**3. Discussion**

group incompatibility [2]. AHTR was a leading cause of transfusion-related mortality from

The blood administration process is challenging in a chaotic, often time-sensitive environment that employs high-volume blood transfusions to prevent a hemorrhaging patient from dying. This combination of factors creates a formidable risk to patient safety [1]. Optimal strategies have been developed to standardize the management of blood transfusions in the

This chapter describes select best practices and identifies system vulnerabilities that may lead to near misses and sentinel events so as to improve patient outcomes and provide an errorfree delivery of blood products. We also review potential solutions that our institution has implemented to decrease the transfusion error risk associated with blood product administration. The case review subsequently underscores how imperative it is to identify the critical

A 48-year-old female presented to the emergency room of a busy community hospital with the chief complaint of a syncopal episode. She had a history of heavy menstrual bleeding caused by multiple uterine fibroids. Upon arrival, the patient was actively bleeding per vagina. She was pale, but alert and oriented, was tachycardia at 120 beats per minute, and had orthostatic hypotension. Her past medical history was significant for a DVT while using combined oral contraceptive pills approximately 20 years before. Her past surgical history included a laparoscopic bilateral tubal ligation. She was scheduled to undergo a hysterectomy later that month secondary

While being evaluated in the emergency room, her initial blood work demonstrated a hemoglobin of 9 g/dL, down from her baseline of 12.5 g/dL. Her gynecologist was consulted who recommended performing the hysterectomy in the acute setting given her ongoing bleeding and contraindication to medical management. The patient agreed and consented for surgery. On the same day, there was a second patient being admitted for a hysterectomy for endometriosis. The patient was known to have two atypical blood antibodies necessitating cross-matched blood to be prepared. The operating staff and blood bank were in close com-

The first patient was taken to the operating room where a total laparoscopic hysterectomy was performed. Secondary to the location of her uterine fibroids, the patient sustained a laceration to her right uterine artery upon manipulation of the uterus to better visualize the uterine vessels. During attempts to control this bleeding, a massive blood transfusion (MBT) was initiated using non-cross-matched O negative blood. The gynecologic surgeons were unable to properly

Upon arrival of the blood products, the patient was immediately transfused. The surgeons completed the hysterectomy but continued to observe significant and diffuse pelvic bleeding. At this

visualize and control the source of bleeding and therefore converted to a laparotomy.

2005 to 2009, second only to transfusion-related acute lung injury (TRALI) [3].

setting of a severe hemorrhage, irrespective of its etiology.

**2. A case review**

126 Vignettes in Patient Safety - Volume 3

steps within the process of blood transfusions so as to prevent error.

to her history of heavy menses and her contraindication to estrogen therapy.

munication for this patient in an event of a hemorrhage.

Hysterectomy is one of the most commonly performed surgical procedures in the United States. Symptomatic uterine fibroids are the leading indication for the procedure, accounting for 52% of this procedure. Abnormal uterine bleeding is the indication for another 42% of hysterectomies [4].

Similarly, maternal hemorrhage is a leading cause of maternal morbidity and mortality worldwide. Its incidence varies widely but is thought to occur in 1–5% of all deliveries [5, 6]. This is a concerning fact since obstetric services are provided in 92% of rural hospitals [7]. These smaller hospitals do not have the same resources as their larger urban counterparts to handle severe hemorrhage from a variety of etiologies. For these smaller institutions, developing a standardized plan to manage emergencies such as postpartum hemorrhage is critical [7]. All surgical and emergency services should devise comprehensive approaches that identify, evaluate, treat, and monitor a hemorrhaging patient in order to stop bleeding at earlier stages, reduce the number of blood products transfused, and to reduce adverse outcomes [1]. The case review above demonstrates the need for an institution to establish and practice sound policies for the emergency preparation, transportation, and administration of blood products.

Root cause analysis (RCA) is a process with the primary aim of identifying any factors that may influence the nature, magnitude, timing, and/or occurrence of an error, keeping in mind that more than one root cause can impact an event. Through such a methodical approach, RCA is commonly employed after an occurrence of an error in order to develop and implement preventative strategies to improve future response and outcomes. Through this chapter, we perform a root cause analysis to systematically identify "root causes" of potential errors in the blood transfusion process.

Medical errors fall into one of two broad classes, errors of omission and errors of commission. An error of omission is one that occurs because an action was not taken, whereas an error of commission occurs because an incorrect action was taken [8]. The clinical vignette described in this chapter is a combination of both types of error. The errors of commission are the blood bank sending the inappropriate blood to the operating room as well as the anesthesiologist administering the wrong blood to the patient. The failure of the blood bank to properly identify which patient was receiving the transfusion as well as the failure of the anesthesiologist to verify that O negative blood was sent to the operating room would be considered acts of omission. These omissions acted in conjunction with the errors of commission to result in a nearly disastrous outcome for the patient involved. Therefore, systems must be in place to combat both of these types of errors in order to keep patients safe in complex medical situations.

non-immune, whereby other biological response modifiers in the transfused blood are believed to be the etiology behind the reaction [14]. These factors include bioactive lipids and sCD40L molecules, both of which are found in stored red cell and platelet components [15]. This may explain why TRALI reactions can also occur in the absence of donor-derived antibodies and why every blood transfusion does not result in TRALI [14]. Another model known as the "threshold model" is believed to support cases, whereby TRALI occurs in otherwise healthy recipients who receive donor blood [13]. This theory postulates that if the second event is significant enough,

Transfusion Error in the Gynecology Patient: A Case Review with Analysis

http://dx.doi.org/10.5772/intechopen.79184

129

Given the high morbidity and mortality associated with TRALI, many preventative measures have been instituted in an attempt to limit these adverse transfusion reactions. These aims include decreasing donor-derived antibodies in blood products with elevated levels of plasma. This is done by obtaining blood from male donors as opposed to females with a history of pregnancy [14]. A history of pregnancy places these patients at an increased risk of exposure to anti-HLA antibodies [14]. In addition, blood donor management strategies include the inabil-

With strategies in place to reduce these transfusion-related adverse events, there remains an additional complication to the blood transfusion process known as mistransfusion [16]. The SHOT program reported that nearly 30% of mistransfusions were a result of hospital laboratory or blood bank error [11]. These sources of error include the selection of the wrong blood sample for testing, inaccurate blood product labeling, technical errors, or incorrectly selected blood components of the wrong specification [11]. As such, an electronic pre-issuing system within the blood bank should be implemented to further reduce transfusion-associated mor-

**1.** print the blood order request form with patient identifier information, blood product number,

**2.** prepare the blood products, print a compatibility label with a bar code, and then attach the

**3.** using a handheld device, sequentially scan the bar codes and include the staff member identifier, the original blood product label, the newly attached bar code label, and the com-

**4.** ask an additional blood bank staff member to verify whether the data on the handheld scanner matches the data on the labels and forms. If they do, the blood product is issued (**Figure 1**).

The SHOT hemovigilance scheme noted that up to 70% of transfusion errors are related to ABO-incompatible transfusion at the clinical bedside, with the majority of these errors attributed to a failure to properly identify the intended transfusion recipient [11, 17]. These sentinel events are unlikely to be caused by a single error in the transfusion process. Instead, a series of errors occurring together allow the opportunity for a sentinel event to occur [18]. Over the past decade, research has suggested moving toward an automated and computerized transfu-

ity of patients who have had TRALI to donate whole blood or apheresis platelets [14].

bidity and mortality [11]. As a best practice, blood bank staff members should

blood type, and the name of the ordering physician;

sion process, with the goal of decreasing human-related error [1, 9].

label to the blood product;

patibility report form; and

then a TRALI reaction may occur without the initial clinical event [13, 14].

Secondary to human error, as many as one in 12,000 blood transfusions are administered to the wrong patient [1, 9, 10]. The serious hazards of transfusion (SHOT) Hemovigilance Program of England reports that mortality risk from transfusion in 2012 was one in 322,580 transfusion blood products while the morbidity rate was one in 21,000 [10, 11]. The transfusion of incorrect blood products, specifically ABO-incompatible blood, with resulting acute hemolytic transfusion reaction is one of the most grave and yet preventable causes of transfusion-associated morbidity and mortality [1, 12].

Transfusion-related acute lung injury (TRALI) is another rare life-threatening adverse event that can present to a recipient of a blood transfusion, as was seen in our patient in the clinical vignette [13]. The incidence is estimated to be approximately one in 5000 transfused units with more recent literature citing one in 12,000 blood products [14]. However, some literature argues that the true incidence is unknown secondary to underdiagnosis and underreporting. Regardless, TRALI rates are affected by patient population with an increased occurrence observed in critically ill patients [13]. As reported by the International Haemovigilance Network, TRALI is one of most common etiologies behind transfusion-related fatalities. Specifically, it remains the leading cause in the United States [13, 14]. The study has cited TRALI-associated mortality ranging from 5 to 8%, but up to 50–60% in critical care patients [13]. Respiratory symptoms typically present within 6 h of a transfusion of any plasma-containing blood products including intravenous immunoglobulin and cryoprecipitate [14]. TRALI is diagnosed based on clinical and radiographic findings indicating new-onset acute lung injury/acute respiratory distress syndrome within these 6 h [13]. Majority of patients require ventilator support with oxygen levels returning to pre-transfusion levels in 48–96 h [14].

Currently, there are two hypothesized theories behind the pathophysiology for TRALI. The "two hit model" is the most widely accepted hypothesis and postulates that TRALI occurs in two steps [14]. The initiating event occurs pre-transfusion and is thought to be related to the clinical condition of the patient such as recent surgery, infection, or burns [14]. This event will result in the activation of the pulmonary endothelium followed by neutrophil sequestration to this endothelium [13–15]. The second step involves the activation of neutrophils adhered to this endothelium. Activation typically occurs by donor-derived antihuman leukocyte antigen (HLA) or antihuman neutrophil antigen antibodies targeting antigens on these surfaces (HNA) [14]. The activated neutrophils then incite endothelial damage which results in capillary leak and pulmonary edema [14, 15]. This second step is postulated to be either immune-mediated or non-immune-mediated. The majority of TRALI is immune-mediated, whereby neutrophil and HLA class I and II antibodies initiate TRALI [15]. However, approximately 15–20% of cases are non-immune, whereby other biological response modifiers in the transfused blood are believed to be the etiology behind the reaction [14]. These factors include bioactive lipids and sCD40L molecules, both of which are found in stored red cell and platelet components [15]. This may explain why TRALI reactions can also occur in the absence of donor-derived antibodies and why every blood transfusion does not result in TRALI [14]. Another model known as the "threshold model" is believed to support cases, whereby TRALI occurs in otherwise healthy recipients who receive donor blood [13]. This theory postulates that if the second event is significant enough, then a TRALI reaction may occur without the initial clinical event [13, 14].

in this chapter is a combination of both types of error. The errors of commission are the blood bank sending the inappropriate blood to the operating room as well as the anesthesiologist administering the wrong blood to the patient. The failure of the blood bank to properly identify which patient was receiving the transfusion as well as the failure of the anesthesiologist to verify that O negative blood was sent to the operating room would be considered acts of omission. These omissions acted in conjunction with the errors of commission to result in a nearly disastrous outcome for the patient involved. Therefore, systems must be in place to combat both of these types of errors in order to keep patients safe in complex medical situations.

Secondary to human error, as many as one in 12,000 blood transfusions are administered to the wrong patient [1, 9, 10]. The serious hazards of transfusion (SHOT) Hemovigilance Program of England reports that mortality risk from transfusion in 2012 was one in 322,580 transfusion blood products while the morbidity rate was one in 21,000 [10, 11]. The transfusion of incorrect blood products, specifically ABO-incompatible blood, with resulting acute hemolytic transfusion reaction is one of the most grave and yet preventable causes of transfusion-associated

Transfusion-related acute lung injury (TRALI) is another rare life-threatening adverse event that can present to a recipient of a blood transfusion, as was seen in our patient in the clinical vignette [13]. The incidence is estimated to be approximately one in 5000 transfused units with more recent literature citing one in 12,000 blood products [14]. However, some literature argues that the true incidence is unknown secondary to underdiagnosis and underreporting. Regardless, TRALI rates are affected by patient population with an increased occurrence observed in critically ill patients [13]. As reported by the International Haemovigilance Network, TRALI is one of most common etiologies behind transfusion-related fatalities. Specifically, it remains the leading cause in the United States [13, 14]. The study has cited TRALI-associated mortality ranging from 5 to 8%, but up to 50–60% in critical care patients [13]. Respiratory symptoms typically present within 6 h of a transfusion of any plasma-containing blood products including intravenous immunoglobulin and cryoprecipitate [14]. TRALI is diagnosed based on clinical and radiographic findings indicating new-onset acute lung injury/acute respiratory distress syndrome within these 6 h [13]. Majority of patients require ventilator support with oxygen

Currently, there are two hypothesized theories behind the pathophysiology for TRALI. The "two hit model" is the most widely accepted hypothesis and postulates that TRALI occurs in two steps [14]. The initiating event occurs pre-transfusion and is thought to be related to the clinical condition of the patient such as recent surgery, infection, or burns [14]. This event will result in the activation of the pulmonary endothelium followed by neutrophil sequestration to this endothelium [13–15]. The second step involves the activation of neutrophils adhered to this endothelium. Activation typically occurs by donor-derived antihuman leukocyte antigen (HLA) or antihuman neutrophil antigen antibodies targeting antigens on these surfaces (HNA) [14]. The activated neutrophils then incite endothelial damage which results in capillary leak and pulmonary edema [14, 15]. This second step is postulated to be either immune-mediated or non-immune-mediated. The majority of TRALI is immune-mediated, whereby neutrophil and HLA class I and II antibodies initiate TRALI [15]. However, approximately 15–20% of cases are

morbidity and mortality [1, 12].

128 Vignettes in Patient Safety - Volume 3

levels returning to pre-transfusion levels in 48–96 h [14].

Given the high morbidity and mortality associated with TRALI, many preventative measures have been instituted in an attempt to limit these adverse transfusion reactions. These aims include decreasing donor-derived antibodies in blood products with elevated levels of plasma. This is done by obtaining blood from male donors as opposed to females with a history of pregnancy [14]. A history of pregnancy places these patients at an increased risk of exposure to anti-HLA antibodies [14]. In addition, blood donor management strategies include the inability of patients who have had TRALI to donate whole blood or apheresis platelets [14].

With strategies in place to reduce these transfusion-related adverse events, there remains an additional complication to the blood transfusion process known as mistransfusion [16]. The SHOT program reported that nearly 30% of mistransfusions were a result of hospital laboratory or blood bank error [11]. These sources of error include the selection of the wrong blood sample for testing, inaccurate blood product labeling, technical errors, or incorrectly selected blood components of the wrong specification [11]. As such, an electronic pre-issuing system within the blood bank should be implemented to further reduce transfusion-associated morbidity and mortality [11]. As a best practice, blood bank staff members should


The SHOT hemovigilance scheme noted that up to 70% of transfusion errors are related to ABO-incompatible transfusion at the clinical bedside, with the majority of these errors attributed to a failure to properly identify the intended transfusion recipient [11, 17]. These sentinel events are unlikely to be caused by a single error in the transfusion process. Instead, a series of errors occurring together allow the opportunity for a sentinel event to occur [18]. Over the past decade, research has suggested moving toward an automated and computerized transfusion process, with the goal of decreasing human-related error [1, 9].

blood product containers [1, 16]. The microchips are scanned by a handheld portable device and uploaded to a program with the operator alerted to a misstep and the program pausing until the error is corrected [16]. RFID can be used to further standardize and monitor blood collection, preparation, and transfusion in order to reduce transfusion-related human error

Transfusion Error in the Gynecology Patient: A Case Review with Analysis

http://dx.doi.org/10.5772/intechopen.79184

131

The operational aspects of EBT present many challenges that can be overcome by planning and employing best practices. These challenges include describing how to recognize, initiate, and alert others to an EBT. A hospital system must also decide on what kind of blood products to store and how to prioritize select products when they are in high demand. In addition, determining where to store blood products and deciding how to transport them to the bedside require careful planning, especially when faced with multiple concurrent patients requiring EBTs. Personnel should also be trained on how to resuscitate patients while waiting

The first step in any EBT starts with the attending provider recognizing the need to transfuse. Common indications for EBT include an elevated Assessment of Blood Consumption (ABC) score, the presence of visible rapid blood loss such as that seen in postpartum hemorrhage, or the observation of even moderate blood loss in the setting of comorbidities such as low cardiopulmonary reserve. This list is not exhaustive. The ABC score is calculated by assigning a score of one to each parameter present: penetrating injury, positive focused assessment sonography for trauma (FAST), systolic blood pressure of 90 mm Hg or less, and an elevated heart rate of at least 120 beats per minute (**Table 1**). An ABC score of 2 or higher is 75% sensi-

After recognizing the need for transfusion, the second step in an EBT is to alert the blood bank. Our institution offers two levels of response to an EBT: emergency blood release (EBR) and an elevated response level of "code crimson" (CC). Either is initiated by a medical provider dialing a simple hotline—"5555" at our institution—that is answered immediately by the emergency operator. Notification of an EBR arrives in the blood bank via an alphanumeric

ED, emergency department; BP, blood pressure; HR, heart rate; FAST, focused assessment with sonography in trauma.

tive and 86% specific for predicting the need for massive transfusion (MT) [19].

Penetrating Mechanism No 0 Yes + 1 ED Systolic BP ≤ 90 mmHg No 0 Yes + 1 ED HR ≥ 120 No 0 Yes + 1 Positive Ultrasound FAST Exam No 0 Yes + 1

and improve patient safety [16].

for the arrival of blood products.

**Table 1.** Assessment of blood consumption score.

**Criteria**

**4. The operational aspects of EBT**

**Figure 1.** Serious hazards of transfusion (SHOT) best practice flowchart.

Implementation of a software-driven bar code tracking system in place of the conventional "nurse to nurse" double check system for the administration of blood products has been identified as a key strategy for improving transfusion safety [9]. The bar code on blood components identifies blood group, blood type, unit of blood, product number, and the date of collection [9, 10]. Several companies offer a bar code electronic identification system (EIS) which may be portable or built into the electronic medical system [11]. A portable handheld scan and print electronic device can be used to verify and document patient identity. Such a device is utilized at our institution [11]. The common components of the pre-transfusion check list to be scanned include the patient name, medical record number, and blood group [11]. If the bar codes between the patient wristband and blood products match, then the handheld device indicates as such [11]. The bar code EIS is linked to a network host computer that can store, search, and send transfusion data [11]. Multiple studies have indicated that the electronic bar code system is effective in reducing human error related to transfusion procedures as it acts as another barrier for error in the transfusion process [10]. In a time-sensitive event such as a massive transfusion protocol, safety checks including barcoding EIS may be omitted. This may reintroduce transfusion-related human error, such as incorrect blood product administration to the recipient.

Current research is exploring the use of smartphone or tablet devices in transfusion medicine with the aim of achieving enhanced integrity of the transfusion process [10]. In addition, systems utilizing radiofrequency identification (RFID) are being analyzed as a new way of integrating technology into blood transfusion best practice. However, high costs for an institution can be a barrier [10, 11]. RFID is a more user-friendly technology and can be applied to improve visual and bar code electronic identification systems [11, 16]. Radiofrequency transponder microchips have been utilized on patient wristbands, blood sample tube labels, and blood product containers [1, 16]. The microchips are scanned by a handheld portable device and uploaded to a program with the operator alerted to a misstep and the program pausing until the error is corrected [16]. RFID can be used to further standardize and monitor blood collection, preparation, and transfusion in order to reduce transfusion-related human error and improve patient safety [16].
