Introduction
Anemia is highly prevalent in the intensive care unit (ICU), with up to 95% of critically ill patients developing subnormal hemoglobin levels by day 3 [
1]. Likewise, 20% to 53% of patients receive red blood cell (RBC) transfusions to correct anemia during their ICU stays [
2]. However, allogenic RBC transfusions carry risks that may adversely affect clinical outcomes [
3,
4]. Evidence suggests that it is safe to adopt a lower transfusion threshold for the general medical/surgical ICU population [
1,
4‐
8]. This has led to a paradigm shift concerning RBC transfusions in the ICU, with most guidelines now recommending hemoglobin levels around 7 g/dl for transfusion in patients without significant comorbidities to minimize exposure to allogenic blood [
9‐
11].
Most studies of transfusion thresholds have focused on a general medical/surgical ICU population but not on specific, and potentially more vulnerable, subpopulations of critically ill patients, such as those with acute neurologic conditions [
12]. Indeed, neurocritically ill patients may represent an exception to the rationale for using low transfusion triggers because impaired oxygen delivery is a crucial modifiable factor in brain ischemia and secondary brain injury [
13,
14]. The optimal hemoglobin level for cerebral oxygen delivery in these patients is still unknown [
15]. Moreover, data on which clinicians have to rely in decision making is discordant, as both anemia and RBC transfusion have been observed to be associated with unfavorable clinical outcomes in neurocritically ill patients [
16‐
18].
Current guidelines for the optimal transfusion threshold in neurocritical care populations are scarce, and their recommendations are conflicting about which threshold to favor [
19,
20]. Several narrative studies have aimed to summarize the topic [
15‐
18], but no systematic review has been designed to address specifically the question of transfusion thresholds in the neurocritical population. We thus undertook a systematic review of comparative studies to evaluate the effects of hemoglobin levels and RBC transfusion strategies on clinical outcomes in adult and pediatric neurocritically ill patients.
Materials and methods
This systematic review was designed in accordance with the PRISMA statement for systematic reviews and meta-analyses [
21]. A study protocol was developed and followed through every step of the review.
Search strategy
We designed a search strategy for Ovid MEDLINE (1949 to the present), the Cochrane Central Register of Controlled Trials (1974 to Issue 1, 2011), as well as Embase and Embase Classic (1974 to the present). Abstracts and conference proceedings were searched in BIOSIS previews (1926 to the present) and Web of Science (1898 to the present), whereas the grey literature was searched by using Google Scholar. We sought both randomized controlled trials (RCTs) and comparative nonrandomized studies, both prospective or retrospective. No restriction based on language, year, or type of publication was applied. Keywords and Medical Subject Headings (MeSH) terms (or their EMTREE equivalents) pertaining to the population (neurocritical care) and to the exposure (hemoglobin levels, RBC transfusion, anemia) were combined to form the search strategy (Additional file
1). We used clinicaltrials.gov, controlled-trials.com, and strokecenter.org websites to identify unpublished and ongoing studies. Reference lists from relevant reviews and included articles were manually searched to identify missed studies. The last iteration of the search process was completed on January 31, 2011.
Selection of studies
We included comparative studies evaluating the effect of hemoglobin levels on clinical outcomes of neurocritically ill patients admitted to an ICU. Studies were included if at least two different hemoglobin thresholds, levels, targets, or RBC transfusion strategies were compared. Neurocritical conditions encompassed but were not limited to subarachnoid hemorrhage (SAH), stroke, traumatic brain injury (TBI), intracerebral hemorrhage (ICH), and any cerebral neurosurgical conditions. Studies on sickle cell anemia and scoliosis surgery were excluded. We also excluded studies in neonates (< 28 days), but all other age groups were considered.
Two independent reviewers (PD, MHT) screened the studies identified from the systematic search. Non-English language articles were translated as required. A Cohen kappa statistic was calculated to quantify the interrater agreement concerning inclusion of studies. In case of discrepancy, a third reviewer (AFT) was involved to settle the disagreement. Search results from Web of Science, from grey literature sources, and from reference lists of identified studies were reviewed and adjudicated by a single reviewer (PD).
Data-collection process
A standardized abstraction form was developed and tested before data collection. Data abstraction was conducted independently, and in duplicate, by two reviewers (PD, MHT). When judged necessary, missing information was requested from corresponding authors.
The primary outcome measure was all-cause mortality at any given time point. Secondary outcomes were neurologic status (irrespective of the scale used), ICU length of stay, hospital length of stay, duration of mechanical ventilation, surrogate measures of brain oxygen delivery, complications (including vasospasm and multiple organ dysfunction score) [
22], and serious adverse events (thromboembolic events, myocardial infarction, pulmonary edema or volume overload, transfusion-related acute lung injury (TRALI), and infection). Data pertaining to the study design were also retrieved, as well as characteristics of patients that could act as confounders and affect the outcomes of interest, including age, sex, disease severity, comorbidities, incidence of hypoxemia, incidence of hypotension, and baseline hemoglobin. Information on blood transfusion and the nature, timing, and frequency of co-interventions (hemodilution, blood-conservation strategies, erythropoietin analogues, and use of other blood products) were recorded.
Assessment of methodologic quality and risk of bias
Two reviewers (PD, MHT) independently evaluated the risk of bias in included studies. We used the Cochrane Collaboration tool for assessing risk of bias in RCTs, which was customized for the focus of the review [
23]. We judged the overall risk of bias of individual studies as low, moderate, high, or unclear [
23]. Additionally, we used the Downs and Black checklist [
24] to assess the methodologic quality of both RCTs and nonrandomized studies. This checklist has been validated for reliability and external validity. We put emphasis on how study authors dealt with confounding factors mentioned earlier in nonrandomized studies. The last item of the Downs and Black checklist is an assessment of the adequacy of the sample size of the study, which we performed assuming a two-sided
P value of 0.05, 80% power, and a 10% relative difference for the main outcome measure.
Statistical analysis
A meta-analytic approach was planned by using Mantel-Haenztel random-effect models, if deemed appropriate. We presented outcome data by using odds ratios with 95% confidence intervals. An odds ratio of less than 1 suggests a lower rate of the event among the patients exposed to lower hemoglobin levels. Continuous data, such as length of stay and physiologic parameters, were reported as mean or median. We summarized continuous data as mean difference with 95% confidence intervals. We converted hematocrit to hemoglobin by using a standard published equation [
25] (Hb [g/dl] = Hct [%]/3). All data were compiled in Review Manager (version 5.0; The Cochrane Collaboration).
A priori sensitivity analyses were planned to explore heterogeneity in study findings, based on age, type of neurocritical condition, risk of bias, and presence of co-interventions.
Discussion
In this systematic review, despite our thorough search of the literature, we identified very few comparative studies of transfusion strategies conducted in different pediatric or adult neurocritically ill populations. Insufficient data exist to refute or confirm a mortality benefit associated with the maintenance of lower or higher hemoglobin level nor to support a consistent effect on organ failure and duration of mechanical ventilation. A potential decrease in hospital and ICU length of stay was observed, but only in studies with high risk of bias. Interestingly, only two studies presented long-term functional outcomes but were not designed to evaluate a plausible clinical effect. These results underscore the paucity of evidence to justify the use of a restrictive or a liberal strategy for RBC transfusions in neurocritically ill patients.
Many theoretic effects of maintaining low hemoglobin levels in neurocritically ill patients have been described in previous experimental studies. Lower hemoglobin concentration is directly related to lower blood viscosity [
37]. In mild anemia, this decrease in viscosity causes an increase in cerebral blood flow (CBF) through a direct rheologic effect and improves cerebral oxygen delivery (DO
2) [
38]. However, more-severe anemia may be detrimental in neurocritically ill patients because the decline in CaO
2 may not be compensated by the usual CBF regulation mechanisms, which are mitigated in brain injury. On clinical grounds, anemia has repeatedly been shown to be associated with unfavorable outcomes in patients with TBI [
39,
40], although other studies have not confirmed this relation [
41,
42]. In patients with SAH, anemia has mostly been associated with unfavorable outcomes [
43‐
45]. Recent microdialysis studies showed that cerebral metabolism becomes impaired at Hb values lower than 9 g/dl [
46,
47].
RBC transfusions are known to improve physiologic measures such as brain oxygen tension in a majority of patients with TBI [
31,
48‐
50]. Retrospective cohort studies in patients with SAH reported an association between the correction of anemia with RBC transfusion and unfavorable outcomes [
51,
52], more complications [
53], and vasospasm [
54]. Both anemia and RBC transfusion have thus been associated with worse clinical outcomes in different neurocritically ill patients.
Interestingly, we did not observe similar findings in our study. This is likely to be explained by the fact that we studied the impact of the exposure to Hb levels and transfusion strategies on outcomes, unlike most previous studies, which evaluated the impact of RBC transfusions (as a risk factor for a specific oucome measure rather than an intervention), regardless of the hemoglobin thresholds or Hb levels. By doing so, we aimed to avoid two potential biases. The first pertains to anemia, which often occurs in sicker patients along with confounding variables such as a greater volume of sampled blood in patients with more severe diseases [
48]. Thus, it is prone to confounding despite adjustment for disease severity. The second is a potential multicolinearity bias concerning RBC transfusion and anemia. These two variables are strongly linked, given that anemic patients are predisposed to receive more RBC transfusions because of the natural tendency of physicians to give transfusions to sicker patients. To separate the respective effects of anemia and RBC transfusion, we opted to focus on differential transfusion strategies. Therefore, our approach aimed to determine a potential inflexion point (typically the mean of a hemoglobin threshold for transfusion) at which the benefits of correcting anemia surpass the detrimental effects of RBC transfusion.
One of the main limitations of our study pertains to the significant inconsistency in observed summary estimates. This may in part be due to the heterogeneity in study designs. Outcomes were assessed at various time points, and the exposure to Hb levels varied across studies. In particular, the overlaping of ranges between lower and higher Hb levels in various study groups limited direct comparison between studies. Moreover, the presence of a mandatory transfusion protocol in RCTs versus the passive observation of different hemoglobin levels in nonrandomized studies can lead to a difference in observed effects. It would be misleading to liken the data obtained from the nonrandomized studies to transfusion strategies. Accordingly, we did not pool results from RCTs and nonrandomized studies to avoid generating a more-precise but biased pooled estimate [
23]. Still, we believe the comparison within each trial between groups of higher and lower Hb levels, whether by a definite transfusion trigger or by observed exposure to different Hb levels, is valid.
Some other concerns may affect our findings. The first one is obviously the scarcity of RCTs in this neurocritically ill population, despite the large number of retrospective studies in this field. The retrieved studies are mainly in the TBI population, with only two of the six studies focusing on the non-TBI population. Therefore, we cannot extrapolate our findings to stroke and ICH.
Second, most included studies were underpowered to evaluate clinically significant outcomes, making the detection of a difference in these outcomes unlikely, if present. Wide confidence intervals around estimates also stem from small sample sizes.
Third, lack of data on many relevant outcomes, such as long-term neurologic functional status or organ dysfunction, precluded the pooling of data. Even more worrying is the lack of systematic reporting of adverse events associated with RBC transfusion, because one of the main reasons to withhold RBC transfusion is to prevent, at least theoretically, these adverse events.
Finally, the methodologic quality of three of the six included studies was not optimal, although RCTs were considered to have a low risk of bias. In particular, issues with blinding and confounding cast a shadow on the robustness of findings.
Conclusions
In our study, we could not refute or confirm a difference in mortality or long-term neurologic outcomes between RBC transfusion strategies in neurocritically ill patients. Considering the lack of evidence regarding these clinically significant outcomes and the risk of bias of studies, no recommendation can be made about which transfusion strategy to favor in neurocritically ill patients; no evidence exists that maintenance of a lower or a higher hemoglobin level is superior in this specific population. Interestingly, despite how common RBC transfusions can be in neurocritically ill patients, there is a paucity of evidence about when it is appropriate to transfuse.
Ultimately, our findings suggest that research in transfusion therapy in neurocritical care is still in its infancy. Future research on the management of anemia and RBC therapy is warranted. We believe such research should assess long-term neurologic functional status, thoroughly seek adverse events, and encompass different neurocritically ill populations, such as traumatic brain injury, subarachnoid hemorrhage, and stroke.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
PD, AFT, FL, RZ, LM, LM, SWE, and DAF contributed to the conception and design of the study. PD and MHT evaluated the eligibility of studies and extracted data. PD and AFT performed and reviewed the analyses. PD and AFT drafted the manuscript. All authors participated in the interpretation of the data, the critical review of the manuscript, and approved the final version.