A review says decision-supported systems could help maternity teams spot worsening glucose patterns and respond faster, but warns that technology must fit clinical workflows and be tested for safety, equity and workload.
Gestational diabetes remains one of the most common metabolic complications of pregnancy, and its effects do not end at delivery. The review published in DovePress argues that women affected by the condition face higher risks of pregnancy complications and are more likely to develop type 2 diabetes and repeat glucose problems later in life. Yet, despite well-established screening and treatment guidance, the paper says everyday care is still difficult to deliver consistently because it demands frequent glucose monitoring, rapid changes to diet and exercise advice, and, in some cases, swift medication adjustments.
The authors describe a system that goes beyond simple digital logging. In their view, many current tools collect blood sugar readings or send reminders but do little to help teams interpret patterns, decide who needs attention first, or standardise escalation. That leaves clinicians still having to sift through large volumes of patient-generated data by hand, a task that can be especially demanding in busy antenatal services with limited specialist support.
Instead, the review sets out what it calls decision-supported collaborative care: a model in which patient readings, clinical context and guideline-based logic are brought together inside a shared workflow. The paper says such systems should combine structured glucose data with factors such as gestational age, weight change, medication use and meal timing, then route recommendations to the right member of the care team. Nurses, diabetes educators, dietitians, obstetricians and endocrinologists would each have defined responsibilities, with the aim of speeding up response times while preserving clinical judgement.
A key point in the review is that intelligence in this setting does not have to mean complex artificial intelligence. The authors say a transparent rules-based system can be valuable if it reliably identifies repeated high readings, recognises fasting and post-meal patterns, and prompts the appropriate action. Predictive models may also help identify women at greater risk of needing insulin or closer follow-up, but the paper argues that these tools are only useful if they lead to a clear care pathway, rather than adding another layer of data with no practical effect.
The review also stresses that implementation matters as much as software design. It says successful systems need clear ownership of each alert, sensible thresholds, and workflow integration that does not add unnecessary clicks or duplicate work. The authors warn that if a system is too sensitive it can trigger alert fatigue, but if it is too strict it may miss women whose glucose control is worsening. Their preferred approach is a tiered triage model that separates routine trends from more urgent patterns and allows thresholds to be refined using real-world service data.
On evaluation, the paper argues that glycaemic measures alone are not enough. Any programme designed around decision-supported collaborative care should also be judged on process measures such as time to review, speed of patient feedback, alert burden, adherence to monitoring, and how work is shared across the team. The authors say studies should also report maternal and neonatal outcomes where possible, including hypertensive disorders, caesarean delivery, large-for-gestational-age birth, neonatal hypoglycaemia and neonatal intensive care admission. Just as importantly, they say researchers should examine safety, sustainability, workload and whether benefits are evenly distributed across different patient groups.
The review concludes that decision-supported collaborative care should be treated as a service model rather than as a piece of technology alone. Its promise lies in linking data, decision logic and team-based action in a way that is timely, transparent and workable in routine maternity care. But the authors also acknowledge that evidence remains mixed, with no single model yet proving clearly superior across clinical, workflow and implementation outcomes. In their view, the next step is more pragmatic research in real antenatal settings, with careful attention to equity, postpartum follow-up and whether these systems can improve care without simply shifting the burden elsewhere.
Disclaimer: This content is for informational purposes only and is not intended to be a substitute for professional medical judgment, advice, diagnosis, or treatment.
