Association Between Anemia and ICU Outcomes
Anemia is a prevalent condition among critically ill patients admitted to intensive care units (ICUs), with nearly 60% of ICU patients presenting hemoglobin levels below 12 g/dL at baseline. Within three days of ICU admission, over 95% of patients develop anemia, either as a pre-existing condition or as a consequence of critical illness. This high prevalence raises significant clinical concerns, as anemia can impair oxygen delivery to vital organs, potentially exacerbating organ dysfunction and influencing patient outcomes. Despite its frequency, the relationship between anemia and ICU outcomes remains debated, with conflicting evidence across different patient subgroups. This meta-analysis aimed to clarify the association between anemia and clinical outcomes in critically ill patients, exploring variations based on primary diagnoses and the timing of anemia development.
Prevalence and Pathophysiological Implications of Anemia in the ICU
Anemia in critically ill patients arises from multiple factors, including chronic diseases, acute blood loss, inflammation, frequent phlebotomy, and impaired erythropoiesis. The study highlighted that 30% of ICU patients had severe anemia (hemoglobin <9 g/dL) at baseline, with nearly all patients becoming anemic during their ICU stay. Severe anemia reduces oxygen-carrying capacity, leading to tissue hypoxia, which may compromise cardiac, renal, and neurological function. For instance, renal tubular cells, heavily reliant on oxidative metabolism, are particularly vulnerable to hypoxic injury during anemia, potentially accelerating acute kidney injury (AKI). These pathophysiological mechanisms underscore the importance of understanding anemia’s role in ICU outcomes.
Study Design and Methodology
The meta-analysis included 28 observational studies (20 cohort, 8 case-control) involving 28,285 critically ill adults. Studies were identified through systematic searches of PubMed, Web of Science, and EMBASE up to September 2020. Anemia was defined as hemoglobin <13 g/dL for men and <12 g/dL for women, either at ICU admission or during the stay. Outcomes assessed included mortality (ICU, hospital, 30-day, 90-day, and 6-year), length of stay (ICU and hospital), and complications such as AKI. Subgroup analyses were conducted based on primary diagnoses, including sepsis, trauma, cancer, cardiac conditions, and chronic obstructive pulmonary disease (COPD).
Key Findings on Mortality
All-Cause Mortality
Pooled data from 17 studies (15,499 participants) revealed a significant association between anemia and increased all-cause mortality in unadjusted analyses (odds ratio [OR]: 2.57, 95% confidence interval [CI]: 1.94–3.40). However, after adjusting for confounders such as age, comorbidities, and illness severity in four studies, this association attenuated (adjusted OR: 1.36, 95% CI: 0.73–2.52), suggesting that anemia’s impact may be mediated by underlying conditions.
Subgroup analyses demonstrated heterogeneity in outcomes:
- High-Risk Subgroups: Patients with sepsis, trauma, cancer, or AKI exhibited significantly elevated mortality risks (OR: 2.57–3.10). For example, in sepsis, anemia was linked to a 2.6-fold higher mortality risk.
- Neutral or Protective Subgroups: No significant association was found in patients with traumatic brain injury (TBI) or COPD. This divergence highlights the influence of primary disease pathophysiology on anemia-related outcomes.
ICU and Hospital Mortality
Two studies (1,158 participants) showed no significant association between anemia and ICU mortality in unadjusted analyses (OR: 1.78, 95% CI: 0.61–5.18). However, a single adjusted analysis (126 participants) indicated a fourfold higher ICU mortality risk (OR: 4.06, 95% CI: 1.30–12.68). Hospital mortality, assessed in four studies (1,967 participants), was significantly higher in anemic patients (OR: 2.22, 95% CI: 1.39–3.56). Subgroup analyses revealed elevated risks in cardiac ICU patients and those with peritonitis but not in COPD patients.
Longitudinal Mortality Outcomes
Anemia’s impact persisted beyond the acute phase:
- 30-Day Mortality: Two studies reported increased risks (relative risk [RR]: 1.79–3.10).
- 90-Day Mortality: Hazard ratios (HR) ranged from 1.68 to 2.60.
- 6-Year Mortality: A single large study (2,145 participants) found a 79% higher mortality risk (OR: 1.79, 95% CI: 1.49–2.13).
Impact on Hospital Stay and Complications
Length of Stay
Two studies found no association between anemia and ICU or hospital stay duration. However, one study reported prolonged ICU stays (mean difference: 8.0 days, 95% CI: 5.93–10.07) in anemic patients, suggesting variability based on patient characteristics or ICU practices.
Acute Kidney Injury
Anemia was associated with a 76% higher risk of AKI progression (HR: 1.76, 95% CI: 1.35–2.30) in one study. However, no significant association was observed at six-month follow-up, indicating that anemia’s renal effects may be acute rather than chronic.
Limitations and Heterogeneity
The included studies exhibited high risk of bias, primarily due to non-representative samples (e.g., single-disease cohorts) and inconsistent anemia definitions. Heterogeneity (I²: 63–75%) in meta-analyses reflected variations in study design, population, and adjustment for confounders. Funnel plots for all-cause mortality suggested no publication bias, though smaller outcomes lacked formal assessment.
Clinical and Research Implications
The findings emphasize anemia as a prognostic marker in specific ICU populations. For high-risk groups like sepsis or cardiac patients, close monitoring and timely interventions (e.g., optimized transfusion strategies, iron supplementation) may improve outcomes. However, the lack of association in TBI or COPD patients suggests that anemia management should be individualized based on underlying conditions.
Future research should address gaps in understanding the temporal relationship between anemia onset and outcomes, the role of anemia duration, and the impact of therapeutic interventions. Standardized definitions of anemia and rigorous adjustment for confounding variables are essential to enhance comparability across studies.
doi.org/10.1097/CM9.0000000000001669
Was this helpful?
0 / 0