Secondary Logo

Journal Logo

Featured Articles: Editorial

A Problem of Too Much Heterogeneity

Downey, Laura A. MD; Guzzetta, Nina A. MD, FAAP

Author Information
doi: 10.1213/ANE.0000000000004719

See Article, p 1594

Systematic reviews and meta-analyses have become increasingly common in the medical literature.1 Their purpose is to integrate the results from multiple studies or data sets to answer predefined clinical questions and determine a single quantitative estimate or summary effect size.2 Systematic reviews can reduce bias and validate clinical recommendations by identifying, appraising, and synthesizing all relevant studies on a particular topic. However, given the numerous studies included in most systematic reviews, some degree of heterogeneity is unavoidable. It may be due to differences in study participants, interventions, measurements, and/or other factors that vary across data sets. Alternatively, heterogeneity may be embedded within the primary studies included in the meta-analysis or from techniques used in the meta-analysis to assess data from the collected studies. Heterogeneity is not necessarily a bad thing and is often adjusted for during the meta-analysis. In a large data set, modest heterogeneity may be welcome to support a wider generalization of the results. However, in a small data set, too much heterogeneity can compromise the internal validity of the study and therein lies the major issue uncovered in this month’s article by Bianchi et al.3

Several systematic reviews and meta-analyses have been conducted in adult cardiac surgical patients to evaluate the use of point-of-care tests (POCTs), particularly the whole blood viscoelastic tests, thromboelastography (TEG), or rotational thromboelastometry (ROTEM).4–7 All of these analyses contain a substantial number of randomized controlled trials (RCTs) with only a few prospective observational or retrospective trials. Three contain over 8000 adult cardiac patients while 1 has over 15,000. The patients included in these reviews represent a relatively homogeneous sample, adults undergoing coronary artery bypass grafting and/or valve surgery. Only one of the reviews performs a subgroup analysis based on the surgical procedure. The 4 meta-analyses come to a similar conclusion: the routine use of viscoelastic testing reduced the number of patients receiving a transfusion and the number of blood products ultimately transfused. Two of the analyses also report a reduction in postoperative acute kidney injury and 1 suggests a reduction in postoperative thromboembolic events when viscoelastic tests are used to guide transfusions in adults after cardiopulmonary bypass (CPB). Of note, the use of these tests did not affect other patient outcomes such as a reduction in ventilation time, intensive care unit (ICU), or hospital length of stay or mortality.

While the utilization of POCTs in adult cardiac surgery has had a significant impact on bleeding and transfusion metrics, the complexity of perioperative anticoagulation and coagulation management in pediatric cardiac surgery has made it more difficult to make a comparable assessment. We know that bleeding after CPB is a serious complication of pediatric cardiac surgery and associated with substantial morbidity and mortality. Multiple studies have confirmed this finding.8–10 Neonates and young infants in particular undergo long, complex congenital cardiac repairs with extensive suture lines and are vulnerable to post-CPB bleeding. Other contributing factors include the significant hemodilution that occurs during CPB in small patients, activation and consumption of already low levels of coagulation factors, baseline and post-CPB platelet dysfunction, and preoperative cyanosis. Mitigating bleeding is currently addressed through transfusion of blood products obtained from adults even though recent evidence suggests that neonatal and adult fibrin networks differ substantially and may not integrate seamlessly into an organized clot structure.11 In addition, pediatric patients often required multiple transfusions to adequately restore hemostasis and these transfusions are also associated with significant risk.8 Although it seems reasonable that pediatric POCT-based algorithms would also aid clinicians in providing specific blood component therapy to reduce bleeding and complications related to transfusions, this has not yet been demonstrated.

The perioperative hemostatic management of pediatric cardiac patients differs from that in adults. The hemostatic system in neonates is dynamic and rapidly evolves over the first 2 years of life.12 Age-specific differences exist in the concentration of pro- and anticoagulation factors, and in thrombin generation and inhibition. The presence of a qualitatively dysfunctional, or “fetal,” form of fibrinogen has also recently been confirmed and exists up to approximately 1 year of age.11 Due to these quantitative and qualitative differences, the established range of “normal” adult values for both conventional coagulation tests and POCTs may not accurately reflect the function of the developing hematologic system. In fact, current data from both TEG and ROTEM suggest that infants <3 months of age display a faster initiation and propagation of clot formation and greater clot strength when compared to older patients.13,14 As a clinician, understanding how each POCT reflects the dynamic changes within the hemostatic system throughout the course of cardiac surgery is essential to optimizing blood component therapy in the pediatric population.

In this month’s issue of Anesthesia & Analgesia, Bianchi et al3 publish a systematic literature review of POCTs in pediatric cardiac surgery patients. The authors conduct a search using 3 databases (Pubmed/Medline, Embase, and the Cochrane Controlled Clinical Trials registry) to identify studies in which POCTs were used to assess some aspect of the coagulation profile in pediatric patients (age <18 years old) undergoing cardiac surgery. They cast a wide net to include 3 main categories of POCTs: activated clotting time (ACT)–based instruments (10 studies); viscoelastic tests (32 studies, including 1 from the previous grouping); and platelet function tests (6 studies). The authors identify 80 articles of which 47 were deemed appropriate for review. Bianchi et al3 should be congratulated on their effort to review the current literature on this important topic. However, due to heterogeneity of both the test being assessed and the patient population, the authors were unable to perform a meta-analysis or make firm recommendations regarding the utility of POCTs for pediatric cardiac patients. This is unfortunate. POCTs appear to be an attractive option for the timely evaluation of hemostasis in children undergoing CPB and thus would hopefully be studied in an organized and methodical manner to best determine their utility. Those of us who perform research focused on improving hemostatic processes in these patients should perhaps shoulder some of the responsibility for the quality, or lack thereof, of research in this arena.

Several obstacles have impeded progress in the development of POCT-based algorithms for pediatric cardiac surgery patients. First, paramount to utilizing POCTs in the pediatric arena is the determination of appropriate age-specific reference ranges. The field will need to establish not 1 but several ranges of normal values in the postnatal period. The Clinical and Laboratory Standards Institute recommends a minimum of 120 individuals to establish a reference interval,14 and most pediatric studies do not come close to this number. Second, single institutional studies are limited by the number of patients presenting to the institution. As a result, pediatric studies often include a wide range of ages (0–18 years), a mixture of cyanotic and acyanotic patients, as well as heterogeneity in the complexity of the surgical procedure. This level of heterogeneity leads to small data sets and difficultly establishing appropriate age- or disease-specific reference ranges or recommendations. Third, standard coagulation tests require relatively large amounts of blood, are time consuming, need to be performed in a central laboratory, and are designed to assess adult coagulation processes. Finally, many commercially available tests and devices are considered “off-label” for use in the pediatric population. Performing well-designed RCTs in neonates and infants is challenging due to important safety concerns regarding the need for institutional and federal approval to use “off-label” devices and medications in this vulnerable population.

As researchers, how can we address the above barriers to move our field forward? To start, POCTs do not account for the changing levels of pro- and anticoagulants in neonates and young infants. Studies are needed to gain a better understanding of these developmental changes and how to integrate them into current POCTs. In addition, we could better utilize large databases, such as the Congenital Cardiac Anesthesia Society-Society of Thoracic Surgeons Congenital Heart Surgery Database (CCAS-STSCHSD), to obtain baseline data. The quality and quantity of data gathered from multiple institutions would allow for larger, more robust studies to define age-specific reference ranges and support both retrospective and longitudinal studies in neonates and infants undergoing cardiac surgery. Next, translational studies focusing on how neonatal fibrinogen and platelets interact with their adult counterparts are important to define target thresholds for POCT-based transfusion algorithms. Only then will we be able to define the role that developmental hemostasis plays in the clinical arena. Finally, we need well-designed randomized clinical trials to assess outcomes when POCTs are used in specified groups of pediatric patients.

While the advantages of POCTs for pediatric cardiac patients may seem obvious, the literature to support their use in children is not robust. Bianchi et al3 conclude that the evidence is “too weak to define point-of-care tests as the gold standard to treat perioperative bleeding.”3 Age-specific differences in the hemostatic system, limited patient numbers, and a paucity of high-quality studies highlight some of the challenges faced in validating POCT-based algorithms in pediatric patients. Therefore, the importance of this systematic review lies not in the answers it provides, but in recognition of the challenges posed by this field of study. For researchers who focus on improving hemostatic therapies for pediatric cardiac patients, this is a call for multi-institutional collaboration and the development of well-designed studies to address the important questions.


Name: Laura A. Downey, MD.

Contribution: This author helped in the literature search, written content, and editing of the manuscript.

Name: Nina A. Guzzetta, MD, FAAP.

Contribution: This author helped in the literature search, written content, and editing of the manuscript.

This manuscript was handled by: Roman M. Sniecinski, MD.



    1. Ioannidis JP. Interpretation of tests of heterogeneity and bias in meta-analysis. J Eval Clin Pract. 2008;14:951–957.
    2. Sacks HS, Berrier J, Reitman D, Ancona-Berk VA, Chalmers TC. Meta-analyses of randomized controlled trials. N Engl J Med. 1987;316:450–455.
    3. Bianchi P, Beccaris C, Norbert M, Dunlop B, Ranucci M. Use of coagulation point-of-care tests in the management of anticoagulation and bleeding in pediatric cardiac surgery: a systematic review. Anesth Analg. 2020;130:1594–1604.
    4. Deppe A-C, Weber C, Zimmermann J, et al. Point-of-care thrombolastography/thromboelastometry-based coagulation management in cardiac surgery: a meta-analysis of 8332 patients. J Surg Res. 2016;203:424–433.
    5. Serraino GF, Murphy GJ. Routine use of viscoelastic blood tests for diagnosis and treatment of coagulopathic bleeding in cardiac surgery: updated systematic review and meta-analysis. Br J Anaesth. 2017;118:823–833.
    6. Lodewyks C, Heinrichs J, Grocott HP, et al. Point-of-care viscoelastic testing in cardiac surgery patients: a systematic review and meta-analysis. Can J Anesth. 2018; 65:1333–1347.
    7. Li C, Zhao Q, Yang K, Jiang L, Yu J. Thromboelastography or rotational thromboelastometry for bleeding management in adults undergoing cardiac surgery: a systematic review with meta-analysis and trial sequential analysis. J Thorac Dis. 2019;11:1170–1181.
    8. Iyengar A, Scipione CN, Sheth P, et al. Association of complications with blood transfusions in pediatric cardiac surgery patients. Ann Thorac Surg. 2013;96:910–916.
    9. Wolf MJ, Maher KO, Kanter KR, Kogon BE, Guzzetta NA, Mahle WT. Early postoperative bleeding is independently associated with increased surgical mortality in infants after cardiopulmonary bypass. J Thorac Cardiovasc Surg. 2014;148:631.e1–636.e1.
    10. Guzzetta NA, Allen NN, Wilson EC, Foster GS, Ehrlich AC, Miller BE. Excessive postoperative bleeding and outcomes in neonates undergoing cardiopulmonary bypass. Anesth Analg. 2015;120:405–410.
    11. Brown AC, Hannan RT, Timmins LH, Fernandez JD, Barker TH, Guzzetta NA. Fibrin network changes in neonates after cardiopulmonary bypass. Anesthesiology. 2016;124:1021–1031.
    12. Andrew M, Paes B, Johnston M. Development of the hemostatic system in the neonate and young infant. Am J Pediatr Hematol Oncol. 1990;12:95–104.
    13. Miller BE, Guzzetta NA, Tosone SR, et al. Tissue factor-activated thromboelastograms in children undergoing cardiac surgery: baseline values and comparisons. Anesth Analg. 2003;97:1289–1293.
    14. Kim JY, Shin YR, Kil HK, Park MR, Lee JW. Reference intervals of thromboelastometric evaluation of coagulation in pediatric patients with congenital heart diseases: a retrospective investigation. Med Sci Monit. 2016;22:3576–3587.
    Copyright © 2020 International Anesthesia Research Society