White Coat Hypertension in Pregnancy: The challenge of combining inconsistent data

In this issue of Hypertension, Professor Khalil and his colleagues¹ took on the daunting task of assaying the available literature to determine the implications of white coat hypertension (WCH) during pregnancy. The task was daunting because of the remarkably incomplete and inconsistent quality of the information they tried to assess. Nonetheless, three important conclusions come from the study. First, accepting the limitations of the available data, the findings largely support the conventional wisdom that WCH may predict preeclampsia and gestational hypertension in later pregnancy and secondly, that the plan of management in the guidelines of the International Society for the Study of Hypertension (ISSHP)² is, with small modifications, a sensible strategy. Third, they indicate the important fact, also illustrated by many other studies, that in order to share data and take advantage of combining many studies “something must be done” to standardize investigations.

White coat hypertension is blood pressure elevation in the presence of care providers that is not present in home settings. In pregnancy, the ISSHP recommends that WCH only be considered as a diagnosis in women before 20 weeks of gestation.² Blood pressure elevation present only with care providers in later pregnancy is considered pregnancy-specific. This discrimination was not adhered to in many of the studies reviewed for this manuscript. Only preeclampsia as an outcome could be compared for WCH before and after 20 weeks. Data on other outcomes were not sufficient for this comparison. Women with WCH before 20 weeks’ showed a 5 fold greater risk of preeclampsia than normotensive women. The risk of preeclampsia was not significantly greater than the risk in normotensive women when WCH was present at any time during pregnancy. In the majority of available data based upon the diagnosis of WCH at any time in pregnancy, these women had a 2–3 fold increased risk of preterm birth and small for gestational age infants compared to normotensive women. These results require a number of cautions but specifically could have been different had they included only women with the diagnosis of WCH before 20 weeks. This implication of WCH prior to 20 weeks would be valuable to guide clinical management since women with transient increases in blood pressure after 20 weeks are already followed closely. Also, It is not possible to determine from the data presented (and likely from the data assayed by the authors) whether the increased risk of adverse outcomes was limited to women with WCH who later developed preeclampsia or whether the risk was present in women with only WCH. Importantly, women with WCH had a lower risk of several adverse outcomes than women with chronic hypertension or gestational hypertension. Thus, although the data supports the conventional wisdom that WCH is associated with an increased risk of preeclampsia it also suggests, contrary to common knowledge, that WCH is associated with adverse perinatal outcomes such as an increase in preterm birth and small for gestational age infants.

Current ISSHP recommendations² are illustrated in Figure 1 with a few modifications (in italics). ISSHP recommends that women with an in-office blood pressure of ≥ 140/90 mm. Hg. detected before 20 weeks’ gestation undergo ambulatory blood pressure monitoring, frequent automated blood pressure measurements during wake and sleep for a 24 hour period. Although ambulatory blood pressure is preferred,³ automated home blood pressure is an acceptable alternative if formal ambulatory blood pressure measurements are not available. Home blood pressure measurements should be obtained twice daily, morning and evening, and averaged over a 7-day period. An average ambulatory or home blood pressure of ≤ 130/80 mm. Hg. when awake (and 115/70 mm. Hg. when sleeping) is normal and the patient then is judged to have WCH. If the waking blood pressure is ≥ 130/80 mm. Hg. or sleeping ≥ 115/70 mm. Hg. the diagnosis is chronic hypertension. In the women with WCH, antihypertensive medication management is not recommended but rather, close monitoring with home blood pressure measurements throughout pregnancy and increased surveillance for the development of preeclampsia or gestational hypertension. The ISSHP does not suggest follow-up of the WCH pregnancy for perinatal outcomes and, more specifically, does not address monitoring for the detection of small for gestational age infants or preterm births in the absence of preeclampsia. Although monitoring for these adverse perinatal outcomes is suggested by the authors, we do not believe the data presented is sufficient to justify this. Also lacking is information on potential benefit of preeclampsia prevention strategies, such as low dose aspirin, for women diagnosed with WCH.

Figure 1.

Diagnosis and follow up of white coat hypertension (WCH): Women with elevated clinic blood pressure prior to 20 weeks’ gestation are diagnosed by ambulatory (ABPM) or home blood pressure (HBPM) monitoring and followed with HBPM. The diagnosis of preeclampsia (PE) mandates delivery or fetal assessment, chronic hypertension may indicate the need for fetal assessment^2,9, while current data do not support fetal assessment with only WCH. (GH = gestational hypertension, italics = modifications from original). Modified from ISSHP recommendations.²

The authors admit the extreme limitations of the data they were able to acquire. They began with 638 studies but finally only included 12 because of the limitations of the excluded studies. Even in the 12 studies considered, in addition to the variation in times during gestation at which the diagnosis of WCH was assessed, the definitions of WCH and preeclampsia and investigated outcomes also varied. Most studies did not include the protocol for ambulatory blood pressure determinations, quality control measures or masking of findings. Using the Newcastle-Ottawa¹ measure of study quality in which 9 is best and 1 is worst, only 3 studies were rated better than 6. The authors point out that the data that they assembled was of low quality by the GRADE scoring criteria of synthesized data due to bias and imprecision.

This is unfortunately the rule (not the exception) in attempts to merge data from several studies to gain insight into physiology, pathophysiology and clinical problems. At a time when the capacity for analysis of large datasets is more powerful than ever before⁴ the quality, heterogeneity and incompatibility of available data strongly limit the ability to extend findings beyond at most a very few studies. What can be done to facilitate data sharing? The first step is mindset. Very few investigators in the past have performed studies with the consideration that the value of their findings would be increased if they could be shared. Further, institutions and funding agencies have not rewarded sharing data and in fact, being one author of many in a large study has been considered of minimal value.⁴ There are, however, steps that are being taken to address these issues. It is becoming increasingly evident that sharing data demands the collection of appropriate data, standardized outcomes and is greatly facilitated by harmonized datasets.⁴ For pregnancy hypertension some of these topics have been addressed and some are in progress. A recent standardized set of data fields,⁵ standardized biological sample collection strategies⁶ and harmonized data base⁷ have become available and efforts are well advanced to standardizing outcomes and the definition of these outcomes⁸ that should be included in such studies. One can hope that these approaches become part of the mind set of investigators that are encouraged by funding agencies, universities and publishers and come to be an integral component of study design. With this approach studies will be able to combine data from more than 12 out of almost 700 studies and economically accelerate our understanding of physiology, pathophysiology and disease management.