecessary. Many studies have shown that restrictive transfusion is safe. In 2010 our institution adopted more restrictive RBC TXN triggers to minimize clinically unnecessary TXN. The adult triggers were A) Hgb <7 g/dl with symptomatic anemia, B) Hgb <9 g/ dl with significant cardiac, neurological, or respiratory disease, C) Blood loss >20% refractory to fluid resuscitation, D) Pre-op for major surgery with anticipated blood loss >1000 ml and Hgb <7 g/dl, E) Massive transfusion, and F) Other. Prospective auditing of orders began concurrently with the new triggers. Prior to issuing, the blood bank checks the patient's most recent Hgb in the EMR and verifies that the criteria are satisfied for all orders using trigger A or B. In 2011 we sought to evaluate the impact of the changes. Study Design/Methods: All RBC TXN orders from 1/1/11 to 12/31/11 were extracted from the hospital EMR. The orders were sorted by the TXN trigger selected by the ordering physician at the time of order placement. Because of the way the message comes from the blood bank computer system into the EMR, it is not possible to electronically extract the Hgb just prior to RBC issuance. Consequently we looked at the Hgb just prior to RBC order as a surrogate for adherence to the RBC TXN triggers. Results/ Findings: From 1/1/11 to 12/31/11, 8582 RBC orders were placed and 8482 RBCs were transfused. 13.2% had no pre-TXN Hgb. 21.7% were placed at Hgb <7 g/dl. 57.3% were placed at Hgb <9 g/dl. 21.1% were placed at Hgb >=10 g/dl. Despite selecting trigger A, and 72.9% of patients had a Hgb >=7 g/ dl at the time of order placement. Trigger F was selected in 26.4% of orders (Table). Conclusion: Despite having dropdown triggers built into order entry, many physicians opt to type in a justification using the "other" TXN trigger F. Roughly 50% of the "other" orders would be more appropriately ordered as TXN trigger D. Using our current prospective auditing procedure, 57.3% of RBC orders undergo lab value verification by the blood bank technologists prior to RBC issuance. By selecting TXN trigger C, D, E, or F, 42.7% of orders bypass prospective auditing thus preventing the blood bank from stopping medically unnecessary TXN. (Table presented)

Background/Case Studies: It is estimated that 25% of the 15+ million RBC units transfused annually in the US are clinically unnecessary. The Transfusion (TXN) Committee is charged with determining acceptable criteria and monitoring compliance with TXN triggers. All blood orders require the physician to select a TXN trigger. In 2010, new triggers were adopted. If "other" is selected, the physician must free text type the justification. Prospective auditing of some triggers began concurrently with the new triggers. After adoption, use of "other" increased. I sought to analyze the typed justifications for using "other." Study Design/Methods: The trigger entered with all orders placed from 1/1/11 to 12/31/11 was extracted from the EMR. "Other" orders were sorted. The typed reasons were analyzed and sorted into 13 categories. The hemoglobin (Hgb) level just prior to order placement was used as a surrogate for appropriateness of TXN. Results/Findings: From 1/1/11 to 12/31/11, there were 2308 RBC orders placed using "other." Although there is a pre-op trigger, 42.9% of orders were categorized as pre-op requests to be on hold for the OR. 12.1% were for active bleeding. 10.4% were for acute blood loss. 8.4% were for patients' undergoing chemotherapy and/or radiation treatment. 9.5% were for patients with symptomatic anemia of which 20.9% had Hgb > 10 g/dl (Table). Conclusion: "Other" was frequently selected even when the patient satisfied a specific approved RBC TXN trigger. Physicians appear to like to type their own reason for TXN rather than selecting it from a dropdown menu. A new TXN trigger with a higher Hgb level (8-9 g/dl) should be created for use in patients undergoing chemotherapy and/or radiation therapy. RBCs were unnecessarily transfused to patients who had Hgb levels > 10 g/dl who were not acutely bleeding or did not have acute blood loss. More education must be provided to the ordering physicians to reduce the number of unnecessary TXNs particularly in non-bleeding patients with Hgb > 10 g/dl. Consideration should be taken to eliminate the "other" category or limit its use. (Table presented)

Background/Case Studies: Most US medical school curriculums devote little attention to teaching transfusion medicine (TXNM). From early stages of their residency, however, physicians are expected to be able to obtain informed consent for transfusion, administer transfusions, and recognize and treat transfusion reactions. Although there are many transfusion standards designed to assure patient safety, they can also lead to a false sense of security. The overall limited knowledge of TXNM by most physicians and reliance on this false sense of security can adversely impact patient safety. 0.5-1% of all transfusions is expected to result in a reaction. In 2008 and 2009 targeted education was done. In 2010, all incoming house staff (HS) were required to take an online tutorial and pass an exam prior starting work on 7-1-10. The impact of education was assessed by evaluating the number of transfusion reactions reported pre- and post-training. Study Design/ Methods: A Transfusion Guidelines tutorial and competency exam, consisting of 25 true/false questions, was created. All HS were required to take and pass (>=80%) the exam. The instruction course and competency exam were accessible to all HS on the medical school's advanced learning exchange. The competency exams were instantly scored and reported to the test taker. The number of transfusion reactions reported between 7-1-10 and 3-31-11 was assessed and compared to the numbers between 4-1-07 and 12-31-08. Results/Findings: Prior to starting work on 7-1-10, all 380 incoming HS took and scored at least 80% on the exam. Approximately 34.5% of the 380/1100 HS in all departments have completed the training. From 4-1-07 to 12-31-08, 89 (average of 12.7 per quarter) transfusion reactions were reported versus the 140-280 that were expected. From 7-1-10 to 3-31-11, 1921 patients were transfused more than 11000 blood products and 23 (average of 7.6 per quarter) transfusion reactions were reported versus the 53-106 that were expected. Conclusion: Formalized education for all HS involved in the transfusion of blood must continue to be conducted. Administrative and Departmental support were necessary to assure existing HS took and passed the exam. Under recognition and/or under reporting of transfusion reactions continues to be a challenge to improving patient safety. Mandated education prior to starting as HS must be followed up by additional and frequent teaching

Introduction : Rapid infusion devices are becoming increasingly popular for the administration of warm fluids and blood in hypovolemic patients. A recently developed system, Thermacor 1200 (Smisson-Cartledge Biomedical LLC, Macon, GA), consists of a central device to which a disposable cartridge of fluid lines attaches. Performance characteristics of this device have yet to be evaluated. We compared the Thermacor 1200 with a currently utilized infusion device, FMS 2000 (Belmont Instrument Corp., Billerica, MA), to evaluate maximum flow rates, accuracy of actual versus set flow rates, fluid warming capabilities, and air bubble elimination. Methods : A ThermaCor 1200 and an FMS 2000, owned by our institution, were evaluated in vitro after being tested for proper functioning. FMS 2000 was tested with the packaged 4.5ft patient line, and Thermacor 1200 with packaged 3ft (TC3) and 6ft (TC6) patient lines. Maximum flow rates of lactated Ringer's (LR) and expired packed red blood cells (PRBCs) were measured with 22, 20, 18, 16, 14 and 8.5F gauge catheters, using a graduated cylinder and stopwatch. Flow rate accuracy was determined by comparing the actual versus displayed flow rates, for LR and PRBCs. Temperature was measured, at various flow rates, with an electronic probe (Wavetek 23XT, San Diego, CA) positioned 3cm from the distal port of the outflow tubing, for LR and PRBCs. Air elimination capability was determined, for LR only, by infusing fluid into an inverted 20mL syringe submerged in a bucket of water, and measuring the resulting air trapped in the syringe. All measurements were repeated six times. Data were analyzed using one-factor ANOVA, and the Tukey multiple comparisons method. Statistical significance was defined as p<0.05. Results : Maximum flow rates were higher with TC3 and TC6 than with FMS 2000 in most instances, especially when using larger catheter bores (Table 1). Flow rates were more accurate with TC3 and TC6 than with FMS 2000 for LR (1.4, 1.6, and 3.5% variance from target rate, respectively; p<.001) and for PRBCs (2.1, 2.6, and 5.9% variance from target rate, respectively; p<.001). Temperatures of delivered fluid were higher with TC3 and TC6 as compared to the FMS 2000 for LR (38.0, 37.8, and 36.8degreeC. respectively; p<.001) and PRBCs (38.2, 38.1, and 37.2degreeC, respectively; p<.001). Air was not detected in fluid infused from either device. Discussion : In this experiment, the performance of the Thermacor 1200, at both lengths of patient line, was superior to that of the FMS 2000 in that it infused LR and PRBCs at higher and more accurate flow rates, at higher temperatures. (Table presented)