Asymptomatic pregnant women returning to the United States from countries experiencing a Zika virus outbreak should be tested for Zika virus: AGAINST: Not so fast
International Journal of Obstetrics and Gynocology, First published: 6 May 2016, DOI: 10.1111/1471-0528.14068
George R. Saade
These articles are part of a collection of articles on the Zika virus, introduced by an Editorial by SS Witkin.
To view this Editorial visit http://dx.doi.org/10.1111/1471-0528.14076.
Updated guidance issued in response to the recent reports of an association between maternal Zika infection and fetal microcephaly included a statement that serological testing for Zika virus can [emphasis added] be offered to asymptomatic pregnant women who travelled to an area with ongoing Zika virus transmission (SMFM Publications Committee Am J Obstet Gynecol 2016;pii:S0002–9378[16]00343-4. doi: 10.1016/j.ajog.2016.02.043; Oduyebo et al. MMWR 2016;65:122–7).
Testing in asymptomatic women focuses on Zika virus serum immunoglobulin M. Interpretation of serology results is complex and subject to false positives and false negatives, particularly when it is used as a screen for maternal to fetal transmission, the ultimate reason for testing. The value of a screening test depends on its predictive performance and on the a priori risk in the target population, in this case asymptomatic pregnant women who travelled to an affected area. There are no data on the
- sensitivity,
- specificity,
- false positives or
- false negatives in such cases.
Moreover, the a priori risk of infection in an asymptomatic pregnant woman who travelled to an affected area, let alone the risk of having an affected fetus, is not known. This a priori risk is probably close to zero. As of 24 February 2016, no confirmed cases of maternal infection were identified among 162 asymptomatic pregnant women tested by the Centers for Disease Control and Prevention (CDC; 95% CI for maternal infection 0% to 2.2%) (Fleming-Dutra et al. MMWR 2016;65:182–7). All nine women with confirmed maternal infection, including four women with likely fetal infection, had a generalised rash, as well as other symptoms in eight of the women (Fleming-Dutra et al. MMWR 2016;65:182–7). In a more recent update released on April 15, 2016, the CDC reported that Zika infection was confirmed in 7 of 2425 pregnant women who were asymptomatic (maternal infection rate 0.3%; 95% CI 0.1% to 0.6%). Only 2 of the confirmed infections were in women who were short-term travellers. The remaining 5 were residing in affected areas at some time during their pregnancy (Dasgupta et al. MMWR 2016;65:395–99). So regardless of the test likelihood ratios, the post-test probability in asymptomatic women is likely to be zero and most positive tests are likely to be false positives. This highlights the unintended consequences and the potential to do more harm with testing. These unintended consequences include risks of amniocentesis, termination of a normal pregnancy and maternal anxiety.
The benefit of a screening test also depends on what is done next. As there is no treatment, the rationale for testing so far has been to assist in counselling women about fetal risks, whether to perform amniocentesis, and whether to continue serial ultrasound. Data on which to base counselling or decision-making for any of these downstream actions, other than the risk of pregnancy loss from amniocentesis, are not readily available. We do not know the rates of fetal infection with either a positive or negative serological test in an asymptomatic woman, and the rates of fetal abnormalities with either a positive or negative amniocentesis.
Some of the same organisations that recommend testing in this situation have rightly recommended against cell-free DNA testing in low-risk women (SMFM Publications Committee Am J Obstet Gynecol 2015;212:711–16). The rationale was that the performance of cell-free DNA in low-risk women is not known. We actually know a lot more about its performance than we know about Zika virus serological testing, and the outcome being screened for with cell-free DNA is as devastating as microcephaly. So why the difference? Could it be a reaction to the panic and widespread media frenzy? Until additional data become available, I suggest that we follow the original guidance from the CDC, which did not recommend serological testing in asymptomatic women (Petersen et al. MMWR 2016;65:30–3).
Vitamin D Life notes that in mid May the CDC started testing asymptomatic US pregnant women
Of US pregnant women who were tested for Zika, 279 were positive – how many are actually infected - May 2016
Zika – Should asymptomatic pregnant travelers be tracked (no) – May 2016