Few moments are more emotionally charged than bleeding in early pregnancy. It’s a clinical grey area and a deeply human one too.
For many women, it brings fear, uncertainty, and the looming question: Am I losing my baby?
Medically referred to as a “threatened miscarriage,” this is a diagnosis given when bleeding occurs in the early stages of a pregnancy that is still viable. This is a common complication in the first trimester of pregnancy, affecting up to one in five pregnancies.
Yet despite its frequency, doctors often struggle to offer clear answers about what comes next.
A doctoral research study, currently being undertaken at the University of Malta, addresses this gap.
This research explores whether a combination of ultrasound and biochemical markers, together with clinical presentations and sociodemographic characteristics, could improve the prediction of pregnancy outcomes following first-trimester bleeding. Drawing on both retrospective data and prospective clinical research, the study is one of the most comprehensive analyses of its kind within the Maltese healthcare setting.
This research has so far examined over 700 cases of women who arrived at Mater Dei Hospital hospital with first-trimester bleeding. It found that only about a third of these pregnancies progressed beyond 22 weeks of pregnancy.
Unsurprisingly, age played a significant role, whereby women aged 35 and over faced a noticeably higher risk of pregnancy loss. Babies born after early bleeding were also more likely to be born early or weigh less than average.
But the study didn’t stop at statistics. It went deeper, extensively reviewing studies from around the world to identify early markers that may indicate whether a pregnancy is likely to continue.
Key ultrasound features, such as a slow fetal heartbeat or irregular yolk sac, and hormone levels in maternal blood were found to be powerful clues.
While these markers have been studied before, much of the previous research has been limited by heterogeneous methodologies.
To build something more reliable, the final phase of the study involved a prospective trial with nearly 180 women, both with and without early pregnancy bleeding.
With a rich dataset of scans, blood results, and clinical and demographic information, researchers developed predictive models combining traditional statistics with machine learning.
The models demonstrated high levels of accuracy, correctly predicting pregnancy outcomes in up to 93% of cases.
But beyond the numbers lies the heartbeat of this research. It’s in the waiting rooms, where uncertainty hangs heavy, in the silence before a scan begins, in the breath held while searching for signs of life, in the quiet strength of those who return after loss, still hoping.
In our national, state hospital, there is currently no dedicated early pregnancy unit. Women facing the possibility of miscarriage wait in shared spaces, sometimes alongside those visibly pregnant or cradling newborns. It is a setting that amplifies distress, and one that speaks volumes about the gaps in care.
Threatened miscarriage is often dismissed as a "wait-and-see" diagnosis.
This work argues, unequivocally, that "wait-and-see" is not good enough. We can do better. And with the right data, interpretation, and compassion, we must do better.
Special thanks to the PhD main supervisor, Prof. Jean Calleja Agius and the co-supervisors, Prof. Paul Bezzina and Prof. Vivien Gibbs.
