'Flash' observational study demonstrates feasibility and quality of national standardised mid-trimester ultrasound protocol
A significant proportion of morphological anomalies can be detected by a detailed ultrasound scan of foetal anatomy at 20+0 to 24+6weeks, also known as the mid-trimester scan. Several studies have shown that the adoption of a systematic scan with a standardised protocol is associated with improved detection rate of anomalies. The quality of stored images reflects the overall quality of foetal anatomical assessment, and regular audit procedures are recognised as a mean of maintaining and improving practice quality.
In France commencing in 2005, the Comité Technique d'Echographie (CTE) and then the Conférence Nationale d'Echographie Obstétricale et Fœtale (CNEOF) have issued recommendations on ultrasound protocols for the principle US scans during pregnancy in each of the three trimesters. These recommendations have been accepted as the professional standard to be adhered to and are referred to in cases of malpractice litigation. The CNEOF 2022 mid-trimester protocol includes a total of 26 standardised views, 23 recommended and 3 additional views. As the views of limbs are often grouped into one image, this results in a total of 24 views, of which 21 are recommended. If the recommended views are not obtained, the additional views can be produced to complement the recommended views that were obtained. The choice of these standardised views was justified by their a priori feasibility in current practice, their ability to illustrate items detailed in the text of the ultrasound report, and to include anatomical structures affected by the most common and severe foetal morphological anomalies. These standardised views are illustrated in the form of ‘silhouettes’ and are deliberately schematic to avoid subjective interpretation of a ‘typical’ ultrasound image.
The aims of this ‘flash’ study were (i) to evaluate this new standardised protocol by assessing the feasibility of performing the selected mid-trimester views in routine practice, (ii) to assess the quality of these images by evaluating the presence of anatomical landmarks and conformity criteria and (iii) to analyse the reliability between self-assessment and peer-assessment of the images in order to prepare for the quality-control programme that should soon accompany the national implementation of these new recommendations in France.
A consensus-based QA scoring system comprising 73 conformity criteria was established with 28 experts using a 3- round Delphi method. Secondly, authors asked operators to record 5 consecutive routine mid-trimester scans. Images were analysed by the sonographer themselves and by a qualified expert according to the scoring system. The frequency of recorded images was calculated for each of the views. Factors associated with missing images per scan were evaluated. The robustness of conformity criteria was assessed by reliability between self-evaluation and peer-evaluation. Main outcome measures: Based on 9849 images, we observed feasibility of the 21 recommended standardised views for midtrimester scan ranging from 88.5% to 100%.
Most conformity criteria (64/73, 88%) were met in over 90% of cases. Gwet's AC1 correlation between expert evaluation (peer-evaluation) and participant evaluation (self-evaluation) was greater than 0.80 for 70/73 (96%) criteria.
Results of this large population-based study demonstrate the successful implementation of a recently developed national ultrasound protocol for the mid-trimester scan. This large-scale audit study is essential, as it enables implementing an objective quality control and practice improvement approach based on self- and peer-evaluation, which is essential for maintaining the quality of screening.
Training artificial intelligence (AI) tools using previously annotated images based on our consensual scoring system, may increase the efficiency of usually time and resource consuming large-scale audits. As technology and AI advance, real-time auditing could notify the operator of incomplete or low-quality studies providing direct and immediate benefit to the quality of care delivered. Development and adoption of standard imaging protocols, and importantly associated quality assessments tools, can reduce the inherent operator dependence of ultrasound and reduce ‘noise’ in obstetrical screening by introducing greater decision hygiene and hopefully reducing errors. They can also, in the era of ‘value-based care’, allow funders to proactively focus funding on higher quality pregnancy care and imaging by better aligning payment and measurable practice quality.
Source: Thierry Bultez, Laurent Julien Salomon, Houman Mahallati; BJOG: An International Journal of Obstetrics & Gynaecology, 2025; 0:1–9 https://doi.org/10.1111/1471-0528.18102
