OBJECTIVE: To review systematically the literature on diagnostic tests and performance of second-trimester sonographic assessment of nasal bone (NB) in identifying fetuses affected by Down syndrome. METHODS: A search of studies involving screening tests for NB evaluation and measurements was carried out in the main international bibliographic databases (MEDLINE, EMBASE and CINAHL). Those considered to be relevant were then subjected to critical reading, following Critical Appraisal Skills Programme (CASP) criteria, by at least three independent observers. All data were extracted and tabulated by two independent investigators. A statistical synthesis of sensitivity, specificity and likelihood ratios was performed using specific software (Meta-DiSc). RESULTS: From an initial list of 852 articles referring to ultrasound markers for Down syndrome, 207 relevant papers were selected. Following exclusions, 21 studies were included in the quantitative synthesis. The pooled estimates of positive and negative likelihood ratios were 40.08 (95% CI, 18.10-88.76) and 0.71 (95% CI, 0.64-0.79), respectively, for absent NB and 15.15 (95% CI, 8.15-28.16) and 0.47 (95% CI, 0.34-0.64), respectively, for hypoplastic NB. No relevant differences were found between the various means of defining nasal hypoplasia (multiples of the median (MoM) or percentiles). The biparietal diameter/nasal bone length (BPD/NBL) ratio showed somewhat higher sensitivity but lower specificity with a threshold effect. CONCLUSIONS: NB absence or hypoplasia show high specificity and low but acceptable sensitivity in identifying fetuses with Down syndrome. Screening performance is better with NB measurements as a function of MoM or percentiles rather than as the BPD/NBL ratio. Classification of women into various risk groups for Down syndrome does not affect diagnostic performance.
OBJECTIVE: To review systematically the literature on diagnostic tests and performance of second-trimester sonographic assessment of nasal bone (NB) in identifying fetuses affected by Down syndrome. METHODS: A search of studies involving screening tests for NB evaluation and measurements was carried out in the main international bibliographic databases (MEDLINE, EMBASE and CINAHL). Those considered to be relevant were then subjected to critical reading, following Critical Appraisal Skills Programme (CASP) criteria, by at least three independent observers. All data were extracted and tabulated by two independent investigators. A statistical synthesis of sensitivity, specificity and likelihood ratios was performed using specific software (Meta-DiSc). RESULTS: From an initial list of 852 articles referring to ultrasound markers for Down syndrome, 207 relevant papers were selected. Following exclusions, 21 studies were included in the quantitative synthesis. The pooled estimates of positive and negative likelihood ratios were 40.08 (95% CI, 18.10-88.76) and 0.71 (95% CI, 0.64-0.79), respectively, for absent NB and 15.15 (95% CI, 8.15-28.16) and 0.47 (95% CI, 0.34-0.64), respectively, for hypoplastic NB. No relevant differences were found between the various means of defining nasal hypoplasia (multiples of the median (MoM) or percentiles). The biparietal diameter/nasal bone length (BPD/NBL) ratio showed somewhat higher sensitivity but lower specificity with a threshold effect. CONCLUSIONS:NB absence or hypoplasia show high specificity and low but acceptable sensitivity in identifying fetuses with Down syndrome. Screening performance is better with NB measurements as a function of MoM or percentiles rather than as the BPD/NBL ratio. Classification of women into various risk groups for Down syndrome does not affect diagnostic performance.