Chapter 11 – Transfusion Therapy and Common Hematologic Problems in the Critically Ill Elderly

Chapter 11 Transfusion Therapy and Common Hematologic Problems in the Critically Ill Elderly

Aryeh Shander , Faraz Syed and Mazyar Javidroozi

Key Points

Physiologic changes with aging are generally expected to increase the risk of anemia while limiting the physiologic adaptations to anemia.
Anemia is common in elderly patients and is often multifactorial.
Anemia is an independent risk factor for worse outcomes and should never be left untreated.
While transfusion is commonly considered to be the default treatment of anemic elderly patients, it is important to consider management strategies other than transfusion.
Allogeneic blood transfusions have been linked to several unfavorable outcomes and must be used only when clearly indicated.
Several control trials have demonstrated that restrictive transfusion strategies are associated with reduced use of blood while achieving similar or better outcomes than liberal transfusion strategies.
Some evidence suggests that certain patient populations (e.g., those with cardiovascular morbidities) might face some risk when managed with restrictive transfusion strategies, but more studies are needed.
Despite the general assumption that elderly patients might benefit from more liberal transfusion strategies, this notion is not supported by the available evidence.
Other strategies including preventive measures – discussed under the concept of patient blood management – should be considered in all critically ill elderly patients who are anemic or at risk of becoming anemic.

Introduction

Advanced age is associated with significant changes within the hematologic and cardiovascular systems (along with other organs), including how these systems interact with one another. When it comes to the clinical manifestations of these changes and their impact on patient outcomes, the lines often become blurred, and the clinical impact of many of the cellular and subcellular changes that are associated with senescence remains debatable.

Aging is associated with marked changes in bone marrow cellularity. These changes are hallmarked by a gradual reduction in the hematopoietic cell populations and associated cytokine levels, while the fat cell counts increase. From a cellular perspective, all populations in the hematopoietic lineage from stem cells to more differentiated cells are affected [1]. As a result, the frequency of hematologic disorders – namely anemia – increases with aging, while the capacity of bone marrow to respond to hematological stress – namely increased production of red blood cells (RBCs) in response to bleeding – declines [2]. Age-related changes affecting coagulation systems are observed in platelets as well as coagulation factors, with a general shift toward hypercoagulability in the aged patients, which contributes to an increased incidence of thrombotic events in this population [3]. It is important to remember that these and many other changes of aging are gradual in nature and that there is no magic age beyond which all the changes appear abruptly. Similarly, there are many elderly patients who may not experience decline in hemoglobin levels or in other functions of the hematologic system [4].

As with many other cancers, the prevalence of hematopoietic malignancies increases with advanced age. This phenomenon is attributed to several factors, including an accumulation of mutations in stem cells over time [5]. It has been shown that hematopoiesis becomes increasingly “clonal” as we age, meaning that fewer populations of hematopoietic stem cells become dominant and give rise to the bulk of blood cells produced in the bone marrow. Clonal hematopoiesis is more common in elderly patients, and it may lead to leukemia in some [1].

Aging affects the immune system and is associated with impairments in function affecting both innate and adaptive immunity [6] (see Chapter 10). The clinical consequences of these changes (collectively referred to as immunosenescence) [7] may emerge as increased infections and infective complications. Sepsis is therefore more common in the elderly, and it is increasingly associated with worse outcomes, including a higher risk of mortality [8]. Aging is generally recognized as a proinflammatory state, and the term inflamm-aging is used to describe the low-grade inflammation that is often present in the aged body [9]. The consequences of this ongoing inflammation are a matter of debate, and some have linked it with the increased frequency of other chronic diseases that are seen more frequently in the elderly, from diabetes and cardiovascular problems to neurodegenerative diseases and malignancies [7].

Aging is also associated with a gradual decline in the physiologic functional capacity of the cardiovascular and respiratory systems (and their reserve capacity). Perfusion is tightly controlled throughout the body to ensure a reliable supply of oxygen to the cells to meet demands. Several aspects of the regulatory mechanisms that are in place to maintain this balance are compromised during aging – as a direct result of the aging process itself or the result of the presence of concurrent ailments that are more common in the elderly (e.g., hypertension, atherosclerosis) [10]. Furthermore, the declining reserve capacity of the cardiovascular and respiratory systems in aging is expected to limit the adaptability of body to changing conditions that limit supply or increase demand, increasing the risk of becoming decompensated in situations that are normally tolerated in younger patients [11].

Widespread changes in other organs also may affect the hematologic system. For example, age-related changes in the gastrointestinal system include disruption of the mucosal defense barrier, which can lead to an increased risk of inflammation and infection. This mucosal disruption and the ensuing inflammatory processes may interfere with absorption of nutrients, causing malnutrition and potentially leading to nutritional anemia [12].

Discussion of the pathophysiologic characteristics and changes associated with aging often leads us to the concept of frailty. While there is no standard definition for frailty, it generally can be described clinically as a state of increased vulnerability to acute stressors that is a result of a progressive decline in the functional status and reserve capacities of various physiologic systems in the body [13]. From a practical point of view, the presence of at least three of the five criteria suggested by Fried et al. is highly suggestive of frailty in an elderly patient: low energy, low grip strength, low physical activity, slowed walking speed, and unintended weight loss [14].

Key Concepts

Anemia in the Elderly

The interaction between aging and the hematologic system is multidimensional and can have far-reaching implications. Advanced age (and frailty) can increase the risk of developing anemia or worsening of this condition, while reduced reserve capacity of other systems (and the expected impaired physiologic adaptations to anemia) [15] might lead to increased susceptibility to hypoxia and may reduce the tolerance to anemia. Additionally, reduced hematopoietic capacity in the elderly may delay the recovery from acute blood loss, for example, following trauma or surgery [2].

Anemia in the elderly is often multifactorial, and etiologies include inflammation, renal insufficiency, blood loss, nutritional deficiencies (iron, folate, vitamin B₁₂), declined hematopoietic capacity, increased cell death (erythrocyte apoptosis or eryptosis), and iatrogenic causes such as drug-induced anemia [16–19] In a substantial number of elderly patients, the cause(s) of anemia might not be readily identified, giving rise to the term unexplained anaemia of the elderly (UAE) or anemia of unknown etiology (AUE) in this population. In a study from the 3rd National Health and Nutrition Examination Survey (NHNES), the primary cause of anemia in older patients was attributed to iron deficiency in one-third of patients and chronic renal disease or chronic inflammation in another one-third of patients, whereas the remaining one-third of anemic patients were classified as having AUE [20]. With increasing knowledge of the causes of anemia, it might be a matter of time until the underlying causes are identified in some of the AUE patients [18]. However, in some other patients, AUE appears to be a distinct entity characterized by a normocytic anemia developing in the absence of nutrition deficiency, inflammation, or renal dysfunction [21]. In a large study comparing elderly patients with AUE with matched cohorts of nonanemic and anemic individuals with determined causes, interleukin 6 (IL-6) and hepcidin were not different, suggesting that inflammation or iron restriction was not a primary cause [22]. Reduced testosterone levels and a blunted erythropoietin response (characterized by raised levels of erythropoietin but not raised as much as expected for the level of anemia present) were seen more frequently in patients with AUE [22].

Anemia is more frequent in elderly populations [19]. Table 11.1 provides a summary of the reported prevalence of anemia among various aged populations [20,23–32]. As can be seen, there is often a trend toward increase prevalence of anemia in older populations.

Table 11.1 Prevalence of Anemia in the Elderly and Its Reported Consequences

Study	Population	Definition of anemia	Prevalence of anemia	Outcomes of anemia
Contreras et al. (2015) [31]	328 people older than 85 years of age living in a community	WHO criteria	24 percent	Anemia was associated with more dependence and higher comorbidity and mortality.
Deal et al. (2009) [25]	A representative sample of 436 community-dwelling women aged 70 to 80 years in the United States	Hb < 12 g/dl	8.8 percent	Anemia was associated with poorer baseline performance and faster rates of decline on cognitive tests.
Guralnik et al. (2004) [20]	US population aged 65 or older from the 3rd National Health and Nutrition Examination Survey (1988–94)	WHO criteria	11.0 percent of men and 10.2 percent of women ≥ 65 years of age; >20 percent in those ≥85 years of age	–
Hong et al. (2013) [29]	2,552 elderly subjects (mean age 76.1 years) participating in the Health, Aging, and Body Composition study	WHO criteria	15.4 percent	Anemia at baseline was independently associated with increased risk of dementia.
Jorgensen et al. (2010) [23]	5,286 residents of Tromsø, Norway, 55–77 years of aged	WHO criteria	3.4 percent	Lower Hb was associated with higher risk of fracture. Anemic men (but not women) had a 2.15 higher risk of nonvertebral fractures than men with high Hb levels.
Juarez-Cedillo et al. (2014) [30]	1,933 older community-dwelling adults enrolled in the Study on Aging and Dementia in Mexico	WHO criteria	8.3 percent	Anemia and low Hb were independently associated with increased risk of frailty.
Nakashima et al. (2012) [27]	Random sample of 124 adults younger than 60 years of age in long-term care facilities in Maringa, Brazil	WHO criteria	29 percent	–
Rosnick et al. (2010) [28]	451 residents from 12 nursing homes with an average age of 83.7 ± 8.2 years	WHO criteria	54 percent	Anemia was associated with worse physical performance; patients with anemia associated with chronic kidney disease had lower self-efficacy and outcome expectations for functional activities than those without anemia.
Samper-Ternen et al. (2011) [26]	5,605 adults older than 60 years of age from the Mexican National Health and Nutrition Survey	WHO criteria	10.3 percent	–
Zakai et al. (2013) [32]	3,758 nonanemic community-dwelling subjects 65 years of age or older	WHO criteria	9 percent developed anemia during the 3-year follow-up	Hb decline and anemia were associated with worsening of cognitive function and mortality.

Note: World Health Organization (WHO) criteria for the definition of anemia is based on a hemoglobin concentration of less than 13 g/dl in adult men and less than 12 g/dl in adult nonpregnant women [24] (Hb = hemoglobin).

Anemia in the elderly is an independent risk factor for unfavorable outcomes. It has been linked to a decline in quality of life [33], cognitive function [29,32,34], activities of daily living [35], mobility [36,37], and strength [37,38]. Anemia also leads to an increased risk of physical impairments [30,31], falls [39,40], depression [41], hospitalization and nursing home placement [42,43], frailty [30], and mortality [31,32,35,42–44]. Table 11.1 summarized the negative outcomes of anemia reported in a number of studies.

Some evidence suggests that the impact of anemia and hemoglobin levels on outcome in the elderly patients is more complicated [21]. In a study of nonhospitalized elderly disabled women, there was a sharp increase in the risk of 5-year mortality with declining levels of hemoglobin. Interestingly, high hemoglobin levels were also found to be associated with increased risk of long-term mortality (although the magnitude of the risk was far smaller compared than that associated with lower levels of hemoglobin) [45]. Another finding in this study was that lowest mortality rates were associated with a hemoglobin level of 14 to 14.5 g/dl, and there was a linear increase in the risk of death as hemoglobin levels decreased from that range. The suggestion that hemoglobin decline rather than presence or absence of anemia might be a more accurate predictor of unfavorable outcomes in the elderly is supported by other studies [32]. The effect of critical illness as a significant risk factor for the development of anemia [46] and anemia as a significant contributor to the worsening of outcome further complicate the issue [17,47]. A similar association between anemia and worse outcome can be seen in critically ill elderly patients [48].

Blood Transfusion Strategies and Outcomes

Blood transfusion has been traditionally considered to be the standard treatment for anemia across populations including the elderly. Transfusion provides a seemingly simple and readily available treatment to quickly raise the hemoglobin level and restore blood oxygen-carrying capacity and hemodynamic status of the patient to avoid tissue hypoxia and ischemia. In reality, a unit of allogeneic red blood cells (RBCs) is a complex and heterogeneous mixture of various cells and bioactive factors, antigens, vesicles, metabolites, and mediators that could be constantly changing [49]. Transfusion of allogeneic blood should be viewed as a live-tissue transplantation in essence and treated with the same level of caution and scrutiny [50].

Historically, infectious risks and complications have been the leading concerns related to allogeneic blood transfusions. Following implementation of various screening, testing, and processing (e.g., pathogen inactivation) strategies, the risk of transmission of infections through donated blood has been reduced to extremely low levels – typically less than 0.1 per million units of blood components for hepatitis C and B and human immunodeficiency virus – in the developed nations [51]. Nonetheless, new infectious agents remain a potential risk given the unavoidable delay between the first emergence of an infection and the time measures to screen for it or methods to inactivate the pathogen are implemented [52]. Despite the wide publicity and significant cost to the healthcare systems [53], the health burden of transfusion-transmitted infections is dwarfed compared with the noninfectious risk of blood transfusion.

Not surprisingly, many of the complications of blood transfusion are rooted in the immunogenicity of allogeneic blood and its interactions with the immune system of the recipient. Examples include alloimmunization [54], immunomodulation [55], febrile reactions [56], transfusion-related acute lung injury (TRALI) [57], and graft-versus-host disease (GVHD) [58].

Ex vivo storage and processing of donated blood units and the changes that occur during storage (storage lesion) are other mechanisms that have been suggested for the negative effects of blood transfusion [59,60]. Nonetheless, clinical trials to date have not been able to demonstrate a clear superiority for the transfusion of fresh blood units versus aged blood units (but still within the accepted shelf life of the component) in terms of improved clinical outcomes [61,62].

The number of studies linking allogeneic blood transfusions with unfavorable clinical outcomes including multiorgan dysfunction, cardiac complications, thromboembolic events, respiratory distress and failure, prolonged ventilator dependency, stroke, renal injury, sepsis and infection, mortality, and prolonged hospital stay in various patient population has been growing continuously [63,64]. This has given rise to efforts to restrict the use of blood transfusion and limit it only to patients in whom it is clearly indicated.

In their landmark study (Transfusion Requirements in Critical Care [TRICC] trial), Hebert et al. randomized anemic critically ill patients to a restrictive transfusion strategy (transfusing when hemoglobin fell below 7 g/dl to maintain hemoglobin between 7 and 9 g/dl) versus a liberal transfusion strategy (transfusing when hemoglobin fell below 10 g/dl to maintain it between 10 and 12 g/dl) [65]. They reported that the restrictive transfusion strategy was associated with reduced mortality compared with a liberal transfusion strategy among patients who were less acutely ill and those who were younger than 55 years of age, while there was a nonsignificant trend toward increased mortality risk among patients with acute cardiac ischemia and infarction [65]. Given that the survival benefits of restricting transfusions was clearly seen in younger patients, one must ask whether there is a potential for harm from restrictive transfusion strategies to elderly critically ill patients.

The results of the TRICC trial have been largely corroborated by trials conducted since then. A meta-analysis of data from over 3,700 participants in 17 clinical trials concluded that a restrictive transfusion strategy can achieve a 37 percent reduction in the risk of being transfused and a 0.75 unit smaller amount of blood being transfused on average per patient. Analysis of the pooled data indicated that using a restrictive transfusion strategy was associated with a significant reduction in the risk of infection [66]. In another meta-analysis of published data from over 6,200 participants in 19 clinical trials, restrictive transfusion strategy was found to be associated with a 30 percent reduction in transfusion rate and 1.19 absolute reduction in the number of units transfused, in addition to reducing the risk of in-hospital mortality rates (relative risk [RR] 0.77, 95 percent confidence interval [CI] 0.62–0.95) [67].

Holst et al. evaluated data from 31 trials in patients undergoing cardiovascular surgery and concluded that adhering to a restrictive transfusion strategy (hemoglobin threshold of 7 to 9 g/dl or hematocrit level of 24–25 percent depending on the study) compared with a less restrictive transfusion strategy (hemoglobin threshold of 8 to 10 g/dl or hematocrit of 30–32 percent) was associated with a significant reduction in transfusion use, while there was no statistically significant impact on the risk of myocardial infarction, stroke, renal failure, or mortality [68].

Among 1,000 patients with septic shock and a hemoglobin level of 9 g/dl or less in the intensive care unit (ICU) who were randomized to receive 1 unit of RBCs when hemoglobin was below 7 g/dl (lower threshold) or 9 g/dl (higher threshold), 90-day mortality, ischemic events, and the need for life support were similar [69]. In post hoc analyses of subgroups of patients with chronic lung disease, hematologic malignancy, metastatic cancers, or postoperative patients, no survival benefit was observed for transfusion at higher versus lower hemoglobin thresholds [70]. Long-term follow-up of these patients did not reveal any significant differences in mortality and health-related quality of life between the study arms [71].

Another group looked at the impact of transfusion strategies on risk of mortality in adult patients undergoing surgeries and critically-ill adult patients and concluded that a liberal transfusion strategy may improve survival in patients undergoing surgery but not in the critically ill [72]. It should be noted that the reported survival benefit of transfusion in this meta-analysis was heavily influenced by two studies that have faced methodologic criticism [73,74].

Ripolles et al. analyzed the data from six trials with a total of 2,156 patients, focusing on critically ill patients and those admitted with acute coronary syndrome, and they observed a trend toward less mortality among the critically ill patients randomized to restrictive transfusion strategies (RR 0.86, 95% CI 0.73–1.01) [75]. Recently, Carson et al. reviewed the data from over 12,500 participants in 31 clinical trials comparing restrictive transfusion triggers (usually hemoglobin 7–8 g/dl) with liberal transfusion triggers (hemoglobin 9–10 g/dl) [76]. Use of restrictive transfusion triggers was associated with 43 percent reduction in transfusion rates, whereas there was no negative impact on 30-day mortality or morbidity [76]. The authors concluded that while more studies are needed in particular patient populations such as those with acute coronary syndrome, acute neurologic disorders, stroke, cancers (including hematologic malignancies), and bone marrow failure, the available evidence supports the notion that allogeneic blood transfusions are generally not needed in patients with hemoglobin levels greater than 7–8 g/dl [76]. The authors did not specifically address elderly patients in their analysis. However, it is expected that with the higher prevalence of the comorbidities in the elderly, more research is needed in this specific population.

The impact of the presence of cardiovascular comorbidities on the outcomes of transfusion strategies was specifically addressed by a meta-analysis of 11 trials on over 3,000 patients undergoing noncardiac surgery [77]. Overall pooled risk of 30-day mortality was not statistically significantly different between the restrictive and liberal transfusion strategies, whereas restrictive transfusion strategy (based on a hemoglobin threshold of 8 g/dl or less) was associated with an increased risk of acute coronary syndrome (pooled risk ratio 1.78; 95% CI 1.18–2.70) [77]. In their subgroup analysis of data from trials on patients with acute myocardial infarction, Ripolles et al. reported a nonsignificant trend toward increased risk of mortality in patients assigned to restrictive transfusion strategies (RR 3.85, 95% CI 0.82–18.0) [75].

In a study on 2,003 patients with a hemoglobin concentration of less than 9 g/dl following cardiac surgery who were randomized to a liberal transfusion strategy (based on a postoperative hemoglobin threshold of less than 9 g/dl) or restrictive transfusion strategy (based on a hemoglobin threshold of less than 7.5 g/dl), the restrictive transfusion strategy was associated with significant reductions in resource use and costs without negatively affecting the primary composite outcome of any serious infection or ischemic events within the 3-month period following surgery [78]. However, all-cause mortality rates were higher in the restrictive transfusion arm (4.2 percent versus 2.6 percent in liberal transfusion group; hazards ratio [HR] 1.64, 95% CI 1.00–2.67). This observation led the investigators to refrain from recommending restrictive transfusion strategies without any reservations for patients following cardiac surgery, and they called for additional studies and meta-analyses of the data from existing studies in this population [78].

Hovaguimian et al. have recently published a meta-analysis of data from trials comparing liberal and restrictive transfusion strategies in which they pooled and analyzed data from trials within four predefined context-specific subgroups: patients undergoing cardiovascular surgery (3,323 patients from eight trials), elderly patients undergoing orthopedic surgery (3,777 patients from nine trials), medical or surgical acute care patients (emergency or critical care unit; 4,129 patents from 10 trials), and patients with acute brain trauma on intracranial bleeding (244 patients from two trials) [79]. The authors concluded that based on their analysis, there was a possibility that restrictive transfusion strategies in patients undergoing cardiovascular surgery were associated with an increased risk of complications related to inadequate oxygen supply (RR 1.09, 95% CI 0.97–1.22) and early mortality (RR 1.39, 95% CI 0.95–2.04). However, no such association was reported in critically ill patients or those with acute cerebral bleeding [79].

While controlled, randomized trials (and the meta-analyses that pool the results from several trials) are generally considered to provide the highest grade of evidence on the safety and efficacy of medical treatments, including transfusions, it is important to recognize the limitation of these studies. There is often the possibility of issues related to outcome-specific risk adjustment in trials (and meta-analyses that draw on them). Reported associations between restrictive transfusion strategies and increased risk of ischemic cardiac events often lack adequate assessment and analysis for the potential role of baseline risk factors for their outcomes. Additionally, clinical trials of transfusion strategies often focus on transfusions performed in a prespecified time period during the course of care and might overlook any transfusions given outside that window. As such, a patient who was assigned to the restrictive transfusion arm of a study focusing on the postoperative period might still have receive liberal transfusions before or during the surgery, confounding the observed association between the outcomes and transfusion strategies. The same issues can persist and even propagate when data from large trials suffering from these issues is pooled in meta-analyses [80].

Based on the available evidence, a transfusion hemoglobin threshold of 7 to 8 g/dl appears to be safe and reasonable for most critically ill patients and in the absence of other indications of transfusion (e.g., symptoms of hypoxia). However, extra caution might be warranted in the presence of certain comorbidities such as ischemic cardiac disease and brain injury, and more studies are needed in these patient populations [81]. This approach is rather based on an abundance of caution and not necessarily supported by available evidence. Hence alternative treatment options (primarily and prevention and treatment of anemia) should be considered whenever possible (see below) [82].

Blood Transfusion in the Elderly

A recurring theme in retrospective transfusion studies is the observation that transfused patients are generally significantly older that their nontransfused peers. When it comes to transfusion decisions, factors most commonly considered by clinicians include the hemoglobin level, comorbidities, and age [83]. Nonetheless, the underlying evidence to support the idea that advanced age, in absence of other comorbidities, is an independent factor that makes the patient more in need of transfusion (or more liberal transfusion strategies) is generally lacking. While elderly patients are among the largest patient populations across various hospital wards and outpatient clinics, the number of studies specifically addressing transfusion in this population is very limited.

Wu et al. conducted a retrospective review of data from approximately 79,000 patients hospitalized with acute myocardial infarction aged 65 years or older [84]. As expected, they observed that lower hemoglobin levels at the time of admission were associated with increased risk of mortality within 30 days. The impact of transfusion on mortality depended on the hemoglobin level: transfusion was associated with a reduced risk of mortality within 30 days in patients who had lower hemoglobin levels at admission (odds ratio [OR] 0.22, 95% CI 0.11–0.45 in patients with hematocrit levels of 5 to 24 percent and OR 0.69, 95% CI 0.53–0.89 in those with admission hematocrit levels of 30.1 to 33 percent) [84]. However, the conclusion that elderly patients with acute coronary artery disease may benefit from transfusion at hematocrit levels as high as 30 percent (or 33 percent on admission) was inconclusive and challenged by others [85].

In a smaller cohort study on 919 elderly patients undergoing surgery for hip fracture, one-third of the patients were transfused during hospital stay [86]. Although transfusion did not affect survival up to 6 months following admission, the risk of infection (of chest, urinary tract, or wound) was significantly higher in transfused patients (hazard ratio [HR] 1.91, 95% CI 1.41–2.59), and transfused patients had significantly longer duration of stay in hospital compared with patients who were not transfused [86]. The association between restrictive transfusion strategies and reduced risk of infection in hospitalized patients has been confirmed in a meta-analysis of 21 randomized, controlled trials (risk ratio [RR] 0.82, 95% CI 0.72–0.95) [87].

Gregersen et al. studied the association between transfusion strategies and various outcomes in 157 patients 65 years of age or older who underwent surgery for hip fracture [88]. They randomized patients to transfusion strategies based on a hemoglobin threshold of less than 9.7 g/dl (restrictive) or less than 11.3 g/dl (liberal) during the 30-day period following surgery. No significant differences were detected between the study groups with regard to overall quality of life (QoL) a month or a year later. However, the authors reported improved recovery of activities of daily living (ADLs) in patients randomized to their liberal transfusion strategy [88]. In a related study including 284 elderly patients following hip fracture surgery, despite the similarity of measures of ADLs, patients randomized to restrictive transfusion strategy had increased mortality at 30 days (HR 2.4, 95 percent CI 1.1–5.2) and 90 days (only among nursing home residents; HR 2.0, 95% CI 1.1–5.2) [89]. Subsequent analysis of the same group of patients indicated that there was no significant difference in the infection rates among patients randomized to liberal versus restrictive transfusion strategies (66 versus 72 percent, respectively; RR 1.08, 95% CI 0.93–1.27) [90]. While these related studies suggest that liberal transfusion strategies might offer some advantages over restrictive transfusion strategies in elderly patients, it must be emphasized that the restrictive transfusion strategy used in these studies was based on a hemoglobin threshold of 9.7 g/dl, which is in the ballpark of liberal hemoglobin thresholds in most other studies. In other words, what is being compared in these studies might be more appropriately labeled “liberal” and “more liberal” transfusion strategies, which resulted in almost 9 of every 10 patients being transfused regardless of study arm allocation [88–90].

In a clinical trial to study the impact of liberal blood transfusion on recovery in elderly patients undergoing surgery for hip fracture, Carson et al. randomized over 2,000 patients 50 years of age or older (mean age 81 years of age) to either a restrictive transfusion strategy (hemoglobin threshold < 8 g/dl or when symptoms of anemia were present) or a liberal transfusion strategy (hemoglobin threshold 10 g/dl) [91]. Almost all patients in the liberal arm were transfused (median of 2 units of RBCs), whereas 41 percent of patients in the restrictive arm received transfusions. The primary outcome of death or inability to walk unassisted across a room on 60-day follow-up occurred in 35.2 and 34.7 percent of the patients in the liberal and restrictive strategy arms, respectively (OR 1.01; 95% CI 0.84–1.22). There was no significant difference in the occurrence of other complications between the study arms [91].

In the meta-analysis by Hovaguimian et al. [79], elderly orthopedic surgery was one of the four context-specific subgroups. Of note, over half the patients whose data were pooled were from the trial by Carson et al. discussed earlier [91]. Based on their pooled data analysis, they concluded that restrictive transfusion strategies were associated with an increased risk of events reflective of inadequate oxygen supply (RR 1.41, 95% CI 1.03–1.92). However, there was no increase in the risk of early mortality (RR 1.09, 95% CI 0.80–1.49). In contrast, elderly patients assigned to restrictive transfusion strategies were less likely to suffer infectious complications (RR 0.75, 95% CI 0.53–1.04) [79].

With limited evidence available on the impact of transfusion strategies on outcomes of elderly patients (and even more limited evidence from critically ill elderly patients), it is difficult to make transfusion recommendations specific to this population. While advanced age remains a traditional factor to consider when making transfusion decisions [83], the actual clinical implication of age with regard to transfusion requirements is not clearly defined yet, and it remains to be established whether older age can independently justify more liberal transfusion strategies. Until then, a hemoglobin threshold of 7 to 8 g/dl for transfusion of blood can be considered for these patients in the absence of other indications of transfusion [21,81]. The presence of comorbidities – which tend to more be prevalent at older age – is a distinct consideration that might justify transfusion at higher hemoglobin thresholds, while more studies are needed to better define the optimal transfusion strategies in these patients. In the meanwhile, currently available transfusion guidelines, including those specifically addressing transfusion in critically ill patients, can be applied in the critically ill elderly patients [92–101] (Table 11.2).

Table 11.2 Summary of Blood Transfusion Guidelines

Guidelines	American Society of Anesthesiology (2006) [96]	Society of Thoracic Surgeons (2007) [92]	Italian Society of Transfusion Medicine and Immunohematology (2011) [93–95]	American Association of Blood Banks (2016) [97]	National Clinical Guideline Center (UK) (2015) [98]	American College of Physicians (2013) [99]	British Committee for Standards in Hematology (2013) [100]	American College of Critical Care Medicine (2009) [101]
Patient population	General surgery	Cardiovascular surgery	General surgery	Hemodynamically stable adult hospitalized patients including critically ill	General patient populations	Adult patients with heart disease	Critically ill patients	Adult trauma and critically ill patients.
Blood transfusion is usually indicated	Hemoglobin < 6 g/dl	Hemoglobin < 6 g/dl	Hemoglobin < 6 g/dl	Hemoglobin ≤ 7 g/dl in general including critically ill	Hemoglobin ≤ 7 g/dl in absence of major bleeding, acute coronary syndrome, or transfusion-dependent chronic anemia	Hemoglobin < 7–8 g/dl in hospitalized patients with coronary heart disease	Hemoglobin ≤ 7 g/dl	Patients with evidence of hemorrhagic shock.
		Hemoglobin < 7 g/dl in postoperative period	Hemoglobin 6–8 g/dl in presence of risk factors	Hemoglobin ≤ 8 g/dl in patients undergoing orthopedic surgery, cardiac surgery, and those with preexisting cardiovascular disease	Hemoglobin ≤ 8 g/dl for patients with acute coronary syndrome		Target hemoglobin 9–10 g/dl in early-onset severe sepsis with evidence of hypoxia	Patients with evidence of acute hemorrhage and hemodynamic instability or inadequate oxygen delivery (DO₂).
		Possibly higher hemoglobin levels when risk of end-organ ischemia exists	Hemoglobin 6–10 g/dl if symptoms of hypoxia are present				Target hemoglobin > 7 g/dl in late-onset severe sepsis	Hemoglobin < 7 g/dl in critically ill patients requiring mechanical ventilation.
							Target hemoglobin 9 g/dl in traumatic brain injury and/or cerebral ischemia	Hemoglobin < 7 g/dl in resuscitated critically ill trauma patients
							Target hemoglobin > 8–10 g/dl in subarachnoid hemorrhage	Hemoglobin < 7 g/dl in critically ill patients with stable cardiac disease.
							Target hemoglobin 8–9 g/dl in acute coronary syndrome	Hemoglobin ≤ 8 g/dl on hospital admission in patients with acute coronary syndrome.
							Target hemoglobin > 7 g/dl in stable angina
Blood transfusion rarely indicated	Hemoglobin > 10 g/dl	Hemoglobin > 10 g/dl	Hemoglobin > 10 g/dl	Hemoglobin > 10 g/dl	–	Hemoglobin > 10 g/dl	Hemoglobin > 9 g/dl	Hemoglobin > 10 g/dl
Areas of uncertainty	Hemoglobin 6–10 g/dl	–	–	Patients with acute coronary syndrome, severe thrombocytopenia (patients treated for hematologic or oncologic reasons who are at risk of bleeding), chronic transfusion-dependent anemia (not recommended due to insufficient evidence)	–	–	–	Patients with sepsis, patients at risk of acute lung injury and acute respiratory distress syndrome, patients with neurologic injury and diseases.
Other factors to consider	Ischemia, extent/rate of bleeding, volume status, risk factors for hypoxia complications	Age, severity of illness, cardiac function, ischemia, extent/rate of blood loss, mixed venous oxygen saturation (SVO₂)	Rate of blood loss, risk factors, symptoms of hypoxia/ischemia	Symptoms of hypoxia (chest pain, orthostatic hypotension, unresponsive tachycardia, heart failure)	–	–	More caution with hemoglobin trigger of 7 g/dl if patient is elderly with significant cardiorespiratory comorbidities (target hemoglobin 7–9 g/dl)	While transfusion at a hemoglobin threshold of < 7 g/dl is as effective as a hemoglobin threshold of < 10 g/dl in patients with hemodynamically stable anemia except those with acute myocardial infarction or unstable myocardial ischemia, use of hemoglobin levels as trigger for transfusion should be avoided; decision for RBC transfusion should be based on an individual patient’s intravascular volume status, evidence of shock, duration and extent of anemia, and cardiopulmonary physiologic parameters.