Refining Early Warning Systems to Combat Maternal Sepsis Across Diverse Settings

The study evaluated 21 early warning systems (EWS) for detecting maternal sepsis using data from 46 countries, finding no single system adequate for universal use. Obstetric-specific tools performed better, but gaps in sensitivity, specificity, and data highlight the need for refined, context-specific diagnostics.

Conducted by a global consortium of research institutions, including the Malawi-Liverpool Wellcome Programme, the University of Liverpool, and the World Health Organization (WHO), the study explored the effectiveness of early warning systems (EWS) in identifying severe maternal outcomes caused by sepsis. Using data from the WHO Global Maternal Sepsis Study (GLOSS), the research analyzed the performance of 21 EWS in over 2,500 women hospitalized for suspected or confirmed infections in 46 countries. Maternal sepsis, a life-threatening condition stemming from infections during or after pregnancy, is a leading cause of maternal deaths globally, responsible for approximately 11% of these fatalities. Despite advances in maternal healthcare, the study highlighted gaps in existing tools, urging the development of refined systems tailored to the unique challenges of diverse healthcare settings.

The Search for Effective Early Warning Systems

Researchers evaluated a range of EWS, including obstetric-specific and sepsis-specific tools such as the Modified Shock Index (MSI), maternity Systemic Inflammatory Response Syndrome (mSIRS), and quick Sequential Organ Failure Assessment in Pregnancy (qSOFA-P). These systems were assessed for their ability to identify women at risk of severe maternal outcomes, including organ dysfunction and infection-related deaths. The MSI and NICE Risk Stratification Tool for postpartum and post-abortion women (NICE-RST-PP/PA) demonstrated high sensitivity but suffered from low specificity, leading to frequent false positives. This could overwhelm healthcare systems with unnecessary alerts. On the other hand, tools like qSOFA-P showed high specificity but were limited by low sensitivity, risking missed diagnoses of at-risk women. Machine learning models analyzing combinations of clinical markers provided promising specificity but also faced limitations in sensitivity, restricting their broader application.

Obstetric-Specific Tools Show Promise

The study revealed that obstetric-specific tools often outperformed general diagnostic systems in identifying maternal sepsis. For instance, mSIRS adjusted its thresholds to account for pregnancy-specific physiological changes, offering higher sensitivity compared to general systems like SIRS, which often triggered alarms unnecessarily. This adjustment reduced the risk of false positives while improving detection accuracy. However, the lack of comprehensive clinical data, particularly in low-resource settings, posed a significant barrier to the effectiveness of these tools. Key parameters such as Glasgow Coma Scale scores, temperature readings, and laboratory markers like lactate levels were missing in over half of the analyzed cases, complicating the evaluation of EWS accuracy. These gaps underscored disparities in healthcare infrastructure and emphasized the need for scalable, context-specific solutions.

Balancing Sensitivity and Specificity for Better Outcomes

One of the critical challenges highlighted by the study was balancing sensitivity and specificity in EWS. High sensitivity ensures that most at-risk patients are identified, but it often leads to excessive false positives, which can overwhelm healthcare systems, particularly in resource-limited settings. Conversely, high specificity reduces false alarms but risks missing critical cases. Tools like the NICE-RST-PP/PA, which include advanced diagnostic components such as lactate measurement, are ideal for high-resource settings with robust laboratory capacity. Simpler tools like the MSI, which rely on vital signs like pulse rate and blood pressure, are better suited for low-resource environments where advanced equipment is unavailable. The study emphasized the importance of tailoring EWS to local healthcare contexts to maximize their utility and effectiveness.

A Call for Innovation in Maternal Health Diagnostics

While the study confirmed the value of EWS in improving maternal care and preventing sepsis-related complications, it also highlighted significant limitations in their current design and application. No single system was found to be universally effective across all settings, and researchers called for continued innovation to refine existing tools and develop new ones. Future efforts should explore alternative diagnostic markers and employ cutting-edge approaches like machine learning to create more accurate and adaptive systems. Training healthcare workers to effectively use EWS and integrate them with clinical judgment is also crucial, especially in settings with limited resources. By addressing these gaps, the global healthcare community can take significant steps toward reducing maternal mortality and improving outcomes for women worldwide.

This study underscores the urgent need for advancements in maternal sepsis diagnostics. The findings reveal that while EWS can play a crucial role in preventing severe maternal outcomes, their effectiveness is limited by gaps in sensitivity, specificity, and data availability. Continued research and innovation are essential to create more reliable, context-specific tools. Until then, combining EWS with expert clinical judgment remains the best approach to addressing the challenges of maternal sepsis in diverse healthcare settings. With the right investments and collaborative efforts, these tools have the potential to save countless lives and transform maternal health outcomes globally.