We thank Professor Wensink for his letter to the editor regarding our article investigating the risk of stillbirth in patients with intrahepatic cholestasis of pregnancy (ICP). Based on our retrospective cohort study, we concluded that ICP was associated with an increased risk of stillbirth as compared to pregnancies without ICP. Furthermore, we found that delivery at 36 weeks’ gestation would minimize overall perinatal mortality. Professor Wensink’s main objections to our study were in regards to the validity of our composite mortality calculation, the fact that we claim that there is a lack of evidence regarding specific bile acid levels and correspondent fetal outcomes, and lastly, the turnaround time of bile acid results.

Our article focused on the clinical question of when to deliver patients with cholestasis, given that at each gestational week one could choose delivery or expectant management. To summarize, we used a well-established life-table analysis method to establish stillbirth risk. Infant death risk was calculated by dividing the number of infant deaths at a given week of gestation by the number of live births at the same week of gestation. These 2 prior calculations were combined to estimate the mortality risk of expectant management, which takes into account the risk of stillbirth during the week the patient remains pregnant, in addition to the risk of infant death if delivery occurs at the subsequent week of gestation.

Our data remained statistically significant when we controlled for confounding variables including maternal age, race, comorbid conditions (diabetes, hypertension), nulliparous status, and limited prenatal care. Despite drawing our ICP data from a large cohort, the mortality outcomes were so rare that our study may be prone to the same instability faced by smaller studies.

There is no definitive evidence yet that stratifies specific bile acid levels to stillbirth risk. Furthermore, there is no definitive evidence that lower bile acid levels significantly reduce or eliminate stillbirth risk. We believe that larger studies are indeed needed to examine this relationship, as our data were unable to do this. Lastly, while the author has access to bile acid levels within the hour, the majority of hospitals and academic centers in the United States do not have this capability. Rural or resource-poor clinical settings have even more difficulty with this, which may lead to delayed diagnoses.

Ultimately, a large, multicenter, prospective cohort study examining the relationship between bile acids and perinatal mortality with the goal of confirming optimal delivery timing is needed. However, it is unlikely that an adequately powered study as described is forthcoming given that ICP is a rare diagnosis and its mortality outcomes are even rarer. This is why our data, despite their limitations, can be used to help clinicians counsel patients with ICP.