BMJ Evidence-Based Medicine | 2019
Placental growth factor testing to assess women with suspected pre-eclampsia
Abstract
Preeclampsia occurs in 2%–8% of pregnancies, and is a significant contributor to maternal morbidity and mortality; however, the exact causes are unclear. Current diagnosis is based on clinical features including hypertension and proteinuria, requiring regular monitoring. Low circulating concentration of placental growth factor (PlGF), an angiogenic factor, has been shown to have a high sensitivity and negative predictive value for preeclampsia, suggesting PlGF could be a useful test in women with suspected preeclampsia. The steppedwedge cluster randomised controlled trial aimed to determine whether knowledge of the circulating concentration of PlGF, integrated with a clinical management algorithm, decreased the time for clinicians to make a diagnosis in suspected preeclampsia, and whether this approach reduced subsequent maternal or perinatal adverse outcomes. The trial was performed in 11 UK maternity units and included 1023 women aged 18 years and older with singleton pregnancy presenting with suspected preeclampsia at 20–36 weeks gestation. At the start of the trial all participants had PlGF measurements, which were concealed from clinicians and women. PlGF measurements were revealed at the initiation time of the intervention for each successive block and used alongside usual care, which consisted of a clinical management algorithm based on the National Institute for Health and Care Excellence guidelines. In the comparator group PlGF tests were concealed. The primary outcome was time from presentation with suspected preeclampsia to documented preeclampsia. The median time to preeclampsia diagnosis was ~2 days shorter in the revealed testing group compared with concealed testing (1.9 days (IQR 0.5–9.2) vs 4.1 days (IQR 0.8–14.7)), respectively. Use of PlGF testing was also associated with fewer maternal severe adverse outcomes, defined as maternal mortality or one or more serious central nervous system, cardiorespiratory, hepatic, renal or haematological morbidity (22 (4%) vs 24 (5%) events; adjusted OR 0.32, 95% CI 0.11 to 0.96; p=0.043). There were five serious events (two eclamptic fits, two strokes and one cardiac arrest in four women, all of whom had low PlGF concentrations) in the concealed testing group, with no similar serious events in the revealed testing group. A higher proportion of women in the concealed testing group than the revealed testing group received transfusion of blood products (14 (3%) vs 9 (2%)). Of the several secondary outcomes assessed, no differences were observed in gestational age at delivery, perinatal adverse outcomes, prevalence of spontaneous vaginal delivery or the use of caesarean sections. However, more women in the intervention group received a scan (77% vs 69%). The use of a stepwedge cluster randomisation in blocks and intentiontotreat analysis strengthens the analysis, although due to the nature of the intervention it was not masked to clinicians. The stepwedge design may have introduced some selection bias due to lower recruitment in the early months relative to later months. Trials to assess the effects of diagnostic tests on clinical care and patient outcomes are uncommon and as a consequence many tests are implemented in practice without robust evidence. Assessing the test in a realworld setting and assessing the effect on the clinical pathway and health outcomes strengthens the generalisability of the results. Based on this research, placental growth factor blood tests for preeclampsia are to be made more widely available in the UK. It should, however, be noted that the trial did not include women with multiple fetuses nor those presenting with lateonset preeclampsia. The study was conducted in a highresource setting and findings may not be applicable to other settings.