The overall intraclass correlation coefficients ranged from 0.881 to 0.996, indicating good consistency between researchers‘ and methodologists’ assessments.

Of the 10 guidelines included, two^14,22 are relevant to multiple pregnancies (including twins) combined with FGR, and eight^15–21,23 provid recommendations for FGR in singleton pregnancies. Four guidelines, SMFM, RCOG, FIGO, and SOGC, state that low-molecular-weight heparin is not recommended for the prevention in women at high risk of FGR. The RCOG, FIGO, SOGC, and CNGOF guidelines suggest that low-dose aspirin can be used for women at risk of pre-eclampsia and/or placental dysfunction to reduce the risk of FGR, and recommend starting aspirin (100–150 mg/d) at 12–16 weeks of gestation. ISUGO, RCOG, CNGOF suggest that FGR should be suspected when there is reduced and static fetal growth, which can be assessed with the help of Abdominal Circumference (AC) and Estimated Fetal Weight (EFW). RCOG, FIGO, ACR, CNGOF suggest that for those who are at high risk of FGR, ultrasound Doppler evaluation should be performed, especially Umbilical Artery, Uterine Artery, and Middle Cerebral Artery Doppler.

Almost all guidelines recommend that patients with a diagnosis of FGR should undergo procedural testing based on the severity of FGR and umbilical artery changes, including cardiotocography (CTG) or computerized CTG (cCTG), umbilical artery Doppler, middle cerebral artery Doppler, and intravenous catheter Doppler,. FIGO guideline strongly suggests that for late-onset FGR, middle cerebral artery Doppler may provide additional fetal deterioration information, but the SMFM guidelines advise against the use of venous catheterisation, middle cerebral artery or uterine artery Doppler assessment for routine clinical management of early-onset or late-onset FGR. SMFM, RCOG, FIGO, SOGC state that prenatal diagnosis is recommended for pregnant women with isolated premature or severe (EFW < 3rd percentile) FGR. Delivery is currently the only definitive treatment, timing of termination should be individualized as appropriate and FGR alone is not an indication for caesarean section. For cCTG showing fetal heart rate variability (STV), altered venous conduit, umbilical artery agenesis or reversed end-diastolic flow velocities, caesarean section is recommended for delivery and should be performed according to week of gestation after the use of glucocorticoids and magnesium sulphate.

FGR is strongly associated with adverse outcomes for maternal and infant outcomes. CPGs can help doctors and patients make clinical decisions for different situations, but their quality varies. Recommendations from high-quality guidelines are usually more reliable and can guide clinical practice to a higher degree. In this study, FGR guidelines were evaluated using the AGREE II, STAR, and RIGHT tools, a total of 10 guidelines were included after a comprehensive search and rigorous screening process.

The results of the AGREE II Instrument showed that of the 10 CPGs, two of them^17–18 were “recommended”, and the eight^{14–16, 19–23} were “suggested with modification”. Guidelines classified as “recommended” are generally considered to be of high quality. “Recommended with modifications” means that the openness and transparency of the guideline development process needs to be improved, and could be improved when the guideline is revised and updated. “Not recommended” means that the quality of the guideline needs to be improved. A previous study²⁴ evaluated quality differences in FGR guidelines using the AGREE II tool, and its findings were consistent with those of the present study, with both the RCOG and FIGO guidelines categorized as “recommended”. When analysised from the perspective of the domains, the “scope and purpose” and “clarity” domains were of high quality, with average scores of over 60%, while the remaining domains all scored lower, and the “editorial independence” domain being the most polarised. The “scope and purpose” domain scored above 60% for each guideline, with a relative weakness being the lack of clarity in defining the target population to which the guidelines apply. The “stakeholder involvement” domain performed poorly overall, with nine guidelines^{14–16, 18–23} achieving a standardized score below 50.00%. The main shortcomings included: (1) failure to incorporate the perspectives and preferences of the target population (e.g., patients, the public), and (2) insufficient description of the intended users of the guidelines and how they should be utilized. The “rigor of development” domain, which focuses on the methodological aspects of guideline development and is closely tied to guideline quality, received lower scores in six guidelines ^{14–16, 18, 22–23}. This was primarily due to: (1) the absence of clearly defined criteria for evidence retrieval, selection, and exclusion; (2) the lack of a specified method for formulating recommendations (e.g., voting, consensus, Delphi method); (3) no indication of external review prior to publication; and (4) insufficient details regarding updates (e.g., timing, criteria).The “clarity of presentation” domain is central to the guidelines and, consequently, achieved the highest average score. The “applicability” domain also scored lower with eight guidelines^{14–16,19−23} scoring < 50% on a standardized basis, mainly due to a lack of description of factors affecting the implementation of the guideline and a lack of consideration of the economic aspects of application. The “editorial independence” domain exhibited the most variability, with scores ranging from 0 to 100. Notably, the ISUOG and CNGOF guidelines failed to disclose sponsorship details and conflicts of interest.

The results of the STAR tool evaluation showed that the average score rates for the three domains of evidence, recommendations, and other aspects were high, all exceeding 60%, with the recommendations domain achieving the highest score rate. All of the guidelines included in this study scored 0 for “registry” and “protocol”, with the main problems being: (1) fail to register or provide a platform prior to publication of the guideline; and (2) fail to write a protocol, which suggests that published guidelines do not pay much attention to registration and writing protocols, a problem that could reduce the overall reliability of the guidelines. These indicate that the published guidelines do not pay much attention to registration and writing protocols, which will reduce the reliability of the guidelines as a whole. “funding”, “guideline development groups”, “consensus method”, and “accessibility” also scored poorly, with none exceeding 30%. The problems in each area were (1) funding sources and roles were not indicated; (2) participants were not grouped (including evidence and consensus groups); (3) recommendation consensus methods and records were not provided; and (4) different user versions of the guidelines were not provided. In terms of the overall quality of the guidelines, the RCOG guidelines had the highest score, which is consistent with the AGREE II tool results.

The results of the RIGHT checklist showed that two guidelines had an overall reporting rate of more than 80%, one of which was from the RCOG, and the results were also consistent with the AGREE II Instrument and the STAR tool. The average reporting rate was above 60% for six among the seven domains, with the only domain, “funding and declaration and management of interests”, having a low average reporting rate, mainly due to: (1) not reporting of sources of funding and roles of the various phases of the guidelines; and (2) not reporting on the evaluation and management of conflicts of interest.

This study presents the results of ten guideline evaluations in the form of a bubble chart (Fig. 2). The closer the guideline is to the upper right corner of the axis and the larger the bubble, the higher the quality of the guideline. The results showed that RCOG had the highest quality and PTGiP had the lowest quality. In this study, each researcher was required to assess a guideline using all three tools simultaneously, minimizing potential discrepancies in understanding the same guideline at different times and thus enhancing comparability. The trend of the results of the three tools is basically similar, with high quality and lower quality presenting consistent results regardless of the tool used, while medium quality articles will show differences, possibly due to the following reasons: (1) the evaluation forms of the three tools are different, and (2) the evaluation content is not entirely consistent. AGREE II is one of the most popular and established guideline assessment tools in use globally²⁵, covering the six domains of scope and purpose, stakeholder involvement, rigor of development, clarity of presentation, applicability, and editorial independence, with each item scored on a 7-point scale, and with the researcher's subjective perceptions influencing the results. Compared to AGREE II, the RIGHT checklist also requires the reporting of basic information in the guidelines. Each sub-item is assessed against the content of the guideline and the results are categorised only as “reported” and “not reported”, but there is no quantitative assessment of the content of the report. The RIGHT checklist is mainly used to guide developers in reporting guidelines, but can also be used to assess the reporting quality of guidelines. The STAR evaluation tool is a newly developed guideline evaluation method in China, which increases the content of registry, protocol, accessibility, etc., and at the same time refines the areas of funding, evidence, recommendations, etc., which improves the narrower focus of most of the current evaluation tools on the quality of the methodology, the quality of the report, etc. STAR can be regarded as an alternative to the comprehensive use of multiple assessment tools. At the same time, STAR assigns weights to domains and projects, which allows for better evaluation and ranking of guideline quality, but at the same time, its score is also affected by the weights.

Treatment of FGR is based on early diagnosis, optimal fetal monitoring and timely delivery. This study summarized the common recommendations of the 10 included guidelines, including the prevention, screening, management and treatment of FGR panel, and most of the guidelines agreed on CTG and ultrasound Doppler management of pregnant women with FGR, the timing of delivery, and what kind of cesarean section to be performed, and were supported by relatively sufficient evidence. In addition, the guidelines recommend women quite smoking and alcohol to reduce the risk of FGR. Early FGR should be monitored and managed in a tertiary ward with the highest level of neonatal care. Individual guidelines give opposing opinions on whether to use middle cerebral artery Doppler and venous catheter Doppler for prediction and management, but the strength of the recommendations is not high. There is no international consensus on the best way to manage FGR, and more research could be conducted on this component in the future, with a view to reducing perinatal mortality and improving prognosis.

This study has some limitations:(1) AGREE II, STAR, RIGHT are assessment tools only and cannot guide the assurance of clinical validity or usefulness, and clinical judgement should always be exercised when applying guideline recommendations in individual patient care. (2) Reporting bias may exist: the researcher's subjective awareness may potentially affect the evaluation results of the guidelines. This study adopted double-blind assessment to reduce bias. (3) Restricted languages and the small number of FGR guidelines show issues that do not fully represent the current international situation.

The quality of FGR guidelines still needs to be further improved. It is recommended that future guideline development and updating should pay attention to clarifying the target populations and settings of guideline use. Guideline authors should involve experts in statistics and guideline methodology to enhance the quality of evidence supporting the guidelines. Additionally, the influencing factors during guideline dissemination to clinical practice, as well as the practical implementation of guidelines in clinical settings, should be considered more comprehensively. Furthermore, the independence of guideline development, including the management of conflicts of interest and funding issues, also requires improvement.

Not Applicable

Availability of data and material