Why was this transfusion given? Identifying clinical indications for blood transfusion in health care data

Loan R. van Hoeven* (Corresponding Author), Aukje L. Kreuger, Kit C.B. Roes, Peter F. Kemper, Hendrik Koffijberg, Floris J. Kranenburg, Jan M.M. Rondeel, Mart P. Janssen

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

25 Downloads (Pure)

Abstract

Background: To enhance the utility of transfusion data for research, ideally every transfusion should be linked to a primary clinical indication. In electronic patient records, many diagnostic and procedural codes are registered, but unfortunately, it is usually not specified which one is the reason for transfusion. Therefore, a method is needed to determine the most likely indication for transfusion in an automated way.

Study design and methods: An algorithm to identify the most likely transfusion indication was developed and evaluated against a gold standard based on the review of medical records for 234 cases by 2 experts. In a second step, information on misclassification was used to fine-tune the initial algorithm. The adapted algorithm predicts, out of all data available, the most likely indication for transfusion using information on medical specialism, surgical procedures, and diagnosis and procedure dates relative to the transfusion date.

Results: The adapted algorithm was able to predict 74.4% of indications in the sample correctly (extrapolated to the full data set 75.5%). A kappa score, which corrects for the number of options to choose from, was found of 0.63. This indicates that the algorithm performs substantially better than chance level.

Conclusion: It is possible to use an automated algorithm to predict the indication for transfusion in terms of procedures and/or diagnoses. Before implementation of the algorithm in other data sets, the obtained results should be externally validated in an independent hospital data set.

Original languageEnglish
Pages (from-to)353-362
Number of pages10
JournalClinical Epidemiology
Volume10
DOIs
Publication statusPublished - 29 Mar 2018

Fingerprint

Blood Transfusion
Delivery of Health Care
Medical Records
Research
Datasets

Keywords

  • Electronic health record data
  • Indication for transfusion
  • Selection algorithm

Cite this

van Hoeven, L. R., Kreuger, A. L., Roes, K. C. B., Kemper, P. F., Koffijberg, H., Kranenburg, F. J., ... Janssen, M. P. (2018). Why was this transfusion given? Identifying clinical indications for blood transfusion in health care data. Clinical Epidemiology, 10, 353-362. https://doi.org/10.2147/CLEP.S147142
van Hoeven, Loan R. ; Kreuger, Aukje L. ; Roes, Kit C.B. ; Kemper, Peter F. ; Koffijberg, Hendrik ; Kranenburg, Floris J. ; Rondeel, Jan M.M. ; Janssen, Mart P. / Why was this transfusion given? Identifying clinical indications for blood transfusion in health care data. In: Clinical Epidemiology. 2018 ; Vol. 10. pp. 353-362.
@article{8aeb5540e9f64952b41c51a574f3f54f,
title = "Why was this transfusion given? Identifying clinical indications for blood transfusion in health care data",
abstract = "Background: To enhance the utility of transfusion data for research, ideally every transfusion should be linked to a primary clinical indication. In electronic patient records, many diagnostic and procedural codes are registered, but unfortunately, it is usually not specified which one is the reason for transfusion. Therefore, a method is needed to determine the most likely indication for transfusion in an automated way.Study design and methods: An algorithm to identify the most likely transfusion indication was developed and evaluated against a gold standard based on the review of medical records for 234 cases by 2 experts. In a second step, information on misclassification was used to fine-tune the initial algorithm. The adapted algorithm predicts, out of all data available, the most likely indication for transfusion using information on medical specialism, surgical procedures, and diagnosis and procedure dates relative to the transfusion date.Results: The adapted algorithm was able to predict 74.4{\%} of indications in the sample correctly (extrapolated to the full data set 75.5{\%}). A kappa score, which corrects for the number of options to choose from, was found of 0.63. This indicates that the algorithm performs substantially better than chance level. Conclusion: It is possible to use an automated algorithm to predict the indication for transfusion in terms of procedures and/or diagnoses. Before implementation of the algorithm in other data sets, the obtained results should be externally validated in an independent hospital data set.",
keywords = "Electronic health record data, Indication for transfusion, Selection algorithm",
author = "{van Hoeven}, {Loan R.} and Kreuger, {Aukje L.} and Roes, {Kit C.B.} and Kemper, {Peter F.} and Hendrik Koffijberg and Kranenburg, {Floris J.} and Rondeel, {Jan M.M.} and Janssen, {Mart P.}",
year = "2018",
month = "3",
day = "29",
doi = "10.2147/CLEP.S147142",
language = "English",
volume = "10",
pages = "353--362",
journal = "Clinical Epidemiology",
issn = "1179-1349",
publisher = "Dove Medical Press Ltd.",

}

van Hoeven, LR, Kreuger, AL, Roes, KCB, Kemper, PF, Koffijberg, H, Kranenburg, FJ, Rondeel, JMM & Janssen, MP 2018, 'Why was this transfusion given? Identifying clinical indications for blood transfusion in health care data', Clinical Epidemiology, vol. 10, pp. 353-362. https://doi.org/10.2147/CLEP.S147142

Why was this transfusion given? Identifying clinical indications for blood transfusion in health care data. / van Hoeven, Loan R. (Corresponding Author); Kreuger, Aukje L.; Roes, Kit C.B.; Kemper, Peter F.; Koffijberg, Hendrik; Kranenburg, Floris J.; Rondeel, Jan M.M.; Janssen, Mart P.

In: Clinical Epidemiology, Vol. 10, 29.03.2018, p. 353-362.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Why was this transfusion given? Identifying clinical indications for blood transfusion in health care data

AU - van Hoeven, Loan R.

AU - Kreuger, Aukje L.

AU - Roes, Kit C.B.

AU - Kemper, Peter F.

AU - Koffijberg, Hendrik

AU - Kranenburg, Floris J.

AU - Rondeel, Jan M.M.

AU - Janssen, Mart P.

PY - 2018/3/29

Y1 - 2018/3/29

N2 - Background: To enhance the utility of transfusion data for research, ideally every transfusion should be linked to a primary clinical indication. In electronic patient records, many diagnostic and procedural codes are registered, but unfortunately, it is usually not specified which one is the reason for transfusion. Therefore, a method is needed to determine the most likely indication for transfusion in an automated way.Study design and methods: An algorithm to identify the most likely transfusion indication was developed and evaluated against a gold standard based on the review of medical records for 234 cases by 2 experts. In a second step, information on misclassification was used to fine-tune the initial algorithm. The adapted algorithm predicts, out of all data available, the most likely indication for transfusion using information on medical specialism, surgical procedures, and diagnosis and procedure dates relative to the transfusion date.Results: The adapted algorithm was able to predict 74.4% of indications in the sample correctly (extrapolated to the full data set 75.5%). A kappa score, which corrects for the number of options to choose from, was found of 0.63. This indicates that the algorithm performs substantially better than chance level. Conclusion: It is possible to use an automated algorithm to predict the indication for transfusion in terms of procedures and/or diagnoses. Before implementation of the algorithm in other data sets, the obtained results should be externally validated in an independent hospital data set.

AB - Background: To enhance the utility of transfusion data for research, ideally every transfusion should be linked to a primary clinical indication. In electronic patient records, many diagnostic and procedural codes are registered, but unfortunately, it is usually not specified which one is the reason for transfusion. Therefore, a method is needed to determine the most likely indication for transfusion in an automated way.Study design and methods: An algorithm to identify the most likely transfusion indication was developed and evaluated against a gold standard based on the review of medical records for 234 cases by 2 experts. In a second step, information on misclassification was used to fine-tune the initial algorithm. The adapted algorithm predicts, out of all data available, the most likely indication for transfusion using information on medical specialism, surgical procedures, and diagnosis and procedure dates relative to the transfusion date.Results: The adapted algorithm was able to predict 74.4% of indications in the sample correctly (extrapolated to the full data set 75.5%). A kappa score, which corrects for the number of options to choose from, was found of 0.63. This indicates that the algorithm performs substantially better than chance level. Conclusion: It is possible to use an automated algorithm to predict the indication for transfusion in terms of procedures and/or diagnoses. Before implementation of the algorithm in other data sets, the obtained results should be externally validated in an independent hospital data set.

KW - Electronic health record data

KW - Indication for transfusion

KW - Selection algorithm

UR - http://www.scopus.com/inward/record.url?scp=85047774476&partnerID=8YFLogxK

U2 - 10.2147/CLEP.S147142

DO - 10.2147/CLEP.S147142

M3 - Article

AN - SCOPUS:85047774476

VL - 10

SP - 353

EP - 362

JO - Clinical Epidemiology

JF - Clinical Epidemiology

SN - 1179-1349

ER -