Lactate-to-albumin ratio, C-reactive protein-to-albumin ratio, and prognostic nutritional index in acute and critical care: A focused mini-review

Abstract

Risk stratification in acute and critical care increasingly relies on accessible biomarkers that capture multiple dimensions of physiological stress. Albumin-based composite biomarkers have gained attention because serum albumin reflects not only nutritional reserve but also systemic inflammation, hepatic synthetic capacity, endothelial permeability, capillary leakage, and resuscitation-related volume shifts. Among these indices, the lactate-to-albumin ratio, C-reactive protein-to-albumin ratio (CAR), and prognostic nutritional index (PNI) are particularly practical because they combine albumin with markers of tissue hypoperfusion, inflammatory burden, and immune-nutritional status. This focused mini-review presents a narrative synthesis of literature identified through targeted searches of PubMed, Web of Science, and Google Scholar using terms related to these indices and acute or critical care populations. Evidence suggests that elevated lactate-to-albumin and CARs and reduced PNI are generally associated with mortality, organ dysfunction, intensive care requirement, postoperative complications, and prolonged hospitalization in emergency department, intensive care, perioperative, infectious, cardiovascular, respiratory, gastrointestinal, trauma, and oncological settings. However, interpretation is limited by disease heterogeneity, variable sampling times, resuscitation-related albumin changes, hepatic dysfunction, chronic inflammation, inconsistent cutoff values, and the predominance of retrospective and observational evidence. These indices should therefore be interpreted as supportive prognostic markers within the broader clinical context rather than as replacements for clinical judgment or validated prognostic models.

Key Words: Albumin; Lactate-to-albumin ratio; C-reactive protein-to-albumin ratio; Prognostic nutritional index; Critical care; Emergency department; Risk stratification; Mortality

Core Tip: The lactate-to-albumin ratio, C-reactive protein-to-albumin ratio, and prognostic nutritional index may help clinicians interpret routinely available laboratory results within a broader biological framework that includes metabolic stress, inflammation, albumin biology, and immune nutritional reserve. Their greatest potential value is not as standalone decision tools, but as adjuncts to bedside assessment and established severity scores. Future studies should determine whether serial measurements, disease specific thresholds, and integration into existing prognostic models improve clinical decision making and patient risk stratification.

INTRODUCTION

Healthcare systems internationally are experiencing a steadily increasing patient load, especially in acute and critical care environments such as emergency departments (ED), intensive care units (ICU) and post-surgical care units[1]. In the context of this increased patient burden, the rapid and accurate identification of high-risk individuals within a heterogeneous patient population is a main goal in acute and critical care[2]. Early risk stratification facilitates the timely selection of appropriate treatment strategies and improves the prediction of adverse clinical outcomes by supporting clinical decision processes[3]. Timely recognition of negative outcomes, including sudden organ dysfunction, cardiac arrest, unplanned ICU admission, and mortality, allows for the timely initiation of targeted interventions such as early fluid resuscitation, antibiotic therapy, or intensive monitoring[4]. Additionally, early prediction not only improves clinical outcomes, but also optimizes resource-dependent processes, including hospital length of stay, ICU capacity, and overall resource utilization, thus reducing the strain on healthcare systems[5].

Over the years, various clinical scoring systems have been developed to aid in risk stratification[6]. In intensive care, systemic inflammatory response syndrome, Sequential Organ Failure Assessment (SOFA), and quick SOFA are commonly used to assess severity and sepsis related risk. In cardiovascular medicine, Thrombolysis in Myocardial Infarction and Global Registry of Acute Coronary Events Scores are widely applied, whereas Medical Early Warning Score and National Early Warning Score are often used in wards to identify early clinical deterioration[7-9]. An ideal prognostic scoring system should accurately distinguish patients at risk of deterioration while not giving false alarms for low-risk or well-prognosis patients[10]. Furthermore, it should be easily achievable in the context of the targeted condition. Therefore, it is important that it has factors such as being easily measurable at the bedside, inexpensive, and easy to repeat[2,3,10].

While current scoring systems are generally very useful, their effectiveness in a clinical context has some limitations. For example, comprehensive scoring systems such as SOFA or Acute Physiology and Chronic Health Evaluation II require numerous parameters, including arterial blood gases, neurological assessments, and specific laboratory evaluations[11,12]. In time-sensitive scenarios such as the critical patient in the ED or ICU, the fast and comprehensive gathering of these facts is often a substantial difficulty. Such postponements might cause high-risk patients to overlook essential treatment options[12,13]. These limitations highlight the need for alternative or complementary tools that can be rapidly assessed at the bedside without incurring significant additional costs[14,15]. Therefore, there has recently been increasing interest in indices that are easy to calculate and suitable for bedside clinical practice, as opposed to these scoring systems[16]. These scores, which are essentially predicted to describe the disease state and support physicians in prognosis by integrating information about inflammatory status, metabolic status, and nutritional status, are being studied in many fields[12,17,18]. While this evidence provides physicians with important information, it is crucial to remember that it is not a standalone decision-making system; rather, it serves as a supporting tool[19-21].

ALBUMIN AS A CENTRAL COMPONENT OF COMPOSITE BIOMARKERS

Albumin is the most abundant protein in human plasma and has traditionally been considered an indicator of nutritional status. However, albumin also serves as a dynamic biomarker with a significant role in the physiology of acute and critical illness[22]. It is a classic negative acute-phase reactant, with levels that decrease rapidly due to suppressed synthesis in conditions such as inflammation, infection, or tissue damage. Inflammation impairs endothelial permeability, subsequently causing capillary leakage, which can lead to albumin escaping into the interstitial space[18,23]. Additional causes of decreased albumin levels include fluid and blood losses associated with intravascular volume depletion, as well as dilution resulting from aggressive fluid resuscitation[24,25]. Therefore, when considering the causes of albumin decline in a critically ill patient, it should not be solely attributed to malnutrition. Consequently, hypoalbuminemia can be affected by and reflects dysfunctions in many physiological systems, including the degree of systemic inflammation, hepatic synthesis capacity, endothelial permeability, body fluid distribution, and physiological reserves[18,26].

Therefore, albumin, a key component of many physiological mechanisms, is similarly central to various clinical scoring systems. While some systems include albumin as a standalone component, others combine it with other components to provide a more comprehensive assessment. For example, albumin is combined with other markers to indicate tissue hypoperfusion, systemic inflammation, and immune nutritional status[27,28]. In this minireview, we will examine three albumin-based indices that have recently attracted significant attention and have been widely discussed: Lactate-to-albumin ratio (LAR), C-reactive protein-to-albumin ratio (CAR), and prognostic nutritional index (PNI). A schematic representation summarizing these indices, their key confounders, and areas of application is presented in Figure 1. A detailed overview of these indices, covering their calculation formulas, primary clinical indications, and key confounding factors, is provided in Table 1.

Figure 1 Biological framework of albumin-based composite biomarkers in acute and critical care. This figure illustrates how inflammation, capillary leakage, hepatic synthesis, fluid shifts, nutritional reserve, physiological stress, and organ dysfunction influence serum albumin and contribute to the interpretation of lactate-to-albumin ratio, C-reactive protein-to-albumin ratio, and prognostic nutritional index. Lactate-to-albumin ratio mainly reflects tissue hypoperfusion and metabolic stress relative to albumin-related reserve, C-reactive protein-to-albumin ratio reflects systemic inflammatory burden relative to albumin-related reserve, and prognostic nutritional index reflects immune-nutritional vulnerability through albumin and lymphocyte count. These indices may assist risk stratification for mortality, intensive care unit admission, organ failure, complications, length of stay, and treatment response, but should be interpreted in light of clinical context and major confounding factors. IL-6: Interleukin-6; TNF-α: Tumor necrosis factor-α; CRP: C-reactive protein; LAR: Lactate-to-albumin ratio; CAR: C-reactive protein-to-albumin ratio; PNI: Prognostic nutritional index; ICU: Intensive care units. The figure was conceptually developed by the authors and generated with the assistance of OpenAI ChatGPT. It was used only as an assistive tool for generating the visual components of the figure. The overall concept, scientific content, organization, structure, labels, abbreviations, and explanatory elements were developed by the authors.

Table 1 Calculation, units, clinical use, strengths/Limitations, main confounders and cutoff values of albumin-based indices.

Index	Formula	Unit or reporting format	Common clinical settings	Strengths and limitations	Main confounders
Lactate to albumin ratio	Serum lactate/serum albumin	Unit dependent ratio. Usually calculated using lactate in mmol/L and albumin in g/dL or g/L. The units used in calculation should be clearly reported	Sepsis and septic shock, emergency department populations, intensive care unit patients, out-of-hospital cardiac arrest, acute pancreatitis, gastrointestinal bleeding, cardiovascular emergencies, respiratory failure, trauma, and postoperative critical care	Reflects the interaction between metabolic stress, tissue hypoperfusion, lactate clearance, and albumin-related physiological reserve. However, its performance varies across disease groups, and cutoff values are not uniform	Hepatic dysfunction, renal dysfunction, impaired lactate clearance, adrenergic stress, vasopressor use, fluid resuscitation, hemodilution, albumin replacement, blood loss, and timing of measurement
C-reactive protein to albumin ratio	C-reactive protein/serum albumin	Unit dependent ratio. Usually calculated using C-reactive protein in mg/L or mg/dL and albumin in g/dL or g/L. The units used in calculation should be clearly reported	Sepsis and septic shock, acute pancreatitis, appendicitis, cholangitis, inflammatory bowel disease, postoperative infection, pneumonia, COVID-19, chronic obstructive pulmonary disease, cardiovascular diseases, trauma, geriatric emergency patients, and malignancy	Combines systemic inflammatory burden with albumin-related biological reserve. However, CRP is nonspecific, and the incremental value of CAR beyond established clinical scores may vary	Chronic inflammatory disease, malignancy, older age, autoimmune disease, sterile inflammation, recent surgery or trauma, hepatic dysfunction, hypoalbuminemia, fluid overload, hemodilution, and timing of CRP measurement
Prognostic nutritional index	10 × serum albumin (g/dL) + 0.005 × total lymphocyte count (/mm3)	Dimensionless score	Oncology, gastrointestinal surgery, perioperative care, sepsis, intensive care unit patients, acute pancreatitis, gastrointestinal bleeding, liver transplantation, pneumonia, COVID-19, cardiovascular diseases, and stroke	Reflects immune-nutritional vulnerability through albumin and lymphocyte count. However, in acute illness, PNI may reflect inflammation and disease severity rather than nutritional status alone. Thresholds may not generalize across populations	Systemic inflammation, acute physiological stress, corticosteroid exposure, infection, malignancy, chronic immune dysfunction, hepatic dysfunction, malnutrition, fluid balance, hemodilution, and baseline lymphocyte abnormalities

CRP: C-reactive protein; COVID-19: Coronavirus disease 2019; CAR: C-reactive protein-to-albumin ratio; PNI: Prognostic nutritional index.

LITERATURE IDENTIFICATION AND REVIEW SCOPE

This article is structured as a focused mini-review with a narrative synthesis of the available literature. Relevant studies were identified through targeted searches of PubMed, Web of Science, and Google Scholar. Search terms included combinations of “lactate-to-albumin ratio”, “C-reactive protein-to-albumin ratio”, “prognostic nutritional index”, “albumin”, “critical care”, “intensive care”, “emergency department”, “sepsis”, “acute pancreatitis”, “gastrointestinal bleeding”, “cardiovascular disease”, “respiratory failure”, “trauma”, “surgery”, and “malignancy”. Priority was given to systematic reviews, meta-analyses, large cohort studies, and clinically relevant studies involving acute or critical care populations. This minireview was not intended as a systematic review or meta-analysis, and no formal PRISMA-based search strategy or quantitative synthesis was performed.

LAR: HYPOPERFUSION, METABOLIC STRESS, AND ALBUMIN RESERVE

When body tissues do not receive sufficient blood flow and oxygen, anaerobic metabolic activity increases, resulting in elevated lactate levels in the blood. Therefore, in critical conditions where reduced oxygen delivery is prominent, such as shock, lactate becomes an important parameter in monitoring these patients[21,29]. However, hypoperfusion is not the only cause of lactate fluctuation[30]. Metabolic conditions such as hepatic insufficiency, metformin-related lactate accumulation, adrenergic stress, impaired excretion, or increased glycolytic activity are also associated with elevated lactate[13,31]. Therefore, lactate should be integrated into the clinical context when evaluating the patient and should not be used alone.

It has been suggested that lactate be integrated with other parameters to increase its predictive power in these conditions. In clinical practice, its integration with albumin is frequently performed[16]. In this composite index, obtained by dividing lactate level by albumin, lactate indicates metabolic stress and hypoperfusion, while albumin provides information about complex components such as systemic inflammation, liver synthesis capacity, and nutritional status[30]. Thus, the two values together provide many insights into the patient’s physiological condition, biological reserve, and inflammatory load[32]. One of the most important advantages of this index is that it is inexpensive and extremely easy to monitor because it is obtained from routine blood tests[33]. Early LAR measurements in critically ill patients can help plan the patient’s individual treatment by showing the patient’s prognosis, and repeat values evaluated after treatment can show valuable information such as the response to treatment and improvements in perfusion[13]. For this purpose, LAR has been used in many critical conditions.

In critically ill ICU patients, where lactate interpretation may be affected by organ dysfunction, altered clearance, and ongoing resuscitation, LAR has been investigated as a prognostic marker that may provide additional information beyond lactate alone[34,35]. Gharipour et al[34] evaluated 6414 ICU patients using the MIMIC-III database. LAR demonstrated moderate performance in predicting 28-day ICU mortality [area under the receiver operating characteristics curve (AUC): 0.69, threshold: 1.01] and outperformed lactate alone (AUC: 0.67). The study further indicated that LAR is a more reliable prognostic tool than lactate in patients with liver dysfunction (AUC: 0.72) and kidney dysfunction (AUC: 0.70), likely due to impaired lactate clearance.

Kokulu and Sert[32] investigated LAR as a predictor of hospital discharge in 235 out-of-hospital cardiac arrest cases. LAR demonstrated superior predictive value in predicting survival (AUC: 0.823, threshold: 1.62) compared to lactate alone (AUC: 0.762) and albumin (AUC: 0.722), and that a LAR below 1.62 is an independent predictor of survival to hospital discharge [odds ratio (OR): 2.55]. In a multicenter national registry study conducted in Japan by Nishimura et al[36], 1270 out-of-hospital cardiac arrest patients who underwent targeted temperature management were investigated. LAR values at ED presentation were categorized into quartiles, and patients in the lowest quartile exhibited significantly higher rates of 30-day favorable neurological survival than those in higher quartiles (adjusted OR values for the second, third, and fourth quartiles were 0.60, 0.24, and 0.12, respectively).

The prognostic value of LAR has been evaluated in large cohorts of sepsis patients. Yoo et al[12] analyzed 3499 sepsis patients in the ED and found that LAR outperformed SOFA and quick SOFA scores in predicting 28-day mortality (AUC: 0.715), and that an LAR value > 1.52 increased the risk of death by 3.75-fold (adjusted OR). In a separate ICU study, Shadvar et al[37] reported that LAR measurements at the 6^th hour of resuscitation in 176 patients were more effective in predicting 28-day mortality (AUC: 0.917), renal replacement therapy (AUC: 0.703), and duration of mechanical ventilation (AUC: 0.714) than other parameters such as lactate clearance. Additionally, Turcato et al[13], using decision tree analyses, confirmed that LAR is an independent risk factor for 30-day mortality in sepsis patients in the ED.

The relationship between LAR and gastrointestinal disorders has been extensively researched in the literature. Liu et al[38] analyzed 539 patients with acute pancreatitis using the MIMIC-IV dataset and reported an AUC of 0.742 for 28-day mortality prediction at a threshold of 1.11, showing an advantage over lactate and albumin. Ağaçkıran and Ağaçkıran[15] evaluated LAR in patients with upper gastrointestinal bleeding and found that it predicted mortality with considerable accuracy (AUC: 0.858) and the need for ICU admission with fair accuracy (AUC: 0.789). LAR also demonstrated statistical advantage over blood urea nitrogen/albumin ratio, a widely used scoring system.

Various studies demonstrate LAR’s predictive role in cardiovascular events. A recent meta-analysis of 8408 cases of acute myocardial infarction by Ul Islam et al[21] confirmed that high LAR values more than doubled the risk of all-cause mortality. In a separate study involving 2816 patients, Chen et al[30] found that LAR demonstrated good predictive ability for in-hospital mortality (AUC: 0.734). Liu et al[33] investigated patients with cardiogenic shock and identified a strong association between LAR and 90-day mortality (hazard ratio: 4.50, AUC: 0.781). Xie et al[39] studied 873 ICU patients diagnosed with cerebrovascular disease and found that an increase in LAR was associated with a 15% increase in mortality. Although the overall predictive performance was moderate (AUC: 0.68), LAR outperformed other parameters, such as the neutrophil-to-lymphocyte ratio and the platelet-to-lymphocyte ratio.

In a study of 769 patients with acute respiratory distress syndrome, Wang et al[31] reported that mortality increased by 11% when LAR exceeded 0.90. The predictive accuracy of LAR (AUC: 0.703) was higher than that of lactate alone (AUC: 0.680). Park et al[14] analyzed 5304 cases of severe trauma and demonstrated that LAR predicted in-hospital mortality with fair accuracy (AUC: 0.740) and the need for massive transfusion with moderate accuracy (AUC: 0.84). In this study, LAR showed better results than the injury severity score and the shock index. In an ICU study of patients undergoing surgery for gastric cancer, LAR demonstrated modest predictive ability for 28-day postoperative mortality (AUC: 0.839, cutoff value: 0.82) and outperformed lactate and albumin alone. LAR also correlated with ICU assessment scores[40].

Overall, when we consider these studies conducted in various acute and critical care settings, higher LAR levels have been associated with adverse outcomes. This correlation appears to be physiologically and biologically valid because LAR predicts poorly performing patients by integrating lactate-related metabolic stress and tissue hypoperfusion with albumin-related inflammation, hepatic synthesis capacity, and physiological reserve[12,33]. However, it should be noted that this relationship varies depending on the patient and disease. For example, in sepsis and shock, trauma, or surgical cases, blood loss, fluid resuscitation, and hemodilution can significantly affect albumin concentrations and make the interpretation of results difficult[12]. In the context of gastrointestinal bleeding, cardiac arrest, and cardiovascular diseases, the timing of lactate and albumin assessments can also affect prognostic validity[15,37]. Consequently, in such critical illnesses, a LAR measurement alone may not adequately reflect dynamic metabolic and inflammatory changes, and follow-up may be necessary. Furthermore, significant heterogeneity in the methodologies of studies (patient inclusion criteria, frequency of the first event, timing of sample collection, units used in laboratory tests, definitions of outcomes, and suggested threshold values) makes comparison between different studies difficult. Therefore, if clinicians are going to use LAR when evaluating patients, it is imperative that they exercise caution in its application, acknowledging that there is no universal threshold value for this variable and interpreting LAR not as a standalone decision-making tool, but as a complementary prognostic indicator.

CAR: SYSTEMIC INFLAMMATION AND NEGATIVE ACUTE-PHASE RESPONSE

C-reactive protein (CRP) is a well established acute phase reactant that is synthesized mainly in the liver in response to infection, trauma, tissue injury, and inflammation[41]. Its serum concentration may rise within hours, which makes it a useful and widely available marker for assessing inflammatory activity in acute clinical settings[42]. In everyday practice, CRP is commonly used to support the diagnosis of infection, estimate disease severity, and follow the response to treatment[12,43]. Its low cost, relatively short half-life, and broad availability make it one of the most frequently requested laboratory tests in EDs and ICUs[44].

Despite these advantages, CRP represents only one part of the host response. It reflects the intensity of inflammation but provides little information about physiological reserve, nutritional status, hepatic synthetic function[41]. For this reason, the CRP to albumin ratio has gained attention as a simple composite index that combines a positive acute phase reactant with albumin, a negative acute phase reactant. By integrating inflammatory burden with albumin related biological reserve, CAR may offer a broader view of disease severity and has been studied as a prognostic marker in several inflammatory and critical illness conditions[45].

The conditions in which CAR is most commonly used are sepsis and septic shock, where inflammation plays a central role. In a meta-analysis by Liu et al[42] involving 3224 patients, high CAR levels were shown to be effective in predicting mortality (AUC: 0.82) and were proven to be directly associated with a poor prognosis. In a large cohort study of sepsis patients in the ED, Yoo et al[12] demonstrated that while CAR levels are a parameter influencing mortality, their effect on prognostic parameters indicating 28-day mortality was lower compared to other parameters (AUC: 0.585). Furthermore, in their study, Ranzani et al[46] demonstrated that CAR levels calculated prior to discharge serve as an important indicator of persistent inflammation even in septic patients who survive intensive care, and that they increase the risk of 90-day long-term mortality (OR 2.18).

Another important group of diseases where CAR is used is gastrointestinal diseases. For example, in patients presenting with suspected appendicitis, CAR achieved higher diagnostic accuracy (AUC: 0.853) than traditional markers[47]. Similarly, in cases of acute cholangitis, it emerges as a parameter predicting poor outcomes and ICU admission[45]. In cases of acute pancreatitis, elevated CAR levels at presentation (> 16.28) were found to increase the risk of 28-day mortality by 19.3-fold, serving as a considerable prognostic tool (AUC: 0.835) directly correlated with the Ranson and Atlanta scores[48]. In another study, CAR dynamics on the 7^th day of treatment in severe ulcerative colitis flare-ups demonstrated good prognostic performance (AUC: 0.713) in predicting treatment response. In patients undergoing emergency major abdominal surgery, CAR values measured at 24 and 48 hours postoperatively were shown to be a valuable and rapid indicator of surgical site infection development[49].

In respiratory system diseases, studies have primarily focused on infectious conditions, such as pneumonia, and inflammatory diseases, such as asthma and chronic obstructive pulmonary disease. In their meta-analysis of coronavirus disease 2019 (COVID-19) patients, Rathore et al[50] demonstrated that high CAR levels at the time of presentation are a strong predictor of severe COVID-19 disease progression (AUC: 0.81) and mortality (AUC: 0.81). In the NHANES study, You et al[51] examined patients with asthma and found that CAR increased all-cause mortality by 2.5-fold. In another study of chronic obstructive pulmonary disease patients, high CAR values were shown to be an independent parameter increasing mortality and length of hospital stay[22].

Cardiovascular diseases represent another specific area where CAR has been studied in the literature. In a study by Karabağ et al[52] evaluating patients with angina pectoris, high CAR levels successfully identified high-risk patients. Çağdaş et al[23] examined patients with acute coronary syndrome and found that elevated CAR levels indicated the high-risk group. In Çınar et al's study[20], an elevated CAR level (> 7.38) in ST-elevation myocardial infarction cases demonstrated high predictive performance (AUC: 0.787) for in-hospital mortality. The meta-analysis conducted by de Liyis et al[53], which included over 5000 patients, found that elevated CAR levels were independent factors associated with poor neurological outcomes (OR: 2.7) and increased mortality (OR: 1.71).

In trauma patients, CAR serves as a significant parameter for identifying critically ill individuals by reflecting severe tissue damage. Salem et al[54] demonstrated that the progressive rise in CAR in critically ill polytrauma patients in the ICU exhibited excellent performance in predicting the development of systemic inflammatory response syndrome (AUC: 0.989) and sepsis (AUC: 0.934). In another study, patients with traumatic brain injury were examined, and elevated CAR levels were shown to indicate a severe neuroinflammatory state, serving as an independent factor for poor neurological outcomes and mortality[55].

Since physiological reserves are limited in geriatric populations, the effects of the inflammatory response and metabolic stress become even more pronounced. Ayrancı et al[56] found that CAR was a fair predictor of mortality (AUC: 0.723) in geriatric patients presenting to the ED. In another study involving 3206 patients, it was shown that a 1-unit increase in CAR approximately tripled short-term mortality, independent of comorbidities[57].

In oncology patients, CAR is considered a strong indicator because it reflects the combined presence of disease-related inflammation and malnutrition[58]. For example, in gastrointestinal malignancies, elevated CAR levels prior to treatment have been shown to significantly reduce overall survival and disease-free survival[44,58].

Although CAR is supported by evidence from several acute and critical care populations, its interpretation has important limitations. CRP is a nonspecific inflammatory marker and may increase not only in infection or sepsis, but also in sterile inflammatory conditions such as trauma, surgery, acute cardiovascular events, autoimmune diseases, and malignancy[41,59]. Therefore, an elevated CAR should not be interpreted as a direct marker of infection, but rather as a composite signal of inflammatory burden and albumin-related physiological reserve[46,49]. Baseline CRP elevation in older adults, patients with chronic inflammatory disorders, and patients with cancer may also influence CAR independently of the acute clinical event. Similarly, low albumin may reflect chronic malnutrition, hepatic dysfunction, inflammation-related capillary leakage, fluid overload, or resuscitation-related hemodilution rather than acute disease severity alone[33,60]. Timing of measurement is also critical because CRP kinetics differ substantially from lactate kinetics[50]. Unlike lactate, which may change rapidly in response to hypoperfusion and resuscitation, CRP usually rises over several hours and may remain elevated after the initial inflammatory trigger has changed[33,41,50]. Therefore, a single CAR value at presentation may not fully reflect the patient’s inflammatory trajectory, and serial measurements may be more informative in selected clinical settings.

In summary, the CAR is a practical and biologically meaningful marker, particularly in clinical conditions where systemic inflammation plays a central role. By combining a positive acute phase reactant with albumin, it may provide a broader estimate of inflammatory burden, biological reserve, and disease severity than either parameter alone. However, its interpretation should remain context dependent. Reported cutoff values vary substantially across disease groups, study populations, timing of measurement, and clinical outcomes, and much of the available evidence is based on retrospective data. Although several studies suggest favorable prognostic performance, the incremental value of CAR beyond established clinical scores or conventional inflammatory markers remains uncertain in some populations. Therefore, CAR should not be used as a stand-alone decision tool. Further prospective and longitudinal studies are needed to clarify its optimal thresholds, dynamic changes, and incremental value when used alongside established clinical scoring systems.

PNI: IMMUNE-NUTRITIONAL RESERVE AND CLINICAL VULNERABILITY

The PNI was initially developed by Onodera et al[61] to assess the preoperative nutritional status and surgical risk of malnourished patients undergoing gastrointestinal surgery, but it has since become a widely used tool in the management of acute and critical illnesses[62]. The PNI is easily calculated using serum albumin and total lymphocyte count, and its formula is as follows: 10 × serum albumin (g/dL) + 0.005 × total lymphocyte count (/mm³). In this way, PNI combines information on nutritional status, mainly reflected by albumin, with immune status, reflected by circulating lymphocyte count[24].

In clinical practice, clearer thresholds have been defined for PNI when performing risk stratification. In general, a PNI value of 50 or higher is considered normal, whereas values below 50 suggest mild malnutrition, values below 45 indicate moderate to severe malnutrition, and values below 40 are usually interpreted as severe malnutrition. Consequently, low values are associated with a poor prognosis characterized by increased complications and mortality[24].

The greatest advantage of PNI is that it provides a rapid, inexpensive, and reproducible risk screening at the bedside without the need for complex nutritional questionnaires, time-consuming tests, or subjective assessments[25]. Since malnutrition and immunosuppression due to decreased lymphocyte counts in critically ill patients directly affect susceptibility to infections, organ failure, delayed wound healing, and length of hospital stay, the use of PNI provides clinicians with the opportunity for early and targeted nutritional/immunological intervention[24]. Although PNI was first used as a surgical risk tool, it is now increasingly applied across a broad range of clinical settings, including oncology, cardiovascular emergencies, sepsis, intensive care, and perioperative medicine, as a simple adjunct for early risk assessment and prognostic evaluation[63].

The most common field in which PNI is used is oncology. The combination of severe inflammation and malnutrition directly affects cancer patient’s survival and treatment adherence[24]. In a systematic review of 3414 cancer patients, Sun et al[64] found that low PNI was associated with poor overall survival (OR: 1.80) and postoperative complications (OR: 2.45).

Zhou et al[25] examined PNI in gastric cancer patients and demonstrated that patients with poor nutritional status had reduced treatment adherence and a poorer prognosis with a higher incidence of toxicities. In a systematic review by Fiflis et al[18] examining 14551 cases of gastroesophageal adenocarcinoma, 5-year survival rates ranged from 39% to 70.6% in the low PNI group, compared to 54.9% to 95.8% in the high PNI group. In a study conducted by Lu et al[27] in patients with pancreatic cancer, PNI values were observed to be an independent predictor of the response to programmed cell death inhibitors and overall survival. In a study by Yang et al[65] involving patients with prostate cancer, biological recurrence was more common in patients with low PNI levels (AUC: 0.789), and postoperative complications were also more frequent in the low PNI group (21.1% vs 38.9%).

In the context of sepsis, an overproduction of pro-inflammatory cytokines precipitates pronounced catabolic processes and a state of protein-energy malnutrition[60]. Research focusing on geriatric sepsis patients aged 65 years and older has revealed a nonlinear (U-shaped) correlation between PNI and 28-day overall mortality; it has been established that both markedly low and exceedingly high PNI scores serve as independent predictors of mortality[66]. Shi et al[28] found that PNI values measured at the 48^th hour of hospitalization emerged as a strong (AUC = 0.871) prognostic marker for predicting the development of severe acute pancreatitis. Cumaoglu et al[47] evaluated PNI in patients with gastrointestinal bleeding and demonstrated an association between low PNI and mortality (OR: 3.0). Li et al[26] examined 30-day mortality and complications in liver transplant patients. While PNI predicted mortality with good accuracy (AUC: 0.781), it performed poorly (AUC: 0.615) in predicting serious complications.

In a study of 488 patients with severe pneumonia, Qiu et al[67] reported that PNI values were significantly higher in survivors than in non-survivors. Moreover, each 1-point increase in PNI was associated with an approximately 7% lower risk of mortality. Similarly, Hu et al[63] reported that higher PNI scores in patients with COVID-19 were associated with a roughly 20% reduction in the risk of developing severe disease.

PNI has also been studied in cardiovascular diseases. In a study by Xu et al[68] involving 734 patients with heart failure, PNI demonstrated poor performance (AUC: 0.654) in predicting major adverse cardiac events. Fındık and Kıyak[69] studied 151 patients with acute ischemic stroke and demonstrated a modest performance (AUC: 0.793) in predicting 90-day mortality. In patients studied by Karagoz et al[70] in the ICU, each 1-point increase in PNI was associated with a 6% reduction in mortality.

Although PNI is widely used as a marker of immune nutritional status, its components may also be influenced by acute illness and systemic inflammation[68,71]. Albumin reflects not only nutritional reserve but also inflammatory activity, hepatic synthetic function, capillary permeability, fluid balance, and hemodilution[60,67,72]. Likewise, lymphocyte counts can vary in response to acute physiological stress, corticosteroid exposure, infection, trauma, malignancy, and underlying immune status[16,66]. Consequently, in acute and critical care settings, PNI may capture a combination of nutritional condition, inflammatory burden, and disease severity[68,70]. This multifactorial nature may in fact contribute to its prognostic utility, although it can make interpretation more complex in certain clinical contexts[71]. In addition, the established threshold values have largely been determined in specific surgical and oncological situations. Different threshold values may be more useful in emergency, intensive care, infection, cardiovascular, or geriatric settings[60]. Although we know that low PNI is associated with adverse risk outcomes, studies on PNI-guided nutrition and treatment guidance are insufficient, and studies are needed to understand whether this improves clinical outcomes[66]. Therefore, future studies could examine threshold values for specific populations and investigate the impact of these values on clinical decision-making processes such as guiding treatment.

When we evaluate PNI, which has been studied in many diseases in acute and critical conditions, in light of these studies, it has been shown that studies provide prognostic information in many situations. This ability is achieved through malnutrition and systemic inflammation. It has found a wide field of study, especially in patients with malignancy-related conditions. Furthermore, studies in sepsis, gastrointestinal emergencies, cardiovascular diseases, and surgical patients have shown that low PNI values are associated with higher mortality and complications. However, these findings should be interpreted with caution. PNI is influenced by numerous factors beyond nutritional status, including acute inflammation, physiological stress, fluid balance, and underlying immune dysfunction. Additionally, significant heterogeneity exists between studies in terms of patient populations, disease severity, measurement timing, outcome definitions, and proposed threshold values, limiting direct comparisons and broad generalizability. Therefore, PNI should not be interpreted as a pure nutritional marker in acute illnesses, and prognostic thresholds should be applied carefully according to the clinical context. Overall, PNI can be considered a practical aid for risk stratification, but it is unlikely to be sufficient as a prognostic marker alone, and further prospective studies are needed to clarify its optimal clinical application and potential role in guiding interventions.

CONCLUSION

Current evidence suggests that LAR, CAR, and PNI are practical, reproducible, and low-cost indices that may support risk stratification in acute and critical care. By combining albumin with markers of metabolic stress, systemic inflammation, and immune nutritional reserve, these indices may provide broader prognostic information than single laboratory parameters alone. Across different clinical contexts, elevated LAR and CAR values and reduced PNI values are generally associated with adverse outcomes, including mortality, organ dysfunction, intensive care requirement, and postoperative complications. However, their predictive utility remains highly context dependent. Most available studies are retrospective and single center in design, and prospective validation studies or meta-analyses are still limited for several disease groups. In addition, differences in patient populations, baseline risk, disease prevalence, timing of measurement, resuscitation status, outcome definitions, laboratory units, and proposed cutoff values limit direct comparison across studies and make universal thresholds difficult to define. Therefore, these indices should not be used as standalone decision tools but should be interpreted together with clinical assessment and established severity scores. Future research should move beyond simple association studies and focus on prospective, multicenter validation in clearly defined acute and critical care populations. Disease specific cutoff values should be standardized according to clinical context, measurement timing, laboratory units, and outcome definitions. In addition, serial measurements and dynamic biomarker trajectories may be more informative than single baseline values, particularly in patients undergoing resuscitation.