Introduction
The COVID-19 pandemic presented an unprecedented challenge to hospital systems, particularly in regions like Lombardy, Italy. Patients admitted during the first wave often exhibited clinical profiles that diverged significantly from standard general ward populations. While individual Early Warning Scores (EWS) are effective for monitoring single patients, they lack the capacity to provide a system-wide overview of clinical stability. This study, published in the Journal of Clinical Monitoring and Computing, introduces a novel heat-map visualization technique to quantify and visualize the "intensity of care" required across hospital wards by comparing COVID-19 patients to historical controls.
Methodology
Researchers at Humanitas Research Hospital in Milan conducted a retrospective Electronic Health Record (EHR) study. They compared two primary cohorts:
- COVID-19 Group: 470 lab-confirmed patients hospitalized between February 21 and April 21, 2020.
- Control Group: 9,241 non-COVID patients hospitalized in the same wards during 2018, representing a standard mix of surgical and medical cases.
"Severe derangement" was defined by specific clinical thresholds, including SpO2 < 90%, respiratory rate anomalies (> 25 or < 10 bpm), and fever (> 37.5 °C). The core innovation was the use of heat maps where rectangle size represents the cumulative number of patients in a ward, and color intensity indicates the frequency of vital sign measurements falling into the "severely deranged" category.
Clinical Signatures and Ward Stratification
The study identified a distinct clinical signature for COVID-19, primarily driven by respiratory instability. The frequency of acute desaturation (SpO2 < 90%) was nearly 8 times higher in COVID-19 wards compared to 2018 controls, while respiratory rate anomalies were 9.5 times more frequent.
Interestingly, a "hypotension paradox" was observed: COVID-19 patients were less likely to experience low blood pressure (rate ratio 0.8) than the standard hospital population, which typically includes more sepsis and post-surgical cases. Ward-level stratification further revealed that even within COVID-designated areas, certain units (Wards 5 and 6) emerged as "hot spots" requiring significantly higher intensity of care due to the use of non-invasive ventilation.
Age and Mortality Outcomes
Age emerged as the most significant predictor of mortality. While the overall mortality for hospitalized COVID-19 patients was 18%, it varied drastically by age group: 1.4% for those under 65 versus 30% for those over 65.
Conclusion
The use of heat maps provides an intuitive, data-driven method for visualizing hospital-wide clinical deterioration. By identifying "hot spots" of instability, hospital administrators can more effectively allocate resources such as nursing staff and respiratory equipment. Future integration of these visualizations into real-time dashboards could shift hospital management from retrospective analysis to proactive surge capacity planning.
Summary updated