Can risk prediction models help us individualise stillbirth prevention?: A systematic review and critical appraisal of published risk models

R Townsend; A Manji; J Allotey; Aep Heazell; L Jorgensen; L A Magee; B W Mol; Kie Snell; R D Riley; J Sandall; Gcs Smith; M Patel; B Thilaganathan; P von Dadelszen; S Thangaratinam; A Khalil

doi:10.1111/1471-0528.16487

Can risk prediction models help us individualise stillbirth prevention? A systematic review and critical appraisal of published risk models

R Townsend, A Manji, J Allotey, Aep Heazell, L Jorgensen, L A Magee, B W Mol, Kie Snell, R D Riley, J Sandall, Gcs Smith, M Patel, B Thilaganathan, P von Dadelszen, S Thangaratinam, A Khalil^* (Corresponding Author)

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

12 Citations (Scopus)

Abstract

BACKGROUND: Stillbirth prevention is an international priority - risk prediction models could individualise care and reduce unnecessary intervention, but their use requires evaluation.

OBJECTIVES: To identify risk prediction models for stillbirth, and assess their potential accuracy and clinical benefit in practice.

SEARCH STRATEGY: MEDLINE, Embase, DH-DATA and AMED databases were searched from inception to June 2019 using terms relevant to stillbirth, perinatal mortality and prediction models. The search was compliant with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.

SELECTION CRITERIA: Studies developing and/or validating prediction models for risk of stillbirth developed for application during pregnancy.

DATA COLLECTION AND ANALYSIS: Study screening and data extraction were conducted in duplicate, using the CHARMS checklist. Risk of bias was appraised using the PROBAST tool.

RESULTS: The search identified 2751 citations. Fourteen studies reporting development of 69 models were included. Variables consistently included were: ethnicity, body mass index, uterine artery Doppler, pregnancy-associated plasma protein and placental growth factor. For almost all models there were significant concerns about risk of bias. Apparent model performance (i.e. in the development dataset) was highest in models developed for use later in pregnancy and including maternal characteristics, and ultrasound and biochemical variables, but few were internally validated and none were externally validated.

CONCLUSIONS: Almost all models identified were at high risk of bias. There are first-trimester models of possible clinical benefit in early risk stratification; these require validation and clinical evaluation. There were few later pregnancy models but, if validated, these could be most relevant to individualised discussions around timing of birth.

TWEETABLE ABSTRACT: Prediction models using maternal factors, blood tests and ultrasound could individualise stillbirth prevention, but existing models are at high risk of bias.

Original language	English
Pages (from-to)	214-224
Number of pages	11
Journal	BJOG : an international journal of obstetrics and gynaecology
Volume	128
Issue number	2
Early online date	13 Oct 2020
DOIs	https://doi.org/10.1111/1471-0528.16487
Publication status	Published - 1 Jan 2021

Bibliographical note

Funding Information:
AH reports grants from Tommy’s and Action Medical Research, outside the submitted work. BWM reports grants from the National Health and Medical Research Council (NHMRC) and personal fees from Obseva, Merck, Merck KGaA, Guerbet and iGenomix, outside the submitted work. GCS reports grants and personal fees from GlaxoSmithKline Research and Development Limited, grants from Sera Prognostics Inc., non‐financial support from Illumina Inc., grants, personal fees and non‐financial support from Roche Diagnostics Ltd, outside the submitted work. In addition, GCS is a named inventor on a patent application submitted by Cambridge Enterprise for a biomarker test to predict human fetal growth restriction pending. JS reports support from the NIHR Applied Research Collaboration South London, outside the submitted work. RT, AK, RR and ST report receipt of a grant from the Stillbirth and Neonatal Death Society (Sands) during the conduct of this study. AM, JA, LJ, LM, PvD, MP and KS have no disclosures to declare. Completed disclosure of interest forms are available to view online as supporting information.

Funding Information:
Jane Sandall is an NIHR Senior Investigator and is also supported by the National Institute for Health Research (NIHR) Applied Research Collaboration South London (NIHR ARC South London) at King’s College Hospital NHS Foundation Trust. The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.

Funding Information:
The authors are collaborators in the IPPIC (International Prediction of Pregnancy Complications) stillbirth project, funded by Sands (the Stillbirth and Neonatal Death Society).

Keywords

Female
Humans
Perinatal Death/etiology
Predictive Value of Tests
Pregnancy
Risk Assessment
Stillbirth
stillbirth
fetal medicine
Epidemiology
serum screening
perinatal
prediction
model
systematic reviews
NUMBER
EVENTS
PREGNANCY
UTERINE ARTERY DOPPLER
GESTATION
PLACENTAL GROWTH-FACTOR

Access to Document

10.1111/1471-0528.16487Licence: Unspecified

Cite this

Townsend, R., Manji, A., Allotey, J., Heazell, A., Jorgensen, L., Magee, L. A., Mol, B. W., Snell, K., Riley, R. D., Sandall, J., Smith, G., Patel, M., Thilaganathan, B., von Dadelszen, P., Thangaratinam, S., & Khalil, A. (2021). Can risk prediction models help us individualise stillbirth prevention? A systematic review and critical appraisal of published risk models. BJOG : an international journal of obstetrics and gynaecology, 128(2), 214-224. https://doi.org/10.1111/1471-0528.16487

Townsend, R, Manji, A, Allotey, J, Heazell, A, Jorgensen, L, Magee, LA, Mol, BW, Snell, K, Riley, RD, Sandall, J, Smith, G, Patel, M, Thilaganathan, B, von Dadelszen, P, Thangaratinam, S & Khalil, A 2021, 'Can risk prediction models help us individualise stillbirth prevention? A systematic review and critical appraisal of published risk models', BJOG : an international journal of obstetrics and gynaecology, vol. 128, no. 2, pp. 214-224. https://doi.org/10.1111/1471-0528.16487

@article{27656a4049bf43288fe046720053078a,

title = "Can risk prediction models help us individualise stillbirth prevention?: A systematic review and critical appraisal of published risk models",

abstract = "BACKGROUND: Stillbirth prevention is an international priority - risk prediction models could individualise care and reduce unnecessary intervention, but their use requires evaluation.OBJECTIVES: To identify risk prediction models for stillbirth, and assess their potential accuracy and clinical benefit in practice.SEARCH STRATEGY: MEDLINE, Embase, DH-DATA and AMED databases were searched from inception to June 2019 using terms relevant to stillbirth, perinatal mortality and prediction models. The search was compliant with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.SELECTION CRITERIA: Studies developing and/or validating prediction models for risk of stillbirth developed for application during pregnancy.DATA COLLECTION AND ANALYSIS: Study screening and data extraction were conducted in duplicate, using the CHARMS checklist. Risk of bias was appraised using the PROBAST tool.RESULTS: The search identified 2751 citations. Fourteen studies reporting development of 69 models were included. Variables consistently included were: ethnicity, body mass index, uterine artery Doppler, pregnancy-associated plasma protein and placental growth factor. For almost all models there were significant concerns about risk of bias. Apparent model performance (i.e. in the development dataset) was highest in models developed for use later in pregnancy and including maternal characteristics, and ultrasound and biochemical variables, but few were internally validated and none were externally validated.CONCLUSIONS: Almost all models identified were at high risk of bias. There are first-trimester models of possible clinical benefit in early risk stratification; these require validation and clinical evaluation. There were few later pregnancy models but, if validated, these could be most relevant to individualised discussions around timing of birth.TWEETABLE ABSTRACT: Prediction models using maternal factors, blood tests and ultrasound could individualise stillbirth prevention, but existing models are at high risk of bias.",

keywords = "Female, Humans, Perinatal Death/etiology, Predictive Value of Tests, Pregnancy, Risk Assessment, Stillbirth, stillbirth, fetal medicine, Epidemiology, serum screening, perinatal, prediction, model, systematic reviews, NUMBER, EVENTS, PREGNANCY, UTERINE ARTERY DOPPLER, GESTATION, PLACENTAL GROWTH-FACTOR",

author = "R Townsend and A Manji and J Allotey and Aep Heazell and L Jorgensen and Magee, {L A} and Mol, {B W} and Kie Snell and Riley, {R D} and J Sandall and Gcs Smith and M Patel and B Thilaganathan and {von Dadelszen}, P and S Thangaratinam and A Khalil",

note = "Funding Information: AH reports grants from Tommy{\textquoteright}s and Action Medical Research, outside the submitted work. BWM reports grants from the National Health and Medical Research Council (NHMRC) and personal fees from Obseva, Merck, Merck KGaA, Guerbet and iGenomix, outside the submitted work. GCS reports grants and personal fees from GlaxoSmithKline Research and Development Limited, grants from Sera Prognostics Inc., non‐financial support from Illumina Inc., grants, personal fees and non‐financial support from Roche Diagnostics Ltd, outside the submitted work. In addition, GCS is a named inventor on a patent application submitted by Cambridge Enterprise for a biomarker test to predict human fetal growth restriction pending. JS reports support from the NIHR Applied Research Collaboration South London, outside the submitted work. RT, AK, RR and ST report receipt of a grant from the Stillbirth and Neonatal Death Society (Sands) during the conduct of this study. AM, JA, LJ, LM, PvD, MP and KS have no disclosures to declare. Completed disclosure of interest forms are available to view online as supporting information. Funding Information: Jane Sandall is an NIHR Senior Investigator and is also supported by the National Institute for Health Research (NIHR) Applied Research Collaboration South London (NIHR ARC South London) at King{\textquoteright}s College Hospital NHS Foundation Trust. The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. Funding Information: The authors are collaborators in the IPPIC (International Prediction of Pregnancy Complications) stillbirth project, funded by Sands (the Stillbirth and Neonatal Death Society). ",

year = "2021",

month = jan,

day = "1",

doi = "10.1111/1471-0528.16487",

language = "English",

volume = "128",

pages = "214--224",

journal = "BJOG : an international journal of obstetrics and gynaecology",

issn = "1470-0328",

publisher = "John Wiley & Sons, Ltd (10.1111)",

number = "2",

}

TY - JOUR

T1 - Can risk prediction models help us individualise stillbirth prevention?

T2 - A systematic review and critical appraisal of published risk models

AU - Townsend, R

AU - Manji, A

AU - Allotey, J

AU - Heazell, Aep

AU - Jorgensen, L

AU - Magee, L A

AU - Mol, B W

AU - Snell, Kie

AU - Riley, R D

AU - Sandall, J

AU - Smith, Gcs

AU - Patel, M

AU - Thilaganathan, B

AU - von Dadelszen, P

AU - Thangaratinam, S

AU - Khalil, A

N1 - Funding Information: AH reports grants from Tommy’s and Action Medical Research, outside the submitted work. BWM reports grants from the National Health and Medical Research Council (NHMRC) and personal fees from Obseva, Merck, Merck KGaA, Guerbet and iGenomix, outside the submitted work. GCS reports grants and personal fees from GlaxoSmithKline Research and Development Limited, grants from Sera Prognostics Inc., non‐financial support from Illumina Inc., grants, personal fees and non‐financial support from Roche Diagnostics Ltd, outside the submitted work. In addition, GCS is a named inventor on a patent application submitted by Cambridge Enterprise for a biomarker test to predict human fetal growth restriction pending. JS reports support from the NIHR Applied Research Collaboration South London, outside the submitted work. RT, AK, RR and ST report receipt of a grant from the Stillbirth and Neonatal Death Society (Sands) during the conduct of this study. AM, JA, LJ, LM, PvD, MP and KS have no disclosures to declare. Completed disclosure of interest forms are available to view online as supporting information. Funding Information: Jane Sandall is an NIHR Senior Investigator and is also supported by the National Institute for Health Research (NIHR) Applied Research Collaboration South London (NIHR ARC South London) at King’s College Hospital NHS Foundation Trust. The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. Funding Information: The authors are collaborators in the IPPIC (International Prediction of Pregnancy Complications) stillbirth project, funded by Sands (the Stillbirth and Neonatal Death Society).

PY - 2021/1/1

Y1 - 2021/1/1

N2 - BACKGROUND: Stillbirth prevention is an international priority - risk prediction models could individualise care and reduce unnecessary intervention, but their use requires evaluation.OBJECTIVES: To identify risk prediction models for stillbirth, and assess their potential accuracy and clinical benefit in practice.SEARCH STRATEGY: MEDLINE, Embase, DH-DATA and AMED databases were searched from inception to June 2019 using terms relevant to stillbirth, perinatal mortality and prediction models. The search was compliant with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.SELECTION CRITERIA: Studies developing and/or validating prediction models for risk of stillbirth developed for application during pregnancy.DATA COLLECTION AND ANALYSIS: Study screening and data extraction were conducted in duplicate, using the CHARMS checklist. Risk of bias was appraised using the PROBAST tool.RESULTS: The search identified 2751 citations. Fourteen studies reporting development of 69 models were included. Variables consistently included were: ethnicity, body mass index, uterine artery Doppler, pregnancy-associated plasma protein and placental growth factor. For almost all models there were significant concerns about risk of bias. Apparent model performance (i.e. in the development dataset) was highest in models developed for use later in pregnancy and including maternal characteristics, and ultrasound and biochemical variables, but few were internally validated and none were externally validated.CONCLUSIONS: Almost all models identified were at high risk of bias. There are first-trimester models of possible clinical benefit in early risk stratification; these require validation and clinical evaluation. There were few later pregnancy models but, if validated, these could be most relevant to individualised discussions around timing of birth.TWEETABLE ABSTRACT: Prediction models using maternal factors, blood tests and ultrasound could individualise stillbirth prevention, but existing models are at high risk of bias.

AB - BACKGROUND: Stillbirth prevention is an international priority - risk prediction models could individualise care and reduce unnecessary intervention, but their use requires evaluation.OBJECTIVES: To identify risk prediction models for stillbirth, and assess their potential accuracy and clinical benefit in practice.SEARCH STRATEGY: MEDLINE, Embase, DH-DATA and AMED databases were searched from inception to June 2019 using terms relevant to stillbirth, perinatal mortality and prediction models. The search was compliant with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.SELECTION CRITERIA: Studies developing and/or validating prediction models for risk of stillbirth developed for application during pregnancy.DATA COLLECTION AND ANALYSIS: Study screening and data extraction were conducted in duplicate, using the CHARMS checklist. Risk of bias was appraised using the PROBAST tool.RESULTS: The search identified 2751 citations. Fourteen studies reporting development of 69 models were included. Variables consistently included were: ethnicity, body mass index, uterine artery Doppler, pregnancy-associated plasma protein and placental growth factor. For almost all models there were significant concerns about risk of bias. Apparent model performance (i.e. in the development dataset) was highest in models developed for use later in pregnancy and including maternal characteristics, and ultrasound and biochemical variables, but few were internally validated and none were externally validated.CONCLUSIONS: Almost all models identified were at high risk of bias. There are first-trimester models of possible clinical benefit in early risk stratification; these require validation and clinical evaluation. There were few later pregnancy models but, if validated, these could be most relevant to individualised discussions around timing of birth.TWEETABLE ABSTRACT: Prediction models using maternal factors, blood tests and ultrasound could individualise stillbirth prevention, but existing models are at high risk of bias.

KW - Female

KW - Humans

KW - Perinatal Death/etiology

KW - Predictive Value of Tests

KW - Pregnancy

KW - Risk Assessment

KW - Stillbirth

KW - stillbirth

KW - fetal medicine

KW - Epidemiology

KW - serum screening

KW - perinatal

KW - prediction

KW - model

KW - systematic reviews

KW - NUMBER

KW - EVENTS

KW - PREGNANCY

KW - UTERINE ARTERY DOPPLER

KW - GESTATION

KW - PLACENTAL GROWTH-FACTOR

UR - http://www.scopus.com/inward/record.url?scp=85092332912&partnerID=8YFLogxK

U2 - 10.1111/1471-0528.16487

DO - 10.1111/1471-0528.16487

M3 - Article

C2 - 32894620

SN - 1470-0328

VL - 128

SP - 214

EP - 224

JO - BJOG : an international journal of obstetrics and gynaecology

JF - BJOG : an international journal of obstetrics and gynaecology

IS - 2

ER -