Skip to main content
Critical Care logoLink to Critical Care
letter
. 2024 Nov 26;28:390. doi: 10.1186/s13054-024-05115-7

Time-dependent intervention in the database study examining the efficacy of whole blood transfusion in traumatic patients

Taisuke Shibata 1, Saburo Minami 1, Atsushi Shiraishi 1,
PMCID: PMC11590400  PMID: 39593156

Author’s response

We read with great interest the database study by Aoki et al. [1], published in Critical Care. As highlighted in this article, whole blood (WB) transfusion for patients with trauma or severe hemorrhage has recently attracted worldwide attention for its promising potential in reducing mortality [24]. This study examines the association between the whole blood rate (WBR), defined as the number of WB units divided by the sum of WB units and packed red blood cells (PRBCs), and 24-h mortality in trauma patients requiring massive transfusion. The results suggest that a higher WBR is associated with improved survival and a lower risk of acute kidney injury (AKI) in patients who received whole blood transfusion within 4 h of hospital arrival. The study utilized generalized estimating equations to adjust for covariates, including clustering, and conducted sensitivity analyses that accounted for the heterogeneity of the lowest WBR group, further enhancing the robustness of the results. Therefore, this study has significant implications for current trauma resuscitation strategies. However, we believe there are some untreated biases, particularly time-dependent intervention bias, which warrant careful consideration.

Firstly, the intervention in this study design is time-dependent; that is, the risk of mortality could continuously influence the physician's decision to administer WB transfusion, and WB transfusion, in turn, could influence the risk of mortality. The risk of mortality and transfusion decisions interact dynamically over time during early trauma care. Component transfusion is routinely administered and readily available at many trauma centers, whereas WB transfusion often depends on the institution's system or regional blood bank, typically making it available later. In many cases, component transfusion is administered first, followed by WB transfusion. Under this assumption, WBR would be zero in the early phase when only component transfusion is being used, and would gradually increase once WB transfusion begins, ranging from zero to one. If, as the authors assume, the intervention, expressed as a continuous variable, affects outcomes, then the time-dependent WBR would impact outcomes differently over time. Thus, evaluating the relationship between a time-dependent intervention like WBR and outcomes at a specific point in time may not fully capture the true relationship.

Additionally, the relationship between WBR and outcomes may introduce immortal time bias. This bias occurs in epidemiological and clinical studies when there is a period during which participants cannot experience the outcome of interest, such as death [5]. Patients with severe trauma are at high risk of early death. Assuming that WBR increases over time, patients with a lower WBR may be those who died early due to severity, while patients with a higher WBR may have survived longer.

Statistical models that account for time-dependent interventions and immortal time bias, such as Marginal Structural Models with Inverse Probability of Treatment Weighting (MSMs with IPTW), may be more appropriate [6, 7]. If the current data source does not sufficiently describe time-dependent interventions, prospective studies may be necessary to generate a more suitable dataset.

In conclusion, prospective studies or database studies that account for time-dependent interventions and biases such as immortal time bias are necessary. Future research incorporating these considerations will ensure results that are both statistically robust and clinically meaningful, ultimately contributing to improved trauma resuscitation strategies.

Acknowledgements

Not capable.

Abbreviations

WB

Whole blood

WBR

Whole blood rate

AKI

Acute kidney injury

TTE

Target trial emulation

Author contributions

All authors contributed to the manuscript conception and design. The first draft of the manuscript was written by Taisuke Shibata and all authors gave critical appraisal to the initial draft. All authors read and approved the final manuscript.

Funding

Not applicable.

Availability of data and materials

Further information on the original manuscript is available from the corresponding authors upon reasonable request. No datasets were generated or analysed during the current study.

Declarations

Ethical approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Aoki M, Abe T, Komori A, Katsura M, Matsushima K. Association between whole blood ratio and risk of mortality in massively transfused trauma patients: retrospective cohort study. Crit Care. 2024;28(1):253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Torres CM, Kent A, Scantling D, Joseph B, Haut ER, Sakran JV. Association of whole blood with survival among patients presenting with severe hemorrhage in US and Canadian adult civilian trauma centers. JAMA Surg. 2023;158(5):532–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Hazelton JP, Ssentongo AE, Oh JS, Ssentongo P, Seamon MJ, Byrne JP, et al. Use of cold-stored whole blood is associated with improved mortality in hemostatic resuscitation of major bleeding: a multicenter study. Ann Surg. 2022;276(4):579–88. [DOI] [PubMed] [Google Scholar]
  • 4.Aoki M, Katsura M, Matsushima K. Association between whole blood transfusion and mortality among injured pediatric patients. Ann Surg. 2024;279(5):880–4. [DOI] [PubMed] [Google Scholar]
  • 5.Yadav K, Lewis RJ. Immortal time bias in observational studies. JAMA. 2021;325(7):686–7. [DOI] [PubMed] [Google Scholar]
  • 6.Robins JM, Hernan MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology. 2000;11(5):550–60. [DOI] [PubMed] [Google Scholar]
  • 7.Hernan MA, Brumback BA, Robins JM. Estimating the causal effect of zidovudine on CD4 count with a marginal structural model for repeated measures. Stat Med. 2002;21(12):1689–709. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Further information on the original manuscript is available from the corresponding authors upon reasonable request. No datasets were generated or analysed during the current study.


Articles from Critical Care are provided here courtesy of BMC

RESOURCES