Use of Doppler ultrasonography to predict pre-eclampsia
=======================================================

* Agustín Conde-Agudelo, MD MPH
* Marshall Lindheimer MD

*   © 2008 Canadian Medical Association or its licensors

We enjoyed Jeltsje Cnossen and colleagues' systematic review of the use of uterine artery Doppler ultrasonography to predict pre-eclampsia.1 They concluded that an increased pulsatility index with notching during the second trimester is the best predictor of pre-eclampsia and strongly recommended the routine use of these measurement parameters in clinical practice. However, this recommendation is based on only 2 studies, one of which included 1757 low-risk women and the other 351 high-risk women. As the incidence of pre-eclampsia is relatively low (0.4%–6.7%), screening tests require high likelihood ratios to adequately predict the disease's probability with positive test results and very low likelihood ratios to confidently exclude the disorder with negative test results.2 An increased pulsatility index with notching produced sufficiently positive likelihood ratios (21.0) in high-risk women but it was inadequate in low-risk populations (7.5); importantly, the negative likelihood ratios were quite poor for both populations (0.59 and 0.82 respectively).

We also have methodologic concerns. First, a valid meta-analysis should be examined for heterogeneity before one considers pooling the results of primary studies to create summary estimates with enhanced precision.3 There is no indication in the review that the heterogeneity of the study results was formally tested. Second, there is a substantial possibility of publication bias in this area of research,4 and there is no indication that this was assessed. Finally, although pooling of sensitivities and specificities instead of likelihood ratios has recently been encouraged,5 we are skeptical and agree with others6 that sensitivities and specificities are inappropriate for meta-analyses as they do not behave independently when pooled from primary studies to generate separate averages.

We therefore suggest that the authors' conclusions are premature. Doppler ultrasonography, although useful for monitoring high-risk pregnancies, should not currently be recommended for routine screening to predict pre-eclampsia.

## Footnotes

*   **Competing interests:** None declared.

## REFERENCES

1.  1. Cnossen JS, Morris RK, ter Riet G, et al. Use of uterine artery Doppler ultrasonography to predict pre-eclampsia and intrauterine growth restriction: a systematic review and bivariable meta-analysis. CMAJ 2008;178:701-11.
    
    [Abstract/FREE Full Text](http://www.cmaj.ca/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiY21haiI7czo1OiJyZXNpZCI7czo5OiIxNzgvNi83MDEiO3M6NDoiYXRvbSI7czoyMToiL2NtYWovMTc5LzEvNTMuMS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

2.  2. Conde-Agudelo A, Villar J, Lindheimer M. World Health Organization systematic review of screening tests for preeclampsia. Obstet Gynecol 2004;104:1367-91.
    
    [CrossRef](http://www.cmaj.ca/lookup/external-ref?access_num=10.1097/01.AOG.0000147599.47713.5d&link_type=DOI) 
    
    [PubMed](http://www.cmaj.ca/lookup/external-ref?access_num=15572504&link_type=MED&atom=%2Fcmaj%2F179%2F1%2F53.1.atom) 
    
    [Web of Science](http://www.cmaj.ca/lookup/external-ref?access_num=000225683400027&link_type=ISI) 

3.  3. Lijmer JG, Bossuyt PM, Heisterkamp SH. Exploring sources of heterogeneity in systematic reviews of diagnostic tests. Stat Med 2002;21:1525-37.
    
    [CrossRef](http://www.cmaj.ca/lookup/external-ref?access_num=10.1002/sim.1185&link_type=DOI) 
    
    [PubMed](http://www.cmaj.ca/lookup/external-ref?access_num=12111918&link_type=MED&atom=%2Fcmaj%2F179%2F1%2F53.1.atom) 
    
    [Web of Science](http://www.cmaj.ca/lookup/external-ref?access_num=000176016900004&link_type=ISI) 

4.  4. Song F, Khan KS, Dinnes J, et al. Asymmetric funnel plots and publication bias in meta-analyses of diagnostic accuracy. Int J Epidemiol 2002;31:88-95.
    
    [Abstract/FREE Full Text](http://www.cmaj.ca/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoiaW50amVwaWQiO3M6NToicmVzaWQiO3M6NzoiMzEvMS84OCI7czo0OiJhdG9tIjtzOjIxOiIvY21hai8xNzkvMS81My4xLmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

5.  5. Zwinderman AH, Bossuyt PM. We should not pool diagnostic likelihood ratios in systematic reviews. Stat Med 2008;27:687-97.
    
    [CrossRef](http://www.cmaj.ca/lookup/external-ref?access_num=10.1002/sim.2992&link_type=DOI) 
    
    [PubMed](http://www.cmaj.ca/lookup/external-ref?access_num=17611957&link_type=MED&atom=%2Fcmaj%2F179%2F1%2F53.1.atom) 
    
    [Web of Science](http://www.cmaj.ca/lookup/external-ref?access_num=000253782500004&link_type=ISI) 

6.  6. Shapiro DE. Issues in combining independent estimates of the sensitivity and specificity of a diagnostic test. *Acad Radiol* 1995;2:S37-S47.