Quality Improvement in Transfusion Medicine:
A Mayo Clinic Process to Reduce Incompatible Transfusions
Safe transfusion of blood requires the use of compatible blood products. Yet, despite rigorous testing requirements, acute hemolytic transfusion reactions still occur in the United States. In 2004, Sandler reported that fatal and nonfatal acute hemolytic transfusion reactions occurred with frequencies of 1 case per 250,000 to 600,000 population and 1 case per 6,000 to 33,000 population, respectively.1 Up to 20% of all incompatible transfusions can be traced to patient misidentification and/or specimen mislabeling errors before a specimen arrives in the laboratory.
Wrong blood in tube (WBIT, see sidebar) incidents can result in a patient ABO mistype, an incompatible transfusion, and an immediate, life-threatening transfusion reaction. In a recent blood bank safety practices survey of 122 clinical laboratories by Grimm et al, for the College of American Pathologists, it was found that WBIT occurs in about 4 in every 10,000 samples.2 Although rare, incompatible transfusions can cause significant patient morbidity and mortality. Patient deaths have been reported with as little as 10 mL of incompatible blood being transfused.3 Laboratory staff play a significant role in preventing transfusion errors by identifying WBIT and mislabeling errors and rejecting these specimens. Most errors are discovered during crosschecking of specimen information on the request forms and specimen tubes before the pretransfusion testings.
Correct patient identification and prevention of transfusion errors related to patient misidentification are both components of the first goal in The Joint Commission National Patient Safety Goals (TJC NPSG) for 2011. The purpose of the defined goals is to promote specific improvements in patient safety by highlighting problematic areas. The goals and recommendations are based on evidence- and expert-based solutions, focusing on a systems approach. To improve the accuracy of patient identification, TJC NPSG goal number 1 recommends, “the use of two patient identifiers when administering blood or blood components; when collecting blood samples or other specimens for clinical testing.”4
Figure 1. Patient sample on the right contains notable hemolysis immediately after infusion of ABO incompatible blood
Wrong Blood in Tube
Wrong blood in tube (WBIT) occurs when a specimen tube is labeled with unique identifiers for a patient, but the blood in the tube is from a different patient. This error can occur when a specimen is mislabeled (collected on the right patient but labeled as a different patient) or is collected from the wrong patient (and labeled as the patient on the order). WBIT errors are often identified by the laboratory when the patient’s ABO typing result is different from his or her historical ABO blood type. Historical ABO type is defined as a patient ABO type performed on a previous specimen, whose result is present in the medical record at the time of receipt of the most recent specimen submitted for testing.
AABB standards require 2 determinations of ABO type for recipients of blood where the computer crossmatch (see sidebar) is used to determine ABO incompatibility with donor blood. This requirement can be achieved by determination of ABO on the current sample and the second, by retesting the same sample, by testing a second current sample, or by comparison with previous records.5
In this article, we describe a process change—requiring a second sample for patients with no historical blood type on file to verify patients’ blood type before transfusion—that increases the ability to detect WBIT, reduces ABO incompatible transfusions, and raises the bar for patient identification safety.
A Mayo Clinic Transfusion Medicine Initiative to Improve the Safety of Blood and Blood Component Transfusions
When a review of our event management database showed an increasing WBIT rate, eliminating WBIT incidents was assigned a top priority for Mayo Clinic Transfusion Medicine. This necessitated a safety practice change to verify a patient’s blood type before transfusion. For patients with no historical blood type on file, ABO/Rh testing performed on a second sample was now required prior to transfusion.
Advocacy for change was carried forward to our institutional leadership to help support the change and ultimately, the change was endorsed by the Institutional Transfusion Committee. A multidisciplinary team was formed to identify key stakeholders, potential roadblocks and barriers, and the resources needed to implement this change. A formal change control procedure was initiated to outline the necessary process that would be needed to effect the change as well as for review and approval by leadership and the affected work units.6
Naturally, as with most major changes, there was resistance to implementing the practice change. The resistance was answered by providing several educational aids and presentations to help those affected by the change understand the impact on patient safety related to the change. A slide-based presentation describing what was changing, why we were doing this change, to whom the change applied, exceptions to the change, who would pay for the change, who approved this change, and what other institutions have made this change was developed and placed on the Mayo Clinic Transfusion Medicine intranet home page. A link to this presentation was written into various practice change memos and distributed to key stakeholders and practice chairs. The memo briefly outlined the same information as in the presentation and guided readers to use the intranet link to obtain more detailed information. The memo also provided the names of contacts if further questions arose. An institutional newsletter and a Mayo Clinic Department of Laboratory Medicine and Pathology newsletter each ran a feature story in the week just prior to implementation describing the new practice change. Procedures and protocols were updated for phlebotomy and transfusion medicine personnel and both groups were trained prior to the implementation date of November 16, 2010. Midmorning was chosen as the time for implementation because the number of patients seen in the phlebotomy areas at this time is generally lower than at other times during the workday. This allowed for potential problems encountered in the computer programming modifications or the new process without compromising patient schedules.
A computer crossmatch (also known as an electronic crossmatch) is defined as the assessment of donor and recipient blood compatibility by substituting a computerized record review for the serologic testing of recipient serum (or plasma) with donor RBCs. The computerized record review follows strict decision rules to determine recipient criteria and donor blood compatibility. This procedure is now used by many institutions in various countries and is considered safe practice. It benefits the laboratory (time savings and less handling of biohazardous material) and the patient (reduced patient specimen volume requirements). A combination of computer programs and carefully developed operating procedures can provide a safe and efficient means of detecting donor-recipient incompatibility without performance of serologic crossmatch. Certain elements are absolutely necessary for the use of computer crossmatch. These include 2 separate ABO/Rh blood group assessments on each patient, past and current history of negative antibody screens, and computer software that does not allow the release of ABO-incompatible blood and warns of Rh incompatibility.
Neonates Less Than 4 Months
There were some exceptions to the requirement. The first exception applied to neonates less than 4 months of age. Per the Mayo Clinic Transfusion Medicine institutional policy, all patients in this group receive universally compatible type O blood. If a WBIT occurred in this population the impact would be negligible from an ABO compatibility standpoint with little to no risk for a hemolytic transfusion reaction. The risk of iatrogenic blood loss outweighs any benefits of a second sample, given the transfusion policy for neonates. However, this exception led to identification of a second major area of concern—pediatric patients who are just outside the 4-month cutoff. There is no good rationale to help ease the concerns for this group of patients other than to say that the second sample requirements are for patients’ safety, a requirement that may save their life.
Specimens Collected from Patients in the Operating Room
The second exception that was identified was a patient in the operating room (OR) with no time to obtain a sample for blood typing prior to entry into the OR. Here, an exception was granted that testing would only be performed on 1 sample drawn in the OR with the recheck of blood type being done on that same sample, similar to the previous process that had been used for verification of blood type. The rationale that was used to justify this exception was that the sample obtained in the OR had been obtained from a patient who had already undergone multiple identification procedures as part of the OR process of being admitted for, and undergoing, a surgical procedure.
Last, as part of the implementation of the new requirements, stakeholders raised concerns about what to do in emergent situations where a second sample could not be obtained. To address this concern, we proposed the use of universally ABO/Rh-compatible type O Rh-negative blood, similar to the approach used by Goodnough and colleagues.7 As was expressed in their article on implementation of a 2-specimen requirement for verification of ABO/Rh for blood transfusion, there was a concern that precious O Rh-negative red cell usage would increase at an unacceptable rate. However, O-negative inventory and transfusion rates at Mayo Clinic were minimally impacted.
Impact of Practice Change
Additional effectiveness assessment data indicate that no WBIT events have been identified by implementation of the second sample requirements either through comparison with historical blood type or through comparison of results from the second sample. We performed 20,593 ABO/Rh blood type tests from November 16, 2010 to February 28, 2011. Based on the frequency of WBIT, one would expect that at least 1, if not more, WBIT events should have been detected. Few complaints have been received from medical personnel or patients related to the requirement for a second sample on patients with no historical blood type on file, indicating good acceptance of the new process as well as effective planning for the implementation and education of stakeholders and patients. One scenario to ponder: What to do in an outlying clinic or in-network physician office laboratory that draws specimens for transfusion that are referred for testing? Often these facilities have only 1 phlebotomist onsite to procure blood samples, especially during off hours. After multiple consultations with the medical staff, it was determined that an acceptable procedure is to have another medical professional, such as a nurse or physician, complete the second identification process of the patient during procurement of the second sample that is drawn by the same phlebotomist. The 2 phlebotomy and identification procedures should be 2 separate events separated by time.
In an effort to minimize WBIT specimens, enhance patient safety, and prevent release of mismatched blood products, we implemented the requirement of ABO/Rh verification for blood transfusion in patients without a historic blood type. While each facility should evaluate its process for obtaining blood specimens for transfusion, this process improvement could greatly enhance patient safety in transfusion medicine. Evaluation of institutional transfusion medicine procedures should include review of WBIT rates and labeling errors. This information can often be obtained through institutional quality or patient safety programs. Review of these data will often indicate the need for a quality improvement initiative to improve patient safety as required by various accrediting and regulatory agencies. Careful planning by a multidisciplinary team in the early stages of change can facilitate a smooth implementation of the requirement for a second sample within the organization and is well worth the investment.
Authored by: Kevin Bundy and James Stubbs, MD
- Sandler SG, Sandler DA. Transfusion Reactions. Available at http://www.emedicine.com/med/topic2297.htm 3/28/2006
- Grimm E, Friedberg RC, Wilkinson DS, et al: Blood Bank Safety Practices: Mislabeled Samples and Wrong Blood in Tube—A Q-Probes Analysis of 122 Clinical Laboratories. Arch Pathol Lab Med 2010;134(8):1108-1115
- Standards for Blood Banks and Transfusion Services. 27th edition, AABB 2011:pp 36
- The Joint Commission Accreditation Program: Hospital National Patient Safety Goals. Effective January 1, 2011, pp 2
- Mazzei CA, Kopko PM, Popovske MA. Noninfectious Complications of Blood Transfusion. In Technical Manual, 16th edition, Edited by JD Roback, MR Combs, BJ Grossman, CD Hillyer, AABB 2008, pp 725-728
- Doppler JM, Foss M, Basiri M, et al: Taking Change Control to the Next Level, Transfusion 2010;50(2S):AP58
- Goodnough LT, Viele M, Fontaine MJ, et al: Implementation of a two specimen requirement for verification of ABO/Rh for blood transfusion, Transfusion 2009; 49:1321-1328