To visualize the relationship between different combinations of mechanical power exposure intensity-duration and death risk in mechanical ventilation patients using a visualization method. Critically ill patients receiving mechanical ventilation were selected from the Medical Information Mart for Intensive Care- IV v1.0 (MIMIC- IV v1.0) database. The patients were divided into four subgroups according to oxygenation index (PaO2/FiO2) including > 300 mmHg (1 mmHg ≈ 0.133 kPa) group, 201-300 mmHg group, 101-200 mmHg group and ≤100 mmHg group. The baseline characteristics, ventilator parameters, and prognostic indicators for different patient populations were collected. For each patient, the mechanical power thresholds from low to high (5-30 J/min, increasing at intervals of 1 J/min) were used to evaluate the different exposures of mechanical power (above the set threshold was recorded as one exposure), and the number of events with different exposure intensity-duration combinations was counted based on their corresponding durations. Based on the 28-day survival/non-survival status, the number of exposures for survivors and non-survivors in each exposure intensity-duration combination was calculated, and the survival odds ratio (OR) for different mechanical power exposure intensity-duration combinations was subsequently computed. Two-dimensional tables were generated with mechanical power exposure duration on the x-axis and exposure intensity on the y-axis, and the heatmap and its corresponding equipotential line view were used to visualize the OR value to assess the risk of death. A total of 5 378 patients receiving mechanical ventilation were enrolled in the study, of whom 2 069 patients in the PaO2/FiO2 > 300 mmHg group, 813 patients in the 201-300 mmHg group, 1 493 patients in the 101-200 mmHg group, and 1 003 patients in the ≤100 mmHg group. The severity scores of patients, including sequential organ failure assessment (SOFA) score and simplified acute physiology score II (SAPS II), gradually increased following the decrease in PaO2/FiO2, and the incidence of co-morbidities also gradually increased. In terms of ventilator parameters, mechanical power was increased gradually with decrease in PaO2/FiO2, measuring 10.4 (7.8, 13.9), 11.3 (8.5, 14.7), 13.6 (10.0, 18.2), and 16.7 (12.5, 22.0) J/min (P < 0.01). In terms of prognosis, 28-day mortality of patients was gradually increased with decrease in PaO2/FiO2 [29.1% (601/2 069), 26.9% (219/813), 28.1% (420/1 493), and 33.3% (334/1 003), respectively, P < 0.05]. In the heatmap, it could be observed that the 28-day death risk of mechanical ventilation patients was gradually increased with increase in mechanical power exposure intensity and long duration, showing two distinct areas: a region near the bottom left corner (representing low mechanical power exposure intensity and short duration) was blue, indicating a greater chance of survival. In contrast, another region near the top right corner (representing high mechanical power exposure intensity and long duration) was red, indicating a higher risk of death. According to the fitted lines of death risk, for the same risk of death, a shorter mechanical power exposure duration was required for higher exposure intensity, while lower mechanical power exposure intensity required a longer exposure duration. The above trend of change was similarly reflected in the overall population and different oxygenation populations. Cumulative mechanical power exposure to higher intensity and/or longer duration is associated with worse outcomes in mechanical ventilation patients. Considering both the mechanical power exposure intensity and duration may help to evaluate the effectiveness of lung protection in mechanical ventilation patients and guide adjustments in mechanical ventilation strategy to reduce the risk of ventilator-induced lung injury.
Read full abstract