|
Second-trimester ultrasound to detect fetuses with Down syndrome: a meta-analysis |
Smith-Bindman R, Hosmer W, Feldstein V A, Deeks J J, Goldberg J D |
|
|
Authors' objectives To determine the accuracy of second-trimester ultrasound in detecting Down syndrome in foetuses.
Searching MEDLINE was searched from January 1980 to February 1999. The search terms including the following: 'choroid plexus cyst', 'nuchal fold thickening', 'echogenic intracardiac focus', 'echogenic bowel', renal pyelectasis', 'shortened humerus', 'shortened femur', 'fetal structural malformations', 'down syndrome', 'trisomy 21', 'prenatal ultrasound', 'chromosomal abnormality', 'diagnostic tests', 'biochemical testing' and 'genetic ultrasound'. The bibliographies of relevant articles were screened.
Study selection Study designs of evaluations included in the reviewDiagnostic accuracy studies were included provided that the original ultrasound interpretation was used. Case studies were excluded, as were review articles, letters, case reports and comments. Studies were also excluded if they reported fewer than five foetuses with a chromosomal abnormality.
Specific interventions included in the reviewStudies were included that recorded second-trimester prenatal findings of ultrasonographic markers. Studies based on first-trimester ultrasound were excluded, as were studies that performed the ultrasound following genetic testing or were primarily focused on a different topic. The following ultrasonographic markers were used: choroid plexus cyst (cysts of any size or number in the cerebral ventricles), nuchal fold thickening (of at least 6 mm), echogenic intracardiac focus (punctate intracardiac echogenic focus within either ventricle), echogenic bowel (echogenicity equal to or greater than bone), renal pyelectasis (anterior-posterior diameter of the renal pelvis of at least 4 mm), shortened humerus, shortened femur, and foetal structural malformations.
Reference standard test against which the new test was comparedOnly studies where there was active ascertainment of all pregnancy outcomes by chromosomal analysis or visual inspection were included.
Participants included in the reviewPregnant women at second-trimester. The mean maternal age was 34 years. Some of the studies included women at increased risk of chromosomal abnormality based on age, serum biochemical testing, or family history of chromosomal abnormality. The overall prevalence of Down syndrome was 1.5%.
Outcomes assessed in the reviewOnly studies from which estimates of sensitivity and specificity could be calculated were included. The outcome measures reported were sensitivity, specificity, and positive and negative likelihood ratios (LRs). Studies were excluded if they obtained outcome data only on foetuses with a specific ultrasound finding or had incomplete follow-up. Three foetal outcomes were considered: unaffected, Down syndrome, or all chromosomal abnormalities when data for Down syndrome alone were unavailable.
How were decisions on the relevance of primary studies made?The studies were selected independently by two reviewers.
Assessment of study quality The authors did not state that they assessed validity.
Data extraction The data were extracted independently by two reviewers, with any discrepancies resolved by reaching consensus with a third author. The study date, maternal risk of chromosomal abnormality, and the presence and type of structural abnormalities were extracted. True-positive, false-positive, true-negative and false-negative results were extracted for each of the specified ultrasonographic markers. Where possible, the data were extracted separately for the markers seen as an isolated abnormality or in combination with foetal structural malformations. The sensitivity, specificity, positive LR, negative LR and 95% confidence intervals were calculated for each study for each of the ultrasound findings.
Methods of synthesis How were the studies combined?Pooled estimates of sensitivity and specificity were calculated for each ultrasound marker using a single-term logistic regression model; the unit of analysis was the study. The positive and negative LRs were calculated for each ultrasound finding, stratified by whether it was seen in isolation or in association with foetal structural malformations. These were then pooled using the random-effects method of DerSimonian and Laird (see Other Publications of Related Interest). Positive and negative predictive values were calculated at a disease prevalence of 1:700 and 1:300 by applying Bayes theorem to the estimated LRs.
The number of women needed to be screened to detect a case of Down syndrome was calculated as the inverse of the product of the prevalence of Down syndrome and the sensitivity of the ultrasound finding. The number of normal foetuses that would be lost for each case of Down syndrome identified was calculated based on the assumption that all screen-positive ultrasound examinations were followed by invasive diagnostic testing, and that the rate of foetal loss per amniocentesis procedure was 0.8%.
How were differences between studies investigated?The consistency of the study results was assessed using the 'goodness-of-fit' deviance statistic. Dispersion was calculated by dividing the deviance statistic by its df. Where dispersions were greater than 1, variability was greater than expected by chance. Where the fit was poor, the logistic regression model was adjusted to account for the overdispersion by multiplying the standard error by the square root of the dispersion statistic. To identify whether sensitivity and specificity estimates might be influenced by study design, sample size or by whether the markers were reported in isolation, terms for these factors were introduced into the logistic regression model. A sensitivity analysis was carried out to assess the impact of one study that contributed 25% of the unaffected foetuses to the meta-analysis.
Results of the review There were 56 studies (n=132,295) in total: 32 diagnostic cohort (n=116,861) and 24 diagnostic case-control studies (n=15,434).
When ultrasonographic markers were observed without associated foetal structural malformations or other markers, the sensitivity for each was low (range: 1 to 16%). The specificity for each of the markers was greater than 95% when the finding was seen as an isolated abnormality. A thickened nuchal fold was the most accurate marker and was associated with approximately a 17-fold increased risk of Down syndrome. The positive LRs for the other markers were lower. A normal finding did not substantially decrease the risk of a foetus having Down syndrome (none of the negative LRs were significant).
The results discussed were heterogeneous for all findings. Only results for nuchal fold are reported in further detail. The percentage of foetuses with Down syndrome correctly identified using a thickened nuchal fold varied from 7 to 75%. There were no differences in the sensitivity based on sample size. There was a significant difference in accuracy based on study design (p=0.008) and whether the marker was seen as an isolated abnormality or in association with foetal structural abnormalities (p<0.001).
Authors' conclusions A thickened nuchal fold in the second trimester may be useful in distinguishing unaffected foetuses from those with Down syndrome. However, the overall sensitivity of this finding is too low for it to be a practical screening test for Down syndrome. When observed without associated structural malformations, the remaining ultrasonographic markers could not discriminate well between unaffected foetuses and those with Down syndrome. Using thse markers as a basis for deciding to offer amniocentesis will result in more foetal losses than cases of Down syndrome detected, and will lead to a decrease in the prenatal detection of foetuses with Down syndrome.
CRD commentary The authors presented a reasonably well-defined review question. The extracted data was clearly tabulated for the included studies, though more information on the participants would have been helpful. However, the review had some limitations. It is possible that relevant studies were missed as the literature search was limited to one database and excluded non-English language papers, and no attempt was made to identify unpublished research. A validity assessment was not carried out, therefore the findings are not discussed in the context of quality. Also, the data were pooled despite evidence of heterogeneity between the studies. The authors' conclusions need to be treated with some caution considering the limited search, the lack of a quality assessment and the pooling of heterogeneous data.
Implications of the review for practice and research The authors did not state any implications for further research and practice.
Funding Supported in part by Mount Zion Health Care Systems; the University of California at San Francisco (UCSF) Academic Senate and the UCSF Research Education and Allocation Committee.
Bibliographic details Smith-Bindman R, Hosmer W, Feldstein V A, Deeks J J, Goldberg J D. Second-trimester ultrasound to detect fetuses with Down syndrome: a meta-analysis. JAMA 2001; 285(8): 1044-1055 Other publications of related interest DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177-88
Indexing Status Subject indexing assigned by NLM MeSH Down Syndrome /ultrasonography; Female; Fetus /anatomy & Humans; Predictive Value of Tests; Pregnancy; Pregnancy Trimester, Second; Sensitivity and Specificity; Ultrasonography, Prenatal; histology AccessionNumber 12002008159 Date bibliographic record published 30/09/2003 Date abstract record published 30/09/2003 Record Status This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn. |
|
|
|