Perioperative Anemia and Transfusion Management in Older Adults with Osteoporotic Hip Fractures: A Narrative Review of Current Evidence and Emerging Strategies

Introduction

Osteoporotic hip fractures (OHFs) represent a major public health challenge, particularly in aging populations, and are associated with substantial morbidity, mortality, and healthcare costs.¹ As reported, the annual incidence in adults aged 45 years and older is approximately 148.7 per 100,000, with a female-to-male ratio approaching 3:1.² The perioperative period for these frail patients is fraught with physiological disturbances, with anemia standing out as a pervasive and consequential complication. The management of perioperative anemia and the judicious use of blood transfusions are central to the multidisciplinary care of patients with OHFs, directly impacting clinical outcomes and resource utilization.³

Anemia is exceedingly common in elderly patients presenting with OHFs. Multiple studies confirm that preoperative anemia is present in approximately 40–60% of patients,^4–7 with moderate-to-severe anemia accounting for over 20% of cases.^8,9 This high prevalence is inextricably linked to the chronic disease burden, nutritional deficiency, and age-related decline in hematopoietic reserve characteristic of this population.¹⁰ Nationwide study reports transfusion rates during admission ranging from approximately 17% to 37%, underscoring the clinical frequency of this issue.¹¹ Postoperatively, the prevalence escalates further due to fracture- and surgery-related blood loss.^12,13 The burden of anemia extends far beyond laboratory values, as it is intricately linked to a spectrum of adverse short- and long-term outcomes.^14–16 Yombi et al confirmed that low Hb at admission was significantly associated with mortality at 90, 180 days, and one year after surgery.¹⁷ Furthermore, allogeneic red blood cell transfusion (ARBCT) itself has also been identified as an independent risk factor for short-term mortality in recent studies.^18,19 Beyond mortality, preoperative anemia has also been independently associated with an increased risk of in-hospital postoperative complications in large OHFs cohorts.²⁰

Besides mortality, the detrimental effects of anemia permeate functional recovery and quality of life. Patients with severe preoperative anemia exhibit poorer physical function, lower Harris Hip Scores, and reduced health-related quality of life (SF-36) at six months post-surgery compared to those with higher Hb levels.²¹ The severity of anemia appears to exhibit a dose-response relationship with adverse outcomes that Hb below 10 g/dL or even below 7.1 g/dL, is associated with a further escalation in the risk of complications.^9,22 In addition, the cumulative time spent with low perioperative Hb levels has also been linked to adverse outcomes including delirium and a trend towards higher mortality.²³ Consistently, a recent observational cohort study (n = 322) demonstrated that anemia at admission was positively associated with delirium, postoperative complications, length of hospital stay and all-cause mortality in hip fracture patients.²⁴ From a healthcare resource perspective, anemia is also a significant risk factor of increased length of hospital stay (LOS). A nonlinear relationship exists, where for Hb levels below 10 g/dL, each 1 g/dL increase is associated with a reduction in LOS of approximately 0.735 days, highlighting a critical threshold for intervention. Above this level, the association is no longer statistically significant, suggesting a floor effect of particular clinical relevance.²⁵ Similarly, anemia also contributes to higher rates of hospital readmission and greater consumption of resources, including intensive care.^26,27

The etiology of perioperative anemia in OHF patients is multifactorial, involving patient-related, fracture-related, and management-related factors. Advanced age, female sex, lower body mass index (BMI), and a higher burden of comorbidities are independent patient-related risk factors for requiring postoperative blood transfusion.^11,19,27,28 Comorbidities such as Parkinson’s disease, type 2 diabetes with high HbA1c variability, and cognitive impairment further compound fracture risk and may independently worsen perioperative anemia.^29,30 Nutritional status, reflected by indices like the Prognostic Nutritional Index (PNI) or Geriatric Nutritional Risk Index (GNRI) is also associated with lower perioperative Hb, higher transfusion rates, and increased mortality.^31,32 Furthermore, the type and location of the fracture significantly influence blood loss and transfusion requirements. Previous studies reported that intertrochanteric and subtrochanteric fractures are associated with markedly greater preoperative Hb decline and higher risk-adjusted incidence of postoperative transfusion compared to femoral neck fractures.^11,33–35 Accordingly, the choice of surgical implant also plays a role that intramedullary nailing for pertrochanteric fractures correlating with higher transfusion rates compared to other fixation methods.³⁶

In summary, perioperative anemia imposes a heavy burden on patients with OHF, exacerbating mortality risk, impairing functional recovery, increasing complication rates, and straining healthcare systems. Its high prevalence and profound impact necessitate a standardized, evidence-based approach to its detection, evaluation, and management throughout the perioperative period. This article is a narrative review intended to provide an expert, integrative synthesis of current concepts, controversies, and emerging strategies in perioperative blood management for older adults with osteoporotic hip fractures, rather than a systematic review. We searched PubMed/MEDLINE, Embase, the Cochrane Library, and Web of Science for English-language articles published up to January 2026, using combinations of terms including “hip fracture”, “osteoporotic”, “perioperative an(a)emia”, “blood transfusion”, “transfusion threshold”, “tranexamic acid”, “iron”, “anticoagulant”, and “patient blood management”. We prioritized randomized controlled trials, systematic reviews and meta-analyses, and large observational and cohort studies, and screened the reference lists of key articles to identify additional relevant publications. Given the breadth of the topic and the marked heterogeneity of the available evidence, a narrative approach was deliberately adopted in order to integrate clinical, pharmacological, and organizational perspectives into a coherent, practice-oriented synthesis; accordingly, no formal quantitative pooling or risk-of-bias assessment was undertaken, and the cited literature was selected to be representative rather than exhaustive. Throughout, we frame perioperative blood management around the three complementary pillars of Patient Blood Management—optimizing erythropoiesis, minimizing blood loss, and enhancing tolerance of anemia, which together provide the conceptual structure for the strategies discussed in the sections that follow.

Preoperative Assessment and Optimization of Anemia

Given the significant burden of perioperative anemia, its systematic assessment and proactive management from the moment of hospital admission become important. While OHF surgery is often an emergency procedure, a strategic window exists for preoperative evaluation and targeted intervention to mitigate risks and improve outcomes. A comprehensive preoperative assessment aims not only to diagnose anemia but also to understand its etiology, identify patients at high risk for requiring ARBCT, and initiate corrective measures within the constraints of an expedited surgical timeline.

Initial Recognition and Etiological Understand

The initial step involves prompt recognition and quantification of anemia. A substantial proportion of elderly OHF patients present with anemia upon admission,^6,7 which is a powerful independent predictor of both short- and long-term mortality, even after adjusting for comorbidities and age.^14–17,37 The assessment begins with a complete blood count to determine the Hb concentration and red cell indices. The World Health Organization (WHO) definition of anemia provides a standard threshold, but clinical significance is graded, with severe anemia being particularly detrimental and linked to worse functional outcomes and increased LOS.^21,25

Beyond the simple Hb value, a thorough preoperative evaluation seeks to identify contributing factors and associated risks. The etiology of anemia in this population is complex, encompassing anemia of chronic disease, nutritional deficiency (iron, vitamin B12, folate), and acute hemorrhage from the fracture.^10,38 Key laboratory parameters include serum ferritin and transferrin saturation to assess iron stores, creatinine to evaluate renal function, and albumin as a marker of nutritional status.^39,40 Comorbidities, quantified by tools such as the Charlson Comorbidity Index (CCI) and modified Elixhauser’s Comorbidity Measure (mECM), are linked with both the presence of anemia and the need for perioperative transfusion.^14,41–43 Other modifiable factors, such as hypocalcemia (corrected calcium <2.11 mmol/L), have also been identified as independent associates of increased perioperative blood loss and transfusion risk.⁴⁴ Furthermore, endocrine responses to trauma, such as Euthyroid Sick Syndrome, may also exacerbate anemia risk.⁴⁵

Risk Stratification: From Traditional Models to Machine Learning

Risk stratification is a cornerstone of effective preoperative planning (Table 1). Several studies have developed and validated predictive models to identify patients at high risk for requiring perioperative ARBCT. Traditional models consistently highlight a core set of independent risk factors including lower preoperative Hb level, advanced age, the presence of specific fracture types, and poorer overall health status.^{11,33,46–48}

Table 1 Risk Factors for Perioperative Blood Transfusion in Elderly Patients with Osteoporotic Hip Fractures

In recent years, machine learning (ML) algorithms have demonstrated a marked superiority over traditional statistical methods in predicting perioperative transfusion risk, capable of capturing complex, non-linear relationships and interaction effects that conventional models may miss. Zhou et al developed a random forest (RF) model that achieved an area under the receiver operating characteristic curve (AUC) of 0.887 in internal validation and 0.834 in external validation, identifying estimated surgical duration, preoperative anemia, advanced age, hypoalbuminemia, and internal fixation as the top five predictive variables.⁴ Guo et al compared eight ML algorithms and found that the XGBoost model performed optimally, with Matthews Correlation Coefficients of 0.828 and 0.939 in training and validation sets, respectively, selecting eight key features including age, fracture timing, fracture type, Hb, albumin, creatinine, ionized calcium, and activated partial thromboplastin time.³⁹ Chen et al developed five ML models with a mean AUC of 0.912 ± 0.021, and SHAP (SHapley Additive exPlanations) analysis confirmed Hb as the most influential predictor.⁵³ Wei et al further developed a novel predictive nomogram specifically for proximal femoral anti-rotation nailing (PFNA), incorporating variables such as albumin and creatinine with high discriminatory power.⁴⁰ Importantly, ML approaches have revealed interaction effects not captured by traditional models, which further improved model performance.⁵³

Despite these advances, the widespread clinical deployment of ML-based prediction models faces considerable challenges. Most published models have been derived from single-center, retrospective datasets and report discriminative performance (AUC) with limited assessment of calibration; relatively few have undergone rigorous external validation, and fewer still have been validated prospectively or across temporally and geographically distinct cohorts, raising concerns regarding overfitting, optimism, and dataset shift. Additional barriers include limited model interpretability, demanding data-quality and missing-data requirements, and uncertain generalizability across diverse healthcare settings. Critically, no model has yet demonstrated improved patient-centered outcomes or clinical utility in a prospective impact study. Future research should therefore prioritize external and prospective validation, transparent calibration reporting, and the development of interpretable, user-friendly clinical decision-support systems integrated into electronic health records before such tools can be recommended for routine practice.^39,40

Preoperative Optimization Strategies

Preoperative optimization, though time-limited, encompasses several evidence-based strategies. The most immediate decision often revolves around preoperative ARBCT. A restrictive transfusion strategy, typically triggered at an Hb threshold of 8 g/dL or for symptomatic anemia, is now widely recommended and has been shown to be safe without increasing mortality or major complications compared to a liberal strategy.⁶⁶ Meta-analysis indicate that preoperative intravenous (IV) iron can enhance Hb recovery, reduce postoperative transfusion rates, and is associated with lower short-term mortality.⁶⁷ In addition, the combination of IV iron and erythropoietin (EPO) appears effective in stimulating erythropoiesis and reducing ARBCT when administered preoperatively.⁶⁸ While a large randomized trial did not show a reduction in transfusion rates with ferric carboxymaltose with or without EPO, it confirmed improved Hb recovery.⁶⁹ Correction of other deficiencies, such as vitamin D and vitamin K, may also form part of a comprehensive nutritional support strategy, though direct evidence for impacting transfusion needs is less robust.^12,70

The timing of surgery itself is a critical variable in the perioperative anemia equation. While expedited surgery is a key standard of care to reduce morbidity and mortality, its relationship with transfusion requirements is nuanced. Recently, several studies suggest that surgery within the first 48 hours may be associated with a higher transfusion rate.^33,43 However, Kim et al found that delaying surgery primarily increases preoperative transfusion needs without necessarily altering the total perioperative transfusion volume.⁷¹ Therefore, the decision must balance the general benefits of early surgery against the opportunity for targeted anemia management in selected high-risk patients.

Frailty, Anemia, and Comprehensive Geriatric Assessment

Anemia and frailty are closely and bidirectionally linked in older adults with OHFs, and their co-occurrence has important physiological implications for perioperative blood management. Frailty denotes a state of diminished physiological reserve and reduced capacity to withstand stressors, and anemia is both a recognized contributor to, and a clinical marker of, this state. Several overlapping mechanisms underlie this relationship. The chronic low-grade inflammation characteristic of aging (“inflammaging”), with elevated interleukin-6 and other pro-inflammatory cytokines, upregulates hepcidin, thereby restricting intestinal iron absorption and sequestering iron within the reticuloendothelial system to produce a functional iron deficiency and anemia of inflammation that is relatively refractory to oral iron. Aging is further associated with a blunted erythropoietin response, reduced hematopoietic reserve, a higher prevalence of nutritional deficiencies (iron, vitamin B12, folate), and impaired renal function, all of which limit the erythropoietic response to the acute blood loss of fracture and surgery. Critically, the reduced cardiovascular and physiological reserve of frail patients lowers their capacity to compensate for a given degree of anemia, narrowing the margin of safety and rendering them more vulnerable to the consequences of both anemia and its treatments. These considerations help explain why a single hemoglobin threshold is inadequate in this population and why frailty status should be formally integrated into perioperative decision-making. Comprehensive geriatric assessment as a multidimensional evaluation of medical, functional, cognitive, nutritional, and social domains, delivered within orthogeriatric co-management models provides a structured means of identifying frailty, anticipating transfusion requirements, and individualizing optimization, and has been associated with improved outcomes after hip fracture.^72–75

The relationship between frailty and perioperative transfusion warrants specific comment. Frailty appears to act both as an independent risk factor and as a marker of shared underlying vulnerability. On the one hand, validated frailty measures independently predict adverse perioperative outcomes after hip fracture surgery: a preoperative modified Frailty Index (mFI) is independently associated with a higher severity of postoperative complications,⁴⁹ and frailty predicts postoperative mortality more strongly than chronological age.⁵⁰ Because frailer patients tend to sustain greater perioperative blood loss and are transfused more frequently, frailty is also closely associated with increased transfusion requirements. On the other hand, frailty overlaps mechanistically with anemia, malnutrition, and comorbidity, such that part of its association with transfusion is likely mediated through these shared pathways; consistent with this, the propensity to be transfused in older surgical patients has been linked more closely to comorbidity burden than to age per se.⁵¹ Frailty should therefore be regarded as both a genuine independent predictor and, in part, a downstream marker of cumulative physiological deficit—an interpretation that reinforces the rationale for formal frailty assessment rather than reliance on any single parameter. More broadly, frailty seldom occurs in isolation but frequently coexists with other geriatric syndromes including sarcopenia, malnutrition, cognitive impairment, and polypharmacy that further erode physiological reserve and recovery capacity; sarcopenia in particular overlaps with both anemia and frailty and is associated with poorer functional recovery and worse outcomes after hip fracture. These considerations reinforce the value of incorporating formal frailty, nutritional, and sarcopenia assessment within comprehensive geriatric evaluation when planning perioperative blood management.

Pharmacological Interventions and Blood Conservation Strategies

Building upon a multidisciplinary preoperative assessment, targeted pharmacological and procedural interventions during the intraoperative and immediate postoperative phases are crucial components of a comprehensive Patient Blood Management (PBM) protocol for OHFs. The goal is to minimize iatrogenic blood loss, optimize hemostasis, and support erythropoiesis, thereby reducing the reliance on ARBCT and its associated risks.

Tranexamic Acid

The most extensively studied and supported pharmacological agent is tranexamic acid (TXA), an antifibrinolytic that inhibits the breakdown of blood clots by blocking the lysine-binding sites on plasminogen. A Cochrane overview of systematic reviews, encompassing data from 21 RCTs and 2148 participants, concluded that TXA probably reduces the need for ARBCT in adults undergoing OHF surgery, with a risk ratio of 0.56, translating to approximately 194 fewer people per 1000 requiring a transfusion.³⁸ This efficacy is consistently demonstrated across various meta-analyses, regardless of fracture type or surgical procedure including in hip hemiarthroplasty,⁷⁶ intertrochanteric fracture surgery,⁷⁷ and femoral neck fractures treated with arthroplasty.⁷⁸ Moreover, the reduction in transfusion requirements is paralleled by a significant decrease in calculated total blood loss, often by over 200 mL,^79,80 and an attenuated postoperative decline in Hb levels.^81,82

Importantly, the safety of TXA in this vulnerable population appears favorable. Multiple high-level syntheses have found no statistically significant increase in the risk of thromboembolic events, including deep vein thrombosis (DVT) and pulmonary embolism (PE), myocardial infarction, cerebrovascular accidents, or mortality when compared to placebo or control.^{38,79–81,83} In a meta-analysis encompassing over 54,000 patients, Augustinus et al reported that 30-day mortality was lower in the TXA group.⁷⁶ One large observational study found a higher incidence of diagnosed PE with TXA use but without an associated increase in 30-day mortality.⁸⁴ Nonetheless, these reassuring data derive largely from randomized trials and meta-analyses that may be underpowered to detect rare thromboembolic events and that frequently excluded patients at the highest thrombotic risk, such as those with recent venous thromboembolism, active arterial or venous thrombosis, or severe renal impairment. The favorable safety profile of TXA should therefore not be over-generalized to these subgroups, in whom dedicated evidence remains sparse and individualized caution is warranted. The interaction between TXA and preoperative chemoprophylaxis for DVT does not appear to alter its effect on transfusion rates, suggesting they can be used concurrently.⁸⁵

Regarding administration, evidence supports both IV and topical routes with similar efficacy and safety.^78,86 Oral TXA (2 g, 2 hours preoperatively) has been shown to be non-inferior to IV TXA (1 g, 15 minutes preoperatively) in terms of total blood loss and transfusion rates, with significantly lower cost, offering a more economical alternative.⁸⁶ The combination of IV and topical administration may provide synergistic hemostatic benefit for intertrochanteric fracture fixation compared to either route alone. While single and multiple dosing regimens show comparable reductions in blood loss,⁷⁹ a single preoperative IV dose of 1 g or 15 mg/kg is widely used and effective.^87,88 Evidence suggests that very early, pre-emptive administration (within 72 hours of injury, especially within 24 hours) may be particularly effective in reducing hidden blood loss (HBL) that occurs prior to surgery.⁸⁹ However, delayed administration beyond 72 hours may not confer the same benefit.⁹⁰ An emerging and clinically important consideration is the influence of nutritional status on TXA efficacy. Xie et al demonstrated that patients with adequate nutritional status, as defined by the GNRI, exhibited significant reductions in transfusion rates and Hb decline with TXA, whereas the effect was markedly attenuated in malnourished patients.⁵² This finding suggests that nutritional status may be a key modulator of TXA responsiveness and warrants consideration when evaluating the expected benefit of antifibrinolytic therapy (Table 2).

Table 2 Summary of Key Evidence for Tranexamic Acid (TXA) in Hip Fracture Surgery

Intravenous Iron Supplementation

In contrast to the strong evidence for TXA, the role of IV iron supplementation in the acute perioperative setting is less definitive. The aforementioned Cochrane overview found only low-certainty evidence from a few small studies suggesting there may be little or no difference in transfusion requirements, infection rates, or mortality with IV iron administration.³⁸ A meta-analysis including six RCTs found that perioperative IV iron had no significant impact on transfusion rates.⁹² The IRON NOF trial similarly found no reduction in the overall proportion of patients transfused, though a possible benefit for younger patients was noted.⁹³ Chen et al⁹⁴ reported that iron supplementation was not associated with a reduced transfusion rate in the overall analysis but did note a significant reduction in transfusion volume and LOS, and a potential reduction in infection risk. Subgroup analyses suggest that preoperative IV iron, particularly at doses of 200–300 mg, may be the most beneficial regimen. Another meta-analysis focusing on acute surgery found that IV iron therapy was associated with a lower 30-day mortality and a reduction in postoperative infections and ARBCT, though these findings were largely driven by observational studies.⁶⁷

The combination of IV iron with EPO has been investigated, with one RCT finding that while it improved Hb recovery at 60 days post-discharge, it did not reduce the immediate need for perioperative ARBCT compared to placebo.⁶⁹ A particularly informative contribution is the HiFIT trial, a multicenter, 2×2 factorial, double-blind RCT, which evaluated ferric derisomaltose (20 mg/kg) and TXA, alone or combined, for reducing ARBCT in OHF patients. Interestingly, neither iron nor TXA alone significantly reduced transfusion rates in this trial, but their combination produced a significant reduction, suggesting a potential synergistic mechanism whereby TXA reduces acute loss while iron supports erythropoietic recovery, which is a key finding that may reshape future multimodal PBM protocols.⁹¹ Therefore, while IV iron appears safe and may aid in longer-term Hb recovery and potentially reduce LOS, its ability to independently and reliably avert perioperative transfusions in OHF requires confirmation from larger, high-quality RCTs. Its greatest value may lie as a component of multimodal PBM, particularly when combined with TXA.

Surgical Technique and Implant Choice

Beyond specific hemostatic drugs, the choice of surgical implant and technique constitutes a critical non-pharmacological blood conservation strategy. For intertrochanteric fractures, the debate between extramedullary (eg, dynamic hip screw, DHS) and intramedullary (eg, cephalomedullary nail, CMN) fixation has implications for blood loss. While previous meta-analyses found no significant difference in transfusion rates between the two,^95,96 more recent and specific analyses suggest nuances based on fracture stability. For stable (AO/OTA 31-A1) intertrochanteric fractures, several studies indicate that DHS fixation is associated with less intraoperative blood loss and lower postoperative transfusion rates compared to CMN.⁵⁷ Cai et al conducted a prospective RCT and confirmed that intramedullary fixation was associated with significantly greater total blood loss and HBL.⁵⁸ For unstable (AO/OTA 31-A2/A3) patterns, the evidence is mixed, with some studies favoring CMN^36,59 and others suggesting DHS with a trochanteric stabilization plate may be a valid alternative.⁹⁷ In addition, when an intramedullary nail is chosen for an intertrochanteric fracture, the use of a short nail is associated with less estimated blood loss, shorter operative time, and a lower transfusion rate compared to a long nail, without compromising healing.^60–62 In regards of femoral neck fractures requiring arthroplasty, hemiarthroplasty (HA) is consistently associated with less operative blood loss and lower transfusion rates than total hip arthroplasty (THA).^63–65

Besides, minimally invasive surgical techniques also contribute to blood conservation. A randomized trial comparing a minimally invasive side plate system to a conventional implant for trochanteric fractures found a significant reduction in perioperative bleeding with the minimally invasive approach.⁹⁸ Furthermore, external fixation, as a minimally invasive alternative for high-risk patients, has been shown to drastically reduce transfusion needs and LOS compared to DHS.⁹⁹

Hidden Blood Loss and Other Perioperative Considerations

A significant and often underappreciated component of perioperative blood loss is HBL, which comprises extravasation into tissues and hemolysis, not captured in surgical suction or drains. In elderly patients with intertrochanteric fractures undergoing intramedullary nailing, HBL can account for over 80% of total perioperative blood loss.^37,54 Factors independently associated with increased perioperative HBL include unstable fracture patterns (AO/OTA 31-A2.2 to A3.3), male sex, hypertension, low bone density, longer operative time (≥60 min), time from injury to operation (<2 days), blood transfusion, and without anticoagulation.^37,55,56,100 Cooperated with early administration of TXA, meticulous surgical technique to minimize soft tissue dissection and operative time, alongside the use of minimally invasive approaches where appropriate, are practical measures to reduce both visible and hidden losses.

Other perioperative strategies hold physiological rationale. Intraoperative hypothermia is a modifiable risk factor for coagulopathy and increased bleeding, maintaining normothermia is therefore a simple yet important component of blood conservation.¹⁰¹ Intraoperative cell salvage is being actively investigated through RCTs like the WHITE 9 study, which aims to determine its clinical and cost-effectiveness in OHF surgery.¹⁰² Local hemostatic agents such as fibrin sealants, have also demonstrated reductions in transfusion rates compared to standard drainage in HA for femoral neck fractures.¹⁰³

Transfusion Strategies: Thresholds, Outcomes, and the Case for Individualization

Despite comprehensive blood conservation strategies, ARBCT remains a common intervention in OHF perioperative care, with reported rates varying widely from 12.1% to over 40% across different settings.^11,104 This variation underscores the absence of a universal consensus on the optimal transfusion trigger, making the development of evidence-based strategies for when to transfuse a critical component of perioperative management.

The Risk-Benefit Analysis of Allogeneic Transfusion

The decision to administer blood is complex, balancing the potential benefits of correcting anemia to improve oxygen delivery against the well-documented risks associated with transfusion, including immunomodulation, transfusion-related acute lung injury (TRALI), acute kidney injury (AKI), transfusion-associated circulatory overload (TACO), and infection.^105–109 Regarding OHF, where patients are typically elderly with multiple comorbidities, this risk-benefit analysis is particularly nuanced. Observational data consistently link perioperative blood transfusion with adverse outcomes such as increased mortality, LOS, and higher readmission rates.^18,19,27 A nationwide study in Scotland found transfusion was independently associated with higher 30- and 60-day mortality, delayed mobilization, and longer LOS. The authors concluded this was most likely a marker of perioperative anemia and patient vulnerability rather than a direct causal effect of transfusion.¹¹ This is supported by analyses using propensity scoring to adjust for the likelihood of receiving a transfusion, which often negate the independent association between transfusion and mortality.¹¹⁰ Furthermore, the characteristics of the blood product itself, such as the storage duration, have not been conclusively shown to affect mortality or morbidity in OHF patients.¹¹¹ Therefore, the focus should shift from viewing transfusion as a primary risk factor to recognizing it as a signal of a high-risk patient recognition, where the underlying anemia and its causes are the primary therapeutic targets.

The Restrictive versus Liberal Threshold Debate

Regarding the debate between restrictive and liberal transfusion thresholds, a restrictive strategy typically involves transfusing only when Hb falls below a specific trigger (eg, 7.0–8.0 g/dL) in asymptomatic, hemodynamically stable patients, while a liberal strategy uses a higher trigger (eg, 9.0–10.0 g/dL). A landmark Cochrane meta-analysis of 61 RCTs demonstrated that a restrictive threshold reduces transfusion rates by 42% without increasing 30-day mortality.¹¹² A Cochrane review specific to OHF surgery and including six trials, found low-quality evidence of no significant difference in 30- or 60-day mortality, functional recovery, or most postoperative morbidities between liberal (Hb ~10 g/dL) and restrictive (Hb ~8 g/dL) thresholds. Notably, this review suggested a possible increased risk of myocardial infarction with a restrictive strategy.¹⁰⁵ The FOCUS trial, as the largest RCT in this field, demonstrated no significant difference in three-year mortality or cause of death between a liberal strategy (maintaining Hb ≥10 g/dL) and a restrictive strategy (transfusion for symptoms or Hb <8 g/dL), supporting the safety of restrictive transfusion in the majority of elderly OHF patients.¹¹³ Another study implementing a protocol with an even more restrictive threshold of <7 g/dL for stable patients demonstrated a significant reduction in blood utilization (transfusion rate from 51% to 33%) and inpatient cardiac morbidity, with no increase in mortality, readmissions, or LOS.¹¹⁴

However, the application of a single Hb threshold for all frail, elderly OHF patients is challenged by evidence suggesting that certain vulnerable subgroups may benefit from a more liberal strategy. The TRIFE RCT in frail anemic patients from nursing homes or sheltered housing found no overall benefit of a liberal strategy (Hb target 11.3 g/dL) over a restrictive one (9.7 g/dL) for physical recovery or mortality at 30 days. However, a per-protocol analysis revealed a higher 30-day mortality with the restrictive strategy, and a prespecified subgroup analysis demonstrated a statistically significant lower 90-day mortality with the liberal strategy among nursing home residents (20% vs. 36%).¹¹⁵ A post-hoc analysis of this trial further demonstrated that a liberal transfusion strategy was associated with a significantly lower risk of postoperative delirium on day 10 compared to a restrictive strategy.¹¹⁶ This is a critical finding given that postoperative delirium is itself a strong independent predictor of increased 90-day mortality. Another RCT in cognitively intact patients found that while a liberal threshold (10.0 g/dL) did not improve ambulation scores compared to a restrictive one (8.0 g/dL), it was associated with fewer cardiovascular complications (2% vs. 10%) and lower in-hospital mortality (0% vs. 8%).¹¹⁷ In a meta-analysis, Gu et al specifically examining orthopedic patients revealed that restrictive strategies increased the risk of cardiovascular events, with the effect being particularly pronounced in OHF surgery.¹¹⁸ These studies collectively highlight that the factors beyond a simple Hb number such as baseline frailty, cognitive status, cardiovascular comorbidity, and the presence of symptoms like delirium or hemodynamic instability, must be integrated into the transfusion decision (Table 3).

Table 3 Comparison of Restrictive versus Liberal Transfusion Strategies: Key Clinical Trials and Meta-Analyses in Hip Fracture Patients

Individualized, Physiologically Guided Transfusion

The evidence above underscores the need for a more individualized, physiologically guided approach. The concept of “symptomatic anemia” is often cited in restrictive protocols but remains poorly defined. Emerging research points to dynamic physiological parameters that could refine decision-making. A multicenter cohort study in acute surgical patients including those with OHF, found that the combination of low Hb (<9.7 g/dL) and a low peripheral perfusion index (PPI<1.5) was strongly associated with increased 90-day mortality and severe complications. Notably, patients with low Hb but high PPI had a significantly better outcome, suggesting that PPI may help stratify risk among anemic patients and identify those who might benefit from earlier transfusion irrespective of the absolute Hb level.¹²⁰ Furthermore, the duration and burden of anemia may be as important as its nadir. A prospective observational study using continuous non-invasive Hb monitoring revealed a significant delay in detecting anemia via intermittent blood sampling and found that the total accumulated time spent with low perioperative Hb (<10 g/dL) was associated with a higher incidence of postoperative delirium, suggesting that minimizing both the depth and duration of anemia, rather than just reacting to a single low value, could be a relevant therapeutic goal.²³ The concept of the oxygen extraction ratio (P(v-a)CO₂/C(a-v)O₂), which reflects the balance between oxygen delivery and consumption, has been proposed as a physiologically rational transfusion trigger in sepsis and may have potential applicability in the perioperative OHF setting.¹²¹ However, it must be critically noted that these physiological parameters and continuous monitoring modalities remain largely exploratory in the orthogeriatric population; their diagnostic accuracy, cost-effectiveness, and definitive impact on clinical hard outcomes require validation in large-scale randomized trials before they can complement or replace standard hemoglobin triggers.

In clinical practice, the transfusion decision should therefore not be reduced to a single Hb threshold. A personalized approach should integrate the absolute Hb level with the patient’s symptom profile including cognitive and cardiovascular status, hemodynamic stability, anticipated further blood loss, trajectory of Hb change, and comorbid disease burden.^114,117,119 The RESULT-NOF feasibility trial, which compared a restrictive (Hb <8 g/dL) and a liberal (Hb <10 g/dL) strategy specifically to prevent myocardial injury, represents ongoing efforts to define optimal thresholds for high-risk subgroups.¹¹⁹

Patient Blood Management Programs

The success of a restrictive transfusion policy is heavily dependent on parallel efforts to reduce bleeding and stimulate erythropoiesis within a formalized PBM approach. The implementation of a formal, multimodal PBM program has been shown to improve outcomes. A study evaluating a revised PBM protocol demonstrated a significant increase in the appropriateness of ARBCT transfusions (from 54% to 95%) alongside a decrease in transfusion rates, LOS, and 30-day readmissions.¹²² Observational data specifically associated IV iron administration on the third postoperative day with a lower 30-day mortality risk compared to no treatment in anemic OHF patients, highlighting that structured, protocol-driven care focusing on the entire patient management can optimize blood use and improve quality of care beyond merely adjusting a transfusion trigger.¹²³ Comprehensive multimodal PBM incorporating perioperative IV iron (with or without rHuEPO) has been associated with reduced ARBCT rates, lower postoperative infection rates, and decreased 30-day mortality in pooled analyses.¹²⁴ The widespread implementation of PBM remains variable across institutions, and harmonizing practice through standardized protocols is a key priority.¹²⁵ An integrated, stage-based PBM pathway summarizing these recommendations is presented in Figure 1.

Figure 1 Multimodal Patient Blood Management (PBM) pathway for older adults with osteoporotic Hip fracture. Abbreviations: Hb, hemoglobin; GNRI, Geriatric Nutritional Risk Index; PNI, Prognostic Nutritional Index; CCI, Charlson Comorbidity Index; BMI, body mass index; ML, machine learning; MCC, Matthews Correlation Coefficient; AUC, area under the receiver operating characteristic curve; OHFs, osteoporotic hip fractures; SHAP, SHapley Additive exPlanations; IV, intravenous; EPO, erythropoietin; PBM, Patient Blood Management; DOACs, direct oral anticoagulants; TXA, tranexamic acid; ITF, intertrochanteric fractures; IMN, intramedullary nail; HA, hemiarthroplasty; THA, total hip arthroplasty; PPI, peripheral perfusion index; TACO, transfusion-associated circulatory; TRALI, transfusion-related acute lung injury; DAAH, days alive and at home. Notes: Arrows indicate the sequential progression of steps and the decision-making flow within the algorithm. ↑ indicates an increase or improvement, and ↓ indicates a reduction or decrease in the respective metric.

Perioperative Management of Patients on Antithrombotic Therapy

A quintessential example of the high-risk patient cohort requiring individualized perioperative blood management is the growing cohort of elderly OHF patients receiving antithrombotic therapy. The management of antiplatelet agents (eg, clopidogrel, aspirin) and oral anticoagulants (vitamin K antagonists, direct oral anticoagulants) presents a significant clinical dilemma: balancing the imperatives of early surgical fixation against the perceived risks of increased perioperative bleeding. This conflict frequently results in surgical delays, which themselves are independently associated with poorer outcomes.^126,127 Contemporary evidence, however, increasingly challenges the necessity of such delays, advocating for a more nuanced, patient-centered approach (Table 4).

Table 4 Perioperative Management of Antithrombotic Therapy in Elderly Hip Fracture Patients: Evidence-Based Recommendations

Patients on Clopidogrel and Antiplatelet Agents

For patients on clopidogrel, systematic reviews and meta-analyses have provided substantial evidence supporting the safety of early surgery. A meta-analysis found that early surgery in patients on clopidogrel was not associated with increased overall or 30-day mortality, while providing a significantly reduced LOS, despite a modest increase in transfusion risk.¹²⁸ Soo et al concluded that early surgery was safe without a statistically significant increase in transfusion odds, recommending against withholding clopidogrel perioperatively due to the elevated risk of cardiovascular events associated with its discontinuation.¹²⁹ Indeed, withdrawal of clopidogrel, particularly for 4 to 8 days, is associated with a significantly increased incidence of in-hospital acute coronary syndrome.¹³⁰ Research indicates that while the timing of bleeding may differ with early surgery leading to more intraoperative loss and delayed surgery to more preoperative loss, the total blood loss and transfusion requirements are often comparable.¹³¹ Furthermore, delaying surgery in patients on dual antiplatelet therapy does not reduce transfusion need but is associated with increased probabilities of major complications and higher 30-day mortality.¹²⁶ For patients on aspirin alone, propensity-matched analyses suggest it does not significantly increase perioperative HBL or transfusion risk, supporting immediate surgery.^132,133

Patients on Direct Oral Anticoagulants

The management of patients on direct oral anticoagulants (DOACs) has also been studied. A systematic review concluded that performing surgery within 48 hours for DOAC-treated patients appears safe, with only a small associated increase in blood transfusion risk and no increase in 30-day mortality, although the underlying evidence is predominantly observational.¹³⁴ Numerous cohort studies are consistent with this finding, reporting that expedited surgery is not associated with increased perioperative blood loss, transfusion rates, or major bleeding complications compared with delayed surgery.^135–139 These studies are, however, susceptible to confounding by indication, as the decision to operate early may itself reflect a lower perceived bleeding risk. Brameier et al found that surgery within 48 hours of the last DOAC dose was associated with lower transfusion requirements (OR 2.39 for delayed surgery), shorter hospital stays (5.9 vs. 7.6 days), and no difference in complication rates.¹³⁸ Kolodychuk et al similarly confirmed that early surgery for DOAC patients did not increase transfusion risk or mortality.¹⁴⁰ While DOAC use is generally associated with a longer time to surgery compared to non-anticoagulated patients,^{127,142,144–146} protocols that do not mandate arbitrary delays can achieve comparable timelines.¹⁴⁷

Patients on Vitamin K Antagonists and Perioperative Optimization

For patients on vitamin K antagonists, the implementation of reversal protocols using vitamin K or prothrombin complex concentrate has proven effective in reducing time to surgery without increasing bleeding risk.^134,141 Yoo et al found no significant association between preoperative anticoagulation reversal and 30-day mortality or other outcomes.¹⁴¹ In addition, Salar et al demonstrated routine preoperative coagulation screening in the absence of known anticoagulant use appears unnecessary and represents a cost-saving opportunity.¹⁴³

Several perioperative management strategies can further optimize outcomes in anticoagulated patients. The concomitant use of TXA is a key blood conservation strategy. Evidence from acetabular fracture surgery suggests that holding preoperative chemoprophylaxis does not alter TXA’s effect on transfusion rates, indicating TXA can be used effectively even in anticoagulated patients.⁸⁵ In OHF, intravenous TXA has been associated with reduced postoperative transfusion rates and 30-day readmissions.¹⁴² Regional techniques like the lateral femoral cutaneous and over the hip block have been shown to be safe in anticoagulated patients, allowing for shorter time to surgery.¹⁴⁸

It is crucial to contextualize the bleeding risk within the broader profile of these patients. Those on antithrombotic agents invariably have significant cardiovascular comorbidities, which explain their higher baseline risk. A Danish nationwide study found that preoperative antiplatelet use was associated with increased risks of transfusion and 30-day mortality, whereas DOACs were linked only to a slightly increased transfusion risk.¹⁴⁹ The HASTE study reported that anticoagulated patients had a longer time to surgery and a higher 30-day mortality rate, but no difference in transfusion rates, suggesting that the delay itself may contribute to the mortality difference.¹²⁷ Therefore, the increased mortality observed in several studies of anticoagulated patients is likely a marker of their greater burden of illness and frailty, rather than a direct consequence of the anticoagulant or its perioperative management, which is analogous to the interpretation of transfusion as a risk marker discussed previously.

Limitations

Several limitations of the current evidence base, and of this review, warrant acknowledgment. First, as a narrative review, this synthesis did not employ a formal systematic search or quantitative pooling, and the selection of studies, although intended to be representative, may be subject to selection bias. Second, much of the available evidence particularly regarding transfusion thresholds in specific subgroups, the perioperative management of anticoagulated patients, and the comparative effects of surgical techniques derives from retrospective and observational studies, which are inherently susceptible to confounding by indication and selection bias; the consistent associations between transfusion or anticoagulant use and adverse outcomes most likely reflect the greater frailty and comorbidity burden of the patients receiving these interventions rather than direct causal effects, an interpretation we have emphasized throughout. Third, prospective randomized validation remains limited for several of the strategies discussed, including individualized, physiologically guided transfusion triggers, the optimal management of antithrombotic therapy, and multimodal PBM bundles, so that a number of recommendations are necessarily provisional. Fourth, considerable clinical and methodological heterogeneity exists across studies with respect to patient populations, definitions of anemia and frailty, hemoglobin thresholds, outcome measures, and follow-up duration, limiting direct comparison and the generalizability of pooled estimates. Fifth, the evidence is least certain precisely in the most vulnerable groups of frail patients, nursing-home residents, and those with significant cardiovascular disease, in whom the balance between the risks of anemia and the risks of transfusion or surgical delay may differ from that in the broader population, yet who remain underrepresented in randomized trials. Finally, machine learning–based transfusion prediction models, although promising, currently share important limitations, most notably a predominance of single-center retrospective derivation, limited external and prospective validation, and an absence of demonstrated improvement in patient-centered outcomes. These limitations should be borne in mind when translating the evidence summarized here into individual clinical decisions.

Clinical Implications, Knowledge Gaps, and Conclusion

The perioperative management of anemia and the judicious use of blood transfusions in older adults with OHFs represent a cornerstone of high-quality, patient-centered care. This population, characterized by frailty, multimorbidity, and a high prevalence of preoperative anemia, is uniquely vulnerable to the detrimental effects of both anemia and its treatments. The accumulated evidence underscores that anemia is not merely a laboratory finding to be corrected, but a critical modifiable risk factor embedded within a complex clinical environment. Therefore, effective management transcends a simple Hb-centric transfusion trigger, demanding an individualized and proactive approach encompassing the entire perioperative management period from admission to recovery.

Core Evidence-Based Recommendations

Current evidence strongly supports the integration of several key principles into structured care pathways. First, systematic preoperative assessment and risk stratification increasingly augmented by ML-based predictive models, enable early identification of high-risk patients and targeted intervention. Second, TXA is an effective blood conservation agent with a favorable safety profile in most patients and is endorsed by major guidelines as a core component of PBM; caution nonetheless remains warranted in individuals at high thromboembolic risk, for whom dedicated evidence is limited. Emerging data further suggest that its combination with IV iron may yield synergistic benefits. Third, a restrictive transfusion strategy (Hb trigger of 7–8 g/dL) is safe for the majority of patients, but an individualized approach incorporating physiological parameters, symptom burden, and frailty status is essential for the most vulnerable subgroups, particularly nursing home residents and those at high risk for delirium or cardiovascular events. Fourth, current evidence, although predominantly observational and therefore susceptible to confounding, suggests that surgery generally need not be delayed solely on the basis of antithrombotic medication status, as the risks of delay appear to outweigh the modest incremental bleeding risk in most patients. This approach should nonetheless be individualized according to each patient’s bleeding and thrombotic risk, the specific agent involved, and local protocols, particularly in the absence of large randomized trials. The establishment of Fast Access to Surgery in Emergency (FASE) strategies and Enhanced Recovery After Surgery (ERAS) protocols demonstrates tangible benefits in streamlining care, reducing time-to-surgery, and potentially minimizing perioperative blood loss. The evolution of orthogeriatric co-management models, now extending into elective orthopedic settings, provides a robust organizational template for integrated, multidisciplinary care.

Unresolved Controversies and Evidence Gaps

Technological innovations continue to refine perioperative blood conservation, including the ongoing evaluation of intraoperative cell salvage in the WHITE 9 trial,¹⁰² and the advent of robot-assisted techniques for fracture fixation showing promise in reducing intraoperative blood loss.¹⁵⁰ The application of ML to develop increasingly accurate and interpretable predictive models for transfusion risk represents a significant advancement towards truly personalized care. Future efforts should prioritize the development of user-friendly clinical decision support systems that can be integrated into electronic health records for real-time risk assessment. The validation of emerging physiological biomarkers such as the PPI and continuous non-invasive Hb monitoring, as dynamic transfusion triggers, also warrants prospective investigation. In addition, a critical future direction is the adoption of more patient-centered outcome measures. Traditional endpoints like mortality and LOS, while important, may not fully capture the functional recovery and quality of life that matter most to older patients. For instance, the metric “days alive and at home” has been validated as a meaningful composite outcome that reflects a patient’s ability to return to their pre-fracture environment and independence.¹⁵¹ Its use in clinical trials and quality audits can shift the focus towards outcomes that align with patient priorities.

Future Perspectives and Methodological Horizons

In conclusion, the management of perioperative anemia in OHF care is a dynamic field at the intersection of geriatrics, hematology, anesthesiology, and surgery. The path forward is characterized by a shift from reactive transfusion to proactive, multimodal PBM embedded within integrated care pathways. Success depends on the seamless collaboration of a multidisciplinary team, the implementation of evidence-based protocols, the strategic adoption of supportive technologies, and a steadfast commitment to measuring what truly matters to patients. As demographic pressures mount, refining and disseminating these comprehensive management strategies will be paramount to improving survival, function, and quality of life for these growing number of older adults suffering from devastating injuries.

Furthermore, a critical paradigm shift must occur in clinical trial methodology: future research should transition from traditional, short-term endpoints (such as in-hospital length of stay or 30-day mortality) toward broader, patient-centered composite metrics. The validation and adoption of metrics like “Days Alive and At Home” (DAAH) will ensure that future blood management strategies align directly with the recovery of independence, functional mobility, and quality of life that matter most to our aging patient population. Notably, functional outcome assessments should move beyond generic mobility scores to encompass culturally specific activities of daily living. As highlighted by recent initiatives,¹⁵² recognizing and measuring distinct activities such as specific prayer movements, is essential for a more granular, inclusive, and culturally competent evaluation of post-fracture recovery.