The performance of risk prediction models for pre‐eclampsia using routinely collected maternal characteristics and comparison with models that include specialised tests and with clinical guideline decision rules: a systematic review

Abstract

Background Risk prediction models may be valuable to identify women at risk of pre-eclampsia to guide aspirin prophylaxis in early pregnancy.Objective To assess the performance of 'simple' risk models for pre-eclampsia that use routinely collected maternal characteristics; compare with 'specialised' models that include specialised tests; and to guideline recommended decision rules.Search strategy MEDLINE, Embase and PubMed were searched to June 2014.Selection criteria We included studies that developed or validated pre-eclampsia risk models using maternal characteristics with or without specialised tests and reported model performance.Data collection and analysis We extracted data on study characteristics; model predictors, validation and performance including area under the curve (AUC), sensitivity and specificity.Main results We identified 29 studies that developed 70 models including 22 simple models. Studies included 151-9149 women with a pre-eclampsia prevalence of 1.2-9.5%. No single predictor was included in all models. Four simple models were externally validated, with a model using parity, pre-eclampsia history, race, chronic hypertension and conception method to predict early-onset pre-eclampsia achieving the highest AUC (0.76, 95% CI 0.74-0.77). Nine studies comparing simple versus specialized models in the same population reported AUC favouring specialised models. A simple model achieved fewer false positives than a guideline recommended risk factor list, but sensitivity to classify risk for aspirin prophylaxis was not assessed.Conclusion Validated simple pre-eclampsia risk models demonstrate good risk discrimination that can be improved with specialised tests. Further research is needed to determine their clinical value to guide aspirin prophylaxis compared with decision rules.Keywords Aspirin, pre-eclampsia, risk factors, risk prediction models, systematic review, validation.Tweetable abstract Pre-eclampsia risk models using maternal factors show good risk discrimination to guide aspirin prophylaxis.Please cite this paper as: Al-Rubaie ZTA, Askie LM, Ray JG, Hudson HM, Lord SJ. The performance of risk prediction models for pre-eclampsia using routinely collected maternal characteristics and comparison with models that include specialised tests and with clinical guideline decision rules: a systematic review.

Estimated reductions in provider-initiated preterm births and hospital length of stay under a universal acetylsalicylic acid prophylaxis strategy: a retrospective cohort study

“…19,30 Currently, there is no accurate means to screen for preeclampsia or intrauterine growth restriction. 17,18 Certainly, although we do not endorse use of ASA prophylaxis in all pregnant women and reiterate that the strategy should be focused at women at higher risk for preeclampsia and/or intra- uterine growth restriction, emerging data about ASA efficacy among women at differing risk may lead to a lowering of the threshold for starting ASA treatment. 22 The rate of preeclampsia and/or intrauterine growth restriction that we observed among all provider-initiated preterm births, 33.6%, is consistent with that noted by other investigators.…”

“…Clinical prediction models for preeclampsia 17 and intrauterine growth restriction 18 typically generate discriminative estimates of only 0.75, which are suboptimal in identifying pregnancies at high risk. In addition, over 50% of cases of preterm preeclampsia may not be preceded by any known risk factor(s).…”

External Validation and Clinical Usefulness of First Trimester Prediction Models for the Risk of Preeclampsia: A Prospective Cohort Study

“…The model of Kenny et al [18] stayed most stable in our external validation cohort. External validation studies of prediction models for PE are scarce [13,14,39,40]. None have simultaneously considered all first trimester prediction models based on routinely collected maternal characteristics for the risk of PE.…”

“…The paper of Syngelaki et al [20], included on the basis of citation lists, was not selected from our systematic search as the title and abstract did not clearly indicate that it addressed prediction models for a wide range of adverse pregnancy outcomes. Syngelaki's model was also not described in existing systematic reviews about prediction models for PE [13,14,39,49]. A limitation in the selection process is that we had to exclude a substantial proportion of prediction models as the algorithm was not available, even after contacting the authors.…”

“…Although specialized tests have been reported to improve predictive performance, a drawback is that these tests are not routinely performed or readily available in general antenatal settings, and generate substantial additional costs [13,14].…”