Jump to comment:

• Bird's eye view misses the benefits of customised assessment
  Jason Gardosi, Oliver Hugh and Emily Butler
  Published on: 4 November 2023

• Published on: 4 November 2023

  Bird's eye view misses the benefits of customised assessment

  • Jason Gardosi, Professor of Maternal and Perinatal Health Perinatal Institute and University of Warwick
  • Other Contributors:
    • Oliver Hugh, Senior Statistician
    • Emily Butler, Midwifery Programme Manager

  Kilpi and colleagues’ [1] comparative analysis of databases from the UK and several Nordic countries concluded that the benefits of customised vs population based standards for determining SGA are limited when assessing risk of stillbirth and other adverse outcomes.

  Regrettably their approach was unlikely to be able to elucidate the advantages of a customised standard. For each country’s cohort, the authors compared adverse outcomes for the lowest 10% of pregnancies by weight-for-gestation. Because of the strong association between SGA and outcomes such as stillbirth [2], this is where most at-risk pregnancies and adverse outcomes will be found, regardless of the standard used. We illustrated this in a comparative analysis of customised GROW and 3 other fetal weight standards (see table 2 in [3]): the relative risk for stillbirth was essentially the same for each standard, and customisation had little effect. The same would apply to the country cohorts in Kilpi et al [1].

  The fact that customisation adds little at country level has already been shown some time ago in the (not referenced) study from 24 countries [4], which combined population average birthweights with the proportionality equation we had developed for GROW. This approach can be useful in providing country specific charts where more data are not available, and ought to be supported [5]. However when there is more detailed information, the focus ought to be on subgroups of the population fo...

  Show More

  Kilpi and colleagues’ [1] comparative analysis of databases from the UK and several Nordic countries concluded that the benefits of customised vs population based standards for determining SGA are limited when assessing risk of stillbirth and other adverse outcomes.

  Regrettably their approach was unlikely to be able to elucidate the advantages of a customised standard. For each country’s cohort, the authors compared adverse outcomes for the lowest 10% of pregnancies by weight-for-gestation. Because of the strong association between SGA and outcomes such as stillbirth [2], this is where most at-risk pregnancies and adverse outcomes will be found, regardless of the standard used. We illustrated this in a comparative analysis of customised GROW and 3 other fetal weight standards (see table 2 in [3]): the relative risk for stillbirth was essentially the same for each standard, and customisation had little effect. The same would apply to the country cohorts in Kilpi et al [1].

  The fact that customisation adds little at country level has already been shown some time ago in the (not referenced) study from 24 countries [4], which combined population average birthweights with the proportionality equation we had developed for GROW. This approach can be useful in providing country specific charts where more data are not available, and ought to be supported [5]. However when there is more detailed information, the focus ought to be on subgroups of the population for whom customisation is intended.

  A good illustration of this is our (also not referenced) study of a large Swedish database, showing the advantages of customised assessment in different maternal size and parity categories [6]. One of the novel findings in that study was that the higher perinatal mortality risk of mothers with increasing body mass index was only associated with SGA when it was determined using a customised birthweight standard. This was recently confirmed in the above mentioned comparison of 4 international fetal weight standards in a cohort of over 2.2 million pregnancies [3]. Uncustomised standards fail to identify this association and present a patient safety risk for high BMI mothers.

  There are other instances in the article in question [1] where inclusion of relevant references would have presented a more balanced, rounded picture. In the Scottish study referred to [7], co-authored by the current study’s senior author, it was claimed that there was a reduction of stillbirths already before implementation of the GAP programme (which includes customised GROW charts, training, standardised protocols and audit [8]). In our (not referenced) response [9], we pointed out that the reported improvement in stillbirth rate was temporary and not sustained, and had coincided with the AFFIRM (reduced fetal movement) trial involving most Scottish units, which was associated with increased use of ultrasound scans [10]. When subsequently the GAP programme did have time to embed, there was a sustained reduction in stillbirths, reflecting the effect seen in English regions [9]. Although the data showing this longitudinal stillbirth trend in Scotland were readily available, they were omitted by the authors of that study [7], presumably as it did not fit the intended narrative.

  Finally, the Design randomised trial [11] is quoted as further evidence that the GAP program is ineffective, without mention of the wide critique and controversy that study has attracted. This included questions about effectiveness of randomisation of clusters, the sharp increase in ultrasound use in the control group only [12], as well as the incomplete implementation and short evaluation period which was terminated well before such a change management program could embed [8]. The effectiveness of GAP is dependent on completeness of implementation, as clearly demonstrated in a 10 year cohort study of Office of National Statistics data [13].

  Despite the failed execution of the Design trial, it was nevertheless able to show a significant reduction in stillbirths in the intervention arm, following earlier delivery of customised SGA cases. This was downplayed by the authors as a chance finding [11], but was in fact an example of how customised charts, at the individual, clinical rather than population level, allow a more confident assessment whether the baby of this mother, in this pregnancy, is at risk because of possible intrauterine growth restriction. This was also the interpretation of one of the authors of the Design trial [14], the only one of 25 who had clinical experience with customised charts.

  We submit therefore that the concluding statement in this article [1], that “widespread adoption of customised centile charts should not be promoted“, might reflect the authors’ opinion, but not the available evidence.

  1. Kilpi F, Jones HE, Magnus MC, Santorelli G, Højsgaard Schmidt LK, Urhoj SK, et al. Association between perinatal mortality and morbidity and customised and non-customised birthweight centiles in Denmark, Finland, Norway, Wales, and England: comparative, population based, record linkage study. bmjmed. 2023 Aug;2(1):e000521. https://bmjmedicine.bmj.com/lookup/doi/10.1136/bmjmed-2023-000521

  2. Gardosi J, Madurasinghe V, Williams M, Malik A, Francis A. Maternal and fetal risk factors for stillbirth: population based study. BMJ. 2013;346(3). https://doi.org/10.1136/bmj.f108

  3. Gardosi J, Hugh O. Stillbirth risk and smallness for gestational age according to Hadlock, Intergrowth-21st, WHO and GROW fetal weight standards: analysis by maternal ethnicity and body mass index. American Journal of Obstetrics & Gynecology. 2023 May 27;0(0). https://www.ajog.org/article/S0002-9378(23)00364-2/fulltext

  4. Mikolajczyk RT, Zhang J, Betran AP, Souza JP, Mori R, Gülmezoglu AM, et al. A global reference for fetal-weight and birthweight percentiles. The Lancet. 2011;377(9780):1855–61. http://www.sciencedirect.com/science/article/pii/S0140673611603644

  5. Gardosi J. Fetal growth standards: individual and global perspectives. The Lancet. 2011 May;377(9780):1812–4. http://linkinghub.elsevier.com/retrieve/pii/S0140673611605072

  6. Gardosi J, Clausson B, Francis A. The value of customised centiles in assessing perinatal mortality risk associated with parity and maternal size. Br J Obstet Gynaecol. 2009;116(10):1356–63. https://doi.org/10.1111/j.1471-0528.2009.02245.x

  7. Iliodromiti S, Smith GCS, Lawlor DA, Pell JP, Nelson SM. UK stillbirth trends in over 11 million births provide no evidence to support effectiveness of Growth Assessment Protocol program. Ultrasound Obstet & Gynecol. 2020;55(5):599–604. https://doi.org/10.1002/uog.21999

  8. Butler E, Hugh O, Gardosi J. Evaluating the Growth Assessment Protocol for stillbirth prevention: progress and challenges. Journal of Perinatal Medicine. 2022;50:737–47. https://www.degruyter.com/document/doi/10.1515/jpm-2022-0209/html

  9. Gardosi J, Turner S, Williams M, Buller S, Hugh O, Francis A. The Growth Assessment Protocol: a major cause of declining stillbirth rates in the UK. Ultrasound Obstet Gynecol. 2020;56(1):117–9. https://doi.org/10.1002/uog.22100

  10. Norman JE, Heazell AEP, Rodriguez A, Weir CJ, Stock SJE, Calderwood CJ, et al. Awareness of fetal movements and care package to reduce fetal mortality (AFFIRM): a stepped wedge, cluster-randomised trial. The Lancet. 2018 Sep; https://linkinghub.elsevier.com/retrieve/pii/S0140673618315435

  11. Vieira MC, Relph S, Muruet-Gutierrez W, Elstad M, Coker B, Moitt N, et al. Evaluation of the Growth Assessment Protocol (GAP) for antenatal detection of small for gestational age: The DESiGN cluster randomised trial. PLOS Medicine. 2022 Jun 21;19(6):e1004004. https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1...

  12. DESIGN Trial - Comments. PLOS Medicine 2022; https://journals.plos.org/plosmedicine/article/comments?id=10.1371/journ...

  13. Hugh O, Williams M, Turner S, Gardosi J. Reduction of stillbirths in England from 2008 to 2017 according to uptake of the Growth Assessment Protocol: 10-year population-based cohort study. Ultrasound in Obstetrics & Gynecology. 2021;57(3):401–8. https://onlinelibrary.wiley.com/doi/abs/10.1002/uog.22187

  14. McCowan L. Reduction in stillbirths in the DESiGN trial. PLOS Medicine. 2022 Jun 23; https://journals.plos.org/plosmedicine/article/comment?id=10.1371/annota...

  Show Less

  Conflict of Interest:
  The authors work for the Perinatal Institute, a not-for-profit social enterprise which provides training and customized growth chart software referred to in this article.

  • Back to top