Thromboembolism is rare in healthy pediatric patients, but it is an increasing problem in children with underlying medical conditions such as cancer. The increase in childhood thromboembolism over the past decade is thought to be due to both heightened awareness of the diagnosis and more invasive technologies used in children with underlying medical conditions.1,2 Understanding thromboembolism and developing safe treatment options for pediatric oncology patients is important due to the increased risk of death, organ dysfunction, and poor oncologic outcome. Consequences of thromboembolism also include increased hospital length of stay and cost.3 Current published evidence on treatment for thromboembolism in the pediatric oncology population is limited, and published guidelines are often extrapolated from adult trials.4,5 For example, in the 2018 guidelines released from the American Society of Hematology for treatment of thromboembolism in the pediatric population, although recommendations were made by a panel of experts, all the recommendations were limited by low or very low certainty in the evidence.6
A key step toward gathering better evidence regarding pediatric thromboembolism in the pediatric oncology population will be the development of validated methods for accurately quantifying thromboembolism diagnosis and outcomes. Much current epidemiological research uses administrative data, or “big data,” to identify cases of interest; however, validity of research findings based on these data depend on the validity of the search parameters. Validity depends on both the proper diagnostic coding by physicians and on the proper choice of search codes by the researchers. Current research in the field has been met with various challenges. For example, recent studies to ascertain the rates of childhood thrombosis have relied on discharge diagnosis codes for identifying thromboembolism cases.2 However, Burles et al. identified pitfalls of using discharge diagnosis code searches, highlighting the extensive presence of false positives and negatives in identifying thromboembolism cases.7 This highlights the fact that healthcare databases, designed primarily for administrative and billing purposes, often lack comprehensive clinical information crucial for research. This includes lack of details on diseases of interest, health outcomes, medications, data on comorbidities, and quality of life.8
Addressing the potential of administrative health care databases as validated sources for data, Doiron et al. discussed the benefits of linking large cohort studies with administrative data to enrich datasets, maximize resource utilization, and facilitate multidisciplinary research.9 Additionally, regular validation studies, evaluating different code combinations or algorithms, are crucial for ensuring data accuracy, particularly in pediatric populations where such studies are limited.10 In the current manuscript, Athale et al. tested the validity of using combinations of ICD and medication codes from large Canadian administrative databases, with a curated oncology database for case verification, to identify thromboembolism diagnoses in children undergoing primary cancer therapy. Multiple query algorithms were tested and validated using the oncology database. The best performing algorithm resulted in a sensitivity of 76% and specificity of 86% for identifying pediatric oncology patients with thromboembolism. Of note, the same analysis improved sensitivity to 84% when using exclusively ICD-10 codes, highlighting the previously reported limitations in using ICD-9 codes for epidemiological research11.
This study demonstrates the validation of search parameters for accurately identifying thromboembolism cases in pediatric populations undergoing cancer therapy using multiple administrative databases in conjunction with a large oncology database. These findings could be instrumental for future epidemiological and outcomes research in this area. Future research will be needed to validate this algorithm in other health care systems. Further validation research can also extend this algorithm to other populations such as neonates, or those with other high-risk conditions for thromboembolism. Such studies would test the algorithm’s generalizability and applicability in diverse clinical settings. Outside of the thromboembolism field this study can serve as a model for validation strategies for big data research in other diseases.
References
Monagle, P. et al. American Society of Hematology 2018 Guidelines for management of venous thromboembolism: treatment of pediatric venous thromboembolism. Blood Adv. 2, 3292–3316 (2018).
Raffini, L. et al. Dramatic increase in venous thromboembolism in children’s hospitals in the United States from 2001 to 2007. Pediatrics 124, 1001–1008 (2009).
Goudie, A. et al. Costs of venous thromboembolism, catheter-associated urinary tract infection, and pressure ulcer. Pediatrics 136, 432–439 (2015).
Wiernikowski, J. T. & Athale, U. H. Thromboembolic complications in children with cancer. Thromb. Res 118, 137–152 (2006).
Law, C. & Raffini, L. A guide to the use of anticoagulant drugs in children. Paediatr. Drugs 17, 105–114 (2015).
Monagle P, et al. American Society of Hematology 2018 Guidelines for management of venous thromboembolism: treatment of pediatric venous thromboembolism. Blood Adv 2, 3292–3316 (2018).
Burles, K., Innes, G., Senior, K., Lang, E. & McRae, A. Limitations of pulmonary embolism ICD-10 codes in emergency department administrative data: let the buyer beware. BMC Med. Res. Methodol. 17, 89 (2017).
Jaffray, J. et al. Development of a risk model for pediatric hospital-acquired thrombosis: a report from the Children’s Hospital-Acquired Thrombosis Consortium. J. Pediatr. 228, 252–259.e1 (2021).
Doiron, D., Raina, P. & Fortier, I. Linkage between cohorts and health care utilization data: meeting of Canadian Stakeholders workshop participants. Linking Canadian population health data: maximizing the potential of cohort and administrative data. Can. J. Public Health 104, e258–e261 (2013).
Ulrich, E. H. et al. A review on the application and limitations of administrative health care data for the study of acute kidney injury epidemiology and outcomes in children. Front. Pediatr. 9, 742888 (2021).
Lau, B. D. et al. ICD-9 Code-Based Venous Thromboembolism Performance Targets Fail to Measure Up. Am J Med Qual 31, 448–453 (2016).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Blankenhorn, K., Mitchell, W.B. Finding pediatric thromboembolism: needles in a big data haystack. Pediatr Res (2024). https://doi.org/10.1038/s41390-024-03186-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41390-024-03186-4