Reader Comments

Post a new comment on this article

DESiGN Trial confirms benefit of GAP program

Posted by jgardosi on 27 Jun 2022 at 14:53 GMT

When in 2015 the Perinatal Institute agreed to support a proposed London based cluster randomised controlled trial to assess the effectiveness of GAP, we did not quite appreciate how many problems the investigators would have to face. Seven years later, we are sorry to see that these problems have resulted in a report [1] with contentious findings and conclusions. It does however also have a number of important learning points:

1. Change management is a challenge in our National Health Service at the best of times. The ‘clusters’ available for the trial were those remaining from the 19 London Trusts that had not previously opted to join GAP. One further Trust withdrew before randomisation, wanting to implement GAP without delay, while two others dropped out following randomisation, and another Trust needed to be recruited from outside London. Our GAP trainers observed that most units in the intervention arm had competing priorities which manifested in partial engagement and slow progress – including delays of up to 12 months between GAP training and go-live. This reduced the effectiveness of staff training.

2. Implementation was incomplete. The authors report problems with our pre-agreed minimal criteria for implementation (Table 1), of which the most important one was the failure of units to record and review their own outcome data, to monitor, benchmark and improve performance - a criterion closely linked with effectiveness [2]. Based on information entered locally into the GROW software database, three units in the intervention arm were able to fulfil this requirement - but only during 2019, after the end of the relatively short ‘outcome period’ designated by the investigators. The average SGA detection rate for these 3 clusters increased from 16.8% (baseline) to 44% in 2019 (Figure 5 in [3]), which was comparable to our national GAP user average (Figure 2 in [3]).

3. The published detection rates are likely to be incorrect. Even with incomplete implementation, the antenatal detection of SGA rates of 20+% (Table 4) seem too low compared to our national data. Similarly, the false positive rate of only 1+% seems too low. The SGA rates of ultrasound EFW data are not given but a quick calculation using the false positive and detection rates in Table 4 shows that the SGA rates of EFWs ranged from 2.4 to 3.7%. This suggests that <3rd centile EFW was used for the analysis of screening performance, whereas according to the methods section (page 7), antenatal detection of SGA was defined as ultrasound-derived EFW <10th centile by customised or population based charts. In our GAP database, the EFW <10th centile SGA rate is about 10%. Clearly, the detection rates will be higher with the 10th centile EFW as the cut-off, and may affect the trial arms differently.

4. The control group was not fit for purpose. As the authors suggest (page 6), it may have been unethical to try to stop the standard care clusters from implementing ‘concomitant strategies for improved detection of SGA and prevention of stillbirths’ such as the NHS England Saving Babies Lives Care Bundle (SBL-CB); however the effects on the ‘standard care’ arm need to be considered. According to Table 3, the proportion of pregnancies with ultrasound EFW >24 weeks in the control cluster rose from 43.7% pre- to 75.7% post randomisation - a 73% increase (effect size, based on table 3 data: 31.9, 95% CI 31.1-32.9; p<0.01). In contrast, there was minimal increase in the intervention arm, from 62.6% to 64.0%.

It is worth noting that the SBL-CB algorithm for fetal growth assessment adopted by NHS England (page 19 in [4]) was an exact copy of the GAP algorithm (S1 Protocol, page 30) which we had developed several years earlier, with the exception that SBL-CB did not specify which growth chart to use. The lack of increase in scans in the intervention arm was likely to be due to fewer referrals based on standardised fundal height measurements plotted on customised charts, in contrast to the various practices in the standard care clusters, as described in (S2 Appendix).

5. SGA detection rates need to be interpreted with caution. The aforementioned steep rise in scans in the ‘standard care’ arm was the likely reason for the rise in SGA detection in that group. According to the ‘SGA by customised centile’ data in Table 4, the rates were 14.9% pre-randomisation and 21.5% in the outcome period – a 44% increase. The detection rate in the intervention arm during the outcome period was not much better: 22.3%. However in light of the sharp rise in standard care arm scans following SBL-CB, it could be considered more appropriate to compare the 22.3% detection rate with the 14.9% pre-randomisation rate as a more valid control, and this shows a modest but significant improvement in the GAP arm (effect size 7.4, CI 4.9-10.0, p<0.01). More importantly, as shown in point 2 above, detection rates in three of the GAP clusters increased to levels comparable with the national GAP user average, once they were able to implement the program more fully. Significant improvement in detection rates compared to respective unit baselines have been observed in NHS GAP sites (Figure 2 in [3]) and in longitudinal studies elsewhere [5] [6]. Based on the limitations of the DESiGN study, the authors cannot conclude that GAP did not increase SGA detection rates.

6. The reduction in stillbirths in the intervention arm is not surprising. It is consistent with the decrease in stillbirths in GAP units when the program was introduced regionally [7], evaluated as part of the SBL-CB [8] or assessed nationally using Office of National Statistics data [2]. The authors dismiss their finding of significantly fewer stillbirths and perinatal deaths in the intervention arm (Table 5) as being only 2 in their list of 26 secondary outcomes. However the other secondary outcomes are not targets of GAP - the main objective of which is prevention of stillbirth. The finding may also not fit the authors’ ‘expected pathway’ via improved SGA detection rates (page 19) – even after the correction we suggest in point 3 above - but can be explained through another key benefit of customised charts: the improved confidence by clinicians to act when required.

A major concern with population based, one-size-fits-all charts is the need to avoid unnecessary intervention for constitutional smallness. Customised charts reduce false positive assessment [9], which is an even bigger task in multicultural populations as in DESiGN, with 44% of mothers non-‘White’ (Table 2). In our South Asian population for example, we found that 56% of EFWs plotted below the 10th centile on Hadlock EFW charts were normal size on customised charts, and these re-classified cases had the same perinatal mortality rate as the non SGA population – confirming that they were only constitutionally small [10]. The second major concern with population based charts is when an SGA plot is not taken seriously, and – for example – explained away because the mother is also small. Customised assessment helps clinicians in their management [3], on deciding whether to deliver, to investigate further, or to reassure and manage expectantly. It is likely that this applied here also in the GAP clusters for the cases that did get recognised as SGA antenatally. The earlier delivery of SGA babies in these units (Table 6) is also consistent with the lower SGA stillbirth rate in the intervention arm.

The authors state repeatedly that this is the first RCT looking at GAP, and portray their study as a ‘pragmatic trial capturing real life challenges’ (page 20). However downplaying the implications of special circumstances (units with competing priorities; short implementation period; change of ‘standard care’ in the control group), and the likely error in the calculated detection rates (section 3 above) limits the external validity of the DESiGN trial findings.

There is nevertheless a lot to be learnt from this trial for future efforts. In the meantime, we suggest that a more valid conclusion of this study would be that: "Despite incomplete implementation, substantial changes in the control arm, and problems in calculating accurate detection rates, introduction of the GAP program and use of customised growth charts resulted in a significant reduction in stillbirths and perinatal deaths".

Jason Gardosi, Emily Butler, Oliver Hugh


References

1. 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...

2. 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://doi.org/10.1002/u...

3. Butler E, Hugh O, Gardosi J. Evaluating the Growth Assessment Protocol for stillbirth prevention: progress and challenges. Journal of Perinatal Medicine. 2022 May 27; https://www.degruyter.com...

4. NHS England. Saving Babies Lives: A care bundle for reducing stillbirth. 2016. https://www.england.nhs.u...

5. Jayawardena L, Sheehan P. Introduction of a customised growth chart protocol increased detection of small for gestational age pregnancies in a tertiary Melbourne hospital. Australian and New Zealand Journal of Obstetrics and Gynaecology. 2018;0(0). https://doi.org/10.1111/a...

6. Cowan FJ, McKinlay CJD, Taylor RS, Wilson J, McAra‐Couper J, Garrett N, et al. Detection of small for gestational age babies and perinatal outcomes following implementation of the Growth Assessment Protocol at a New Zealand tertiary facility: An observational intervention study. Aust N Z J Obstet Gynaecol. 2020 Dec 19;ajo.13283. https://doi.org/10.1111/a...

7. Gardosi J, Giddings S, Clifford S, Wood L, Francis A. Association between reduced stillbirth rates in England and regional uptake of accreditation training in customised fetal growth assessment. BMJ open. 2013;3(12):e003942. http://dx.doi.org/10.1136...

8. Widdows K, Roberts S, Camacho E, Heazell A. Evaluation of the implementation of the Saving Babies’ Lives Care Bundle in early adopter NHS Trusts in England (SPIRE). Manchester, UK: Maternal and Fetal Health Research Centre, University of Manchester; 2018 Jul. https://bit.ly/36rEqkd

9. Mongelli M, Gardosi J. Reduction of false-positive diagnosis of fetal growth restriction by application of customized fetal growth standards. Obstetrics & Gynecology. 1996;88(5):844–8. https://doi.org/10.1016/0...(96)00285-2

10. Giddings S, Clifford S, Madurasinghe V, Gardosi J. PFM.69 Customised vs uncustomised ultrasound charts in the assessment of perinatal mortality risk in the South Asian maternity population. Archives of Disease in Childhood - Fetal and Neonatal Edition. 2014 Jun 1;99(Suppl 1). http://fn.bmj.com/content...

Competing interests declared: The authors work for the Perinatal Institute, a not-for-profit social enterprise which provides the GAP program and customised GROW charts.

RE: DESiGN Trial confirms benefit of GAP program _ the response from the trial team

dpasupathy623 replied to jgardosi on 13 Jul 2022 at 08:30 GMT

Response to Perinatal Institute comment (https://journals.plos.org...):

We thank Gardosi et al for their comments on our manuscript ‘Evaluation of the Growth Assessment Protocol (GAP) for antenatal detection of small for gestational age: The DESiGN cluster randomised trial’,[1] and respond below to each of their concerns.

Detection of SGA

The calculation of detection rates for SGA in the DESiGN trial used the estimated fetal weight (EFW) <10th centile, as described.[1] Our findings concur with other studies where selective screening, as used in both arms of this trial, had a much lower false positive rate (1-2%) compared to universal screening with 3rd trimester scan (10%).[2]

Challenges in implementation

Competing priorities of the UK National Health Service are not specific to the maternity units participating in the DESiGN trial and sites implementing GAP outside the DESiGN trial also implemented incompletely.[3, 4] The pre-agreed minimum criteria for implementation in the DESiGN trial did not specify any level of reporting of outcomes, even though we appreciate this may be a feature subsequently introduced by Perinatal Institute to define complete GAP implementation. The response by Gardosi et al that only 3 out of 5 clusters achieved their reporting criterion for complete implementation in a longer timeframe than pre-specified in DESiGN emphasises these challenges. The results for full implementation of GAP at any time point, should only be compared to full implementation of standard care, as challenges to implementation are not specific to GAP. Our pragmatic trial describes what can be achieved with GAP in addition to standard care practice with current available resources within the National Health Service.

Standard care (control arm)

The control group in the DESiGN study was standard antenatal care. This is ‘fit for purpose’ and the gold-standard in RCTs such as ours where there is clinical equipoise and no appropriate placebo. We acknowledge that standard maternity care practice in the UK has evolved during the DESiGN trial (2015-2022, conception to publication), particularly with implementation of NHS England Saving Babies’ Lives Care Bundle in 2016. The pre-specified analysis plan accounted for this by focusing on the comparison between study arms in the outcome period for primary and secondary outcomes. Therefore, comparison of practice between pre-randomization and outcome period illustrate and contextualise the findings and cannot be used to assess the effect of GAP. Current standard care practice in the UK, as of 2022, does include Saving Babies’ Lives Care Bundle and comparison to practice without these recommendations would not be practical or ethical

Stillbirth

Our RCT was registered prior to starting the study (ISRCTN registry, ISRCTN67698474), and a protocol written and published[5]. Stillbirth is a rare outcome which the trial was underpowered to assess.[5] Furthermore, the statistical difference for stillbirth observed could be a false positive since we chose to make no allowance for multiple testing of secondary outcomes (we retained the ‘standard’ significance level p<0.05; and the observed p-value was p=0.03). Hence, our opinion is that the finding in relation to stillbirth in the GAP arm, in the context of multiple secondary outcomes and lack of effect on the primary outcome of SGA detection cannot be considered robust or causal.

Dharmintra Pasupathy, Matias C Vieira, Sophie Relph, Annette Briley, Kirstie Coxon, Deborah A Lawlor, Christoph Lees, Asma Khalil, Jane Sandall

1. 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 Med. 2022;19(6):e1004004.
2. Sovio U, White IR, Dacey A, Pasupathy D, Smith GCS. Screening for fetal growth restriction with universal third trimester ultrasonography in nulliparous women in the Pregnancy Outcome Prediction (POP) study: a prospective cohort study. Lancet. 2015;386(10008):2089-97.
3. Widdows K, Roberts SA, Camacho EM, Heazell AEP. Evaluation of the implementation of the Saving Babies’ Lives Care Bundle in early adopter NHS Trusts in England. Maternal and Fetal Health Research Centre, University of Manchester, Manchester, UK. 2018.
4. 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 Obstet Gynecol. 2021;57(3):401-8.
5. Vieira MC, Relph S, Copas A, Healey A, Coxon K, Alagna A, et al. The DESiGN trial (DEtection of Small for Gestational age Neonate), evaluating the effect of the Growth Assessment Protocol (GAP): study protocol for a randomised controlled trial. Trials. 2019;20(1):154.

No competing interests declared.

RE: RE: DESiGN Trial confirms benefit of GAP program: Perinatal Institute's response to the Trial team

jgardosi replied to dpasupathy623 on 25 Jul 2022 at 20:17 GMT

We thank Pasupathy and colleagues for their response to our critique of the DESiGN trial, although it failed to resolve the main points of contention. We will attempt to add clarity by a further examination of the evidence.

1. SGA detection rates

We are not aware of test positive rates in selective screening being as low as those observed in the Trial (Table 4), and the authors in their response have not quoted any references to back up their claim. Routine, high ascertainment recording of antenatal detection shows consistently much higher rates, as did those reported from GAP units in the SPIRE study [1]. In the GAP program the information is based on mandatory review of the antenatal record and data entry after each delivery, before the birthweight centile is calculated by the GROW software. For example, in Quarter 3 of 2018/19, which was during the DESiGN Trial’s ‘outcome period’, the national average SGA detection rate in GAP units was 42.4%, with a false positive rate of 7% (see Fig 2 in [2]), which represents a ‘test positive’ (EFW<10th centile) rate of 11.3%. This is three times higher than the 3.7% test positive rate reported in the intervention arm of the DESiGN Trial (Table 4).

If the low test positive rate is not explained by use of an EFW<3rd centile cut-off, then it is more likely to be due to poor data quality and incomplete ascertainment. The high rate of missing variables and extensive need for imputation has already been commented on with concern by reviewers in the Peer Review section of the publication. In one instance, ultrasound measurement data for a whole cluster’s baseline in the intervention arm was imputed, against reviewer’s advice [3].

The effect of missing data on detection rate is not known. In Appendix S3, the authors include a series of tables of ‘Available Case Analyses’ to assess missingness related to data presented in Table 2 (clinical and sociodemographic characteristics), Table 4 (screening performance) and Table 5 (secondary clinical outcome). Curiously, there is no analysis of missing data relating to Table 3 (utilisation of ultrasound services), other than that two clusters in the intervention group are excluded, without any explanation. In the description of the data collection method (‘Additional File 3’ in [4]), ‘number of ultrasound scans >24 weeks gestation’ and ‘SGA detected’ are included as variables in the within-cluster multiple imputation model, but no information is provided on accuracy and level of ascertainment. Missing scan reports, or missing information to obtain or calculate the EFW value within a report, can of course affect the accuracy of detection rates in either arm of the trial.

2. Challenges in implementation

Routine recording of outcome is an essential criterion for implementation and has no specified minimum (Table 1). For analysis we do require more that 75% cases recorded before the data are considered reliable, but the expectation is to record ALL births, as emphasised during training. Local monitoring of performance is an essential element of GAP, as it allows maternity services to identify hurdles and improve local referral pathways. In 2017, we reported that 65 of the 94 GAP units in the UK (69%) fulfilled this criterion [5]; today, it is 97 of 113 (86%) [6]. In the intervention arm of the DESiGN Trial, it was 0%.

3. Control

The control arm clearly did NOT maintain standard care. As events unfolded, the trial became a study of implementation of the Saving Babies Lives Care Bundle with and without the GAP program. The authors state that the purpose of collecting pre-randomisation data was to ‘contextualise’ the findings but this information is essentially ignored. Why did the randomisation process result in two substantially different summary clusters / cohorts, as evidenced in Table 2, and as also highlighted by peer reviewers? Why the sharp increase in 3rd trimester EFW scans in the control arm alone, and no increase in the intervention arm? In the end, ‘Cluster Randomised Controlled Trial’ seems a grandiose term for a study which had problems with each component of that title: the Clusters had missing data requiring multiple imputations; Randomisation resulted in substantially different cohorts; the Control arm, meant to represent standard care, had an unexplained increase in number of ultrasound scans, not mirrored in the intervention arm; and the Trial was too short to allow the intervention arm to achieve the minimal implementation criteria of the GAP program.

4. Reduction in stillbirths

The one remaining clear difference between the (non)-‘standard care’ and intervention arms was that customised charts were used in the latter. As we have shown above, the actual rate of SGA detection overall and in each arm is uncertain. However even if the rates were the same, the risk represented by the identified cases were not: Customised assessment improves the strength of association of SGA with adverse outcome when comparing fetal [7,8, 9] and neonatal standards including term [10, 11, 12, 13], and reduces SGA due to constitutional variation [14]. Therefore, while there is variable overlap, customised SGA identifies many additional cases that are at risk, and decreases false positives that have no increased risk. This results in greater confidence in the significance of the plotted measurement - a clinical benefit well known by users of customised charts – and translates here to earlier delivery of SGA babies and fewer stillbirths, in the GAP arm of the trial only (Tables 5 & 6).

Discrediting these significant findings because they did not happen via the presumed pathway of increased SGA detection ignores the above evidence of stronger association of customised SGA with adverse outcome. Discrediting it on the basis that stillbirth was only one of 26 secondary outcomes firstly ignores the reduction in gestational age and birthweight in the intervention arm as additionally significant and logically linked outcomes. Secondly, most other outcomes listed represent complications which would be unlikely to show differences, unless SGA and induction rates were substantially different between arms. Thirdly, there were no plans for correction for multiple testing in the pre-analysis protocol, and post hoc claims that results could be due to multiple testing should not be used to try to explain away a secondary outcome that does not correspond to the authors’ expectations.

5. Implication and relevance.

The DESiGN trial is being presented at conferences and in social media as ‘the first RCT’ and the headline that GAP has no effect. Its questionable results and interpretation need to be seen against solid evidence from observational studies of GAP implementation in various settings, which reported positive effects on SGA detection rate [15,16], stillbirth prevention [17,18] or both [1,5,19].

PLOS Medicine ought to be complimented for making the review process transparent, which allow readers to see the reservations of the reviewers, and how they were (or were not) addressed. Having supported the trial from the outset, we are not against the findings being published, but contest the interpretation chosen by the authors. We would like to conclude with a quote from the Academic Editor [20], which summarises what to us is the real (and only) value of the DESiGN Trial: “This was a very challenging study and unfortunately has limited implications for practice due to the significant logistical challenges. That said it includes some very important messages and articulates many of the challenges associated with complex intervention implementation, and there are important lessons for future studies in this area”.

Jason Gardosi, Emily Butler, Oliver Hugh


References

1. Widdows K, Roberts S, Camacho E, Heazell A. Evaluation of the implementation of the Saving Babies’ Lives Care Bundle in early adopter NHS Trusts in England (SPIRE). Manchester, UK: Maternal and Fetal Health Research Centre, University of Manchester; 2018. https://bit.ly/36rEqkd
2. Butler E, Hugh O, Gardosi J. Evaluating the Growth Assessment Protocol for stillbirth prevention: progress and challenges. Journal of Perinatal Medicine. 2022;50(6). https://doi.org/10.1515/j...
3. DESiGN Trial Peer Review History – Reviewer No 4 and Author response 20 Nov 2021 https://journals.plos.org...
4. Relph S, Elstad M, Coker B, Vieira MC, Moitt N, Gutierrez WM, et al. Using electronic patient records to assess the effect of a complex antenatal intervention in a cluster randomised controlled trial—data management experience from the DESiGN Trial team. Trials. 2021;22(1):195. https://doi.org/10.1186/s...
5. 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 Obstet Gynecol. 2021;57(3):401–8. https://doi.org/10.1002/u...
6. Perinatal Institute. Growth Assessment Protocol - Uptake. 2022. http://www.perinatal.org....
7. 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...
8. Francis A, Hugh O, Gardosi J. Stillbirth risk and small-for-gestational-age rate in subgroups according to maternal size: comparison of GROW, WHO and IG21 fetal growth standards. Ultrasound Obstet Gynecol. 2020;56(S1):19–19. https://doi.org/10.1002/u...
9. Giddings S, Clifford S, Madurasinghe V, Gardosi J. PFM.69 Customised vs uncustomised ultrasound charts in the assessment of perinatal mortality risk in the South Asian maternity population. Archives of Disease in Childhood - Fetal and Neonatal Edition. 2014;99(Suppl 1):A104. http://dx.doi.org/10.1136...
10. McCowan LM, Harding JE, Stewart AW. Customised birthweight centiles predict SGA pregnancies with perinatal morbidity. Br J Obstet Gynaecol. 2005;112(8):1026–33. https://doi.org/10.1111/j...
11. Figueras F, Figueras J, Meler E, Eixarch E, Coll O, Gratacos E, et al. Customised birthweight standards accurately predict perinatal morbidity. Archives Dis Child - Fetal and Neonatal Edition. 2007;92(4):F277–80. http://dx.doi.org/10.1136...
12. Francis A, Hugh O, Gardosi J. Customized vs INTERGROWTH-21st standards for the assessment of birthweight and stillbirth risk at term. Am J Obstet Gynecol. 2018;218(2):S692–9. https://doi.org/10.1016/j...
13. Fay E, Hugh O, Francis A, Katz R, Sitcov K, Souter V, Gardosi J. Customized GROW vs INTERGROWTH-21st birthweight standards to identify small for gestational age associated perinatal outcomes at term. Am J Obstet Gynecol MFM. 2022;4(2):100545. https://doi.org/10.1016/j...
14. Mongelli M, Gardosi J. Reduction of false-positive diagnosis of fetal growth restriction by application of customized fetal growth standards. Obstet Gynecol. 1996;88(5):844–8. https://doi.org/10.1016/0...(96)00285-2
15. Jayawardena L, Sheehan P. Introduction of a customised growth chart protocol increased detection of small for gestational age pregnancies in a tertiary Melbourne hospital. Aust N Z J Obstet Gynaecol. 2019;59(4):493–500. https://doi.org/10.1111/a...
16. Cowan FJ, McKinlay CJD, Taylor RS, Wilson J, McAra‐Couper J, Garrett N, et al. Detection of small for gestational age babies and perinatal outcomes following implementation of the Growth Assessment Protocol at a New Zealand tertiary facility: An observational intervention study. Aust N Z J Obstet Gynaecol. 2020 Dec 19;ajo.13283. https://doi.org/10.1111/a...
17. Gardosi J, Giddings S, Clifford S, Wood L, Francis A. Association between reduced stillbirth rates in England and regional uptake of accreditation training in customised fetal growth assessment. BMJOpen. 2013;3(12):e003942. http://dx.doi.org/10.1136...
18. Gardosi J, Turner S, Williams M, Buller S, Hugh O, Francis A. The Growth Assessment Protocol: a major cause of the declining stillbirth rates in the UK. Ultrasound Obstet Gynecol 2020; https://doi.org/10.1002/u...
19. Ravula PC, Veluganti S, Gopireddy MMR, Aziz N. Impact of introduction of the growth assessment protocol in a South Indian tertiary hospital on SGA detection, stillbirth rate and neonatal outcome. J Perinatal Med. 2022; 50(6). https://doi.org/10.1515/j...
20. DESiGN Trial Peer Review History: Academic Editor - Decision Letter 27 Oct 2021 https://journals.plos.org...

Competing interests declared: The authors work for the Perinatal Institute, a not-for-profit social enterprise which provides the GAP program and customised GROW charts.