Re: Selective serotonin reuptake inhibitors during pregnancy and risk of persistent pulmonary hypertension in the newborn: population based cohort study from the five Nordic countries
This large Nordic study on use of selective serotonin reuptake inhibitors (SSRI) in pregnancy and risk of pulmonary hypertension (PPHN) in the new born showed that the risk of PPHN was only slightly increased after SSRI exposure in early pregnancy (unadjusted odds ratio 1.6; adjusted 1.4). In late pregnancy, the risk of PPHN was more than doublet (unadjusted odds ratio 2.5; adjusted 2.1). Timing of exposure was based on the date that SSRIs were dispensed at the pharmacy. Use in early pregnancy was defined as an SSRI dispensed in the time period from 90 days before the start of pregnancy until the eight gestational week of pregnancy. This is done to include women that were dispensed SSRI prior to pregnancy but who may have used the drug during the early stages of pregnancy.
The study was based on a linkage of population-based registries in the five Nordic countries. The Norwegian Prescription Database was one of the nationwide databases that the Nordic SSRI study was based on. In Norway we also have a large population based pregnancy cohort; the Norwegian Mother and Child Cohort (MoBa) including approximately 100 000 pregnancies. In MoBa data on drug use were self-reported two times during pregnancy and when the child was 6 month for the last part of pregnancy. Combining information from these two data sources on drug exposure has given us the opportunity to validate SSRI exposure data from the prescription database by assessing the comparability of these data with the self-reported data on drug use in pregnant women from MoBa.
The prevalence of SSRI use in the Nordic countries is relatively low. As pointed out by Greenlad specificity is more important than high sensitivity for the measured risk association when the prevalence of exposure is low. Thus it is important to minimize the number of truly unexposed SSRI users in the exposed group in studies of the safety of SSRIs in pregnancy. Women who were dispensed SSRI before pregnancy may be truly unexposed if they stop using the drug before pregnancy but may have been misclassified as exposed.
We have recently published a validity study based on the two data sources mentioned above. We have shown that expansion of the time windows for dispensed drugs in the prescription databases to include intervals 90 days before pregnancy led to higher sensitivity but lower specificity of drug exposure compared 60, 30 and 0 days expansion before pregnancy. Application of the observed sensitivities and specificities in our study on the data in the Nordic study demonstrated that the degree of underestimation of risk associations increased as the time windows was expanded before pregnancy. Including the 90 days’ time window for use in early pregnancy the calculated odds ratio corrected for misclassification was 2.6 compared with the reported OR by 1.6 in the study of Kieler at al. Similarly, OR by use of SSRI in late pregnancy was 2.5 in the Nordic study compared to the OR corrected for misclassification of 2.7 Based on these results it is possible that the study by Kieler at al has underestimated the risk of SSRI exposure in early pregnancy.
In the Summary of Product Characteristics (SPC) updated in November 2012 it is stated that epidemiological data suggests increased risk of PPHN, particularly emphasizing risk associated with exposure late in pregnancy. This is also mentioned in the British National Formulary and in corresponding references in other European countries. Taking into account that most women are exposed to SSRIs in early pregnancy it would be very interesting to have seen the results from additional analysis using a definition of exposure including fewer or no days prior to pregnancy.
1. Kieler H, Artama M, Engeland A, et al. Selective serotonin reuptake inhibitors during pregnancy and risk of persistent pulmonary hypertension in the newborn: population based cohort study from the five Nordic countries. BMJ 2012;344:d8012 doi: 10.1136/bmj.d8012
2. Greenland S. Basic methods for sensitivity analysis of biases. Int.J.Epidemiol. 1996;25(6):1107-16
3. Skurtveit S, Selmer R, Tverdal A, et al. Drug exposure: inclusion of dispensed drugs before pregnancy may lead to underestimation of risk associations. J.Clin.Epidemiol. 2013;66(9):964-72 doi: S0895-4356(13)00126-1
4. The electronic Medicines Compendium. Summary of Product Characteristics Cipramil Tablets. Secondary Summary of Product Characteristics Cipramil Tablets 19.11.2012 http://medicines.org.uk/emc/medicine/1070/SPC/Cipramil+Tablets/#PREGNANCY.
5. British National Formulary, BMJ Group, Royal Parmaceutical Socoiety of Great Britain. Selective Serotonin Re-uptake inhibitors. 2013. http://www.medicinescomplete.com/mc/bnf/current/PHP2413-selective-seroto....
Competing interests: No competing interests