Article
Higher mortality in female versus male critically ill patients at comparable thresholds of mechanical power: necessity of normalization to functional lung size
Dario von Wedel, Simone Redaelli, Boris Jung, et al
Intensive Care Med 2025; 51: 624-626
Published: 23 January 2025
Dear Editor,
Mechanical power (MP) is an integrative target to minimize risk of ventilator-induced lung injury (VILI) [1]. Estimating the energy delivered by the ventilator to the lung, MP is calculated from multiple ventilator parameters and traditionally reported in J/min [1]. Injurious effects per J/min of MP have been suggested to depend on the surface area of the lung (i.e., functional lung size) to which the energy is applied [2]. Disproportionate effects could arise among patients with different functional lung sizes. One important example may be female versus male patients, as females on average have smaller lungs [3].
Based on these physiological differences, we hypothesized that (1) the association between MP (in J/min) and mortality is modified by patient sex, affecting females more markedly due to smaller functional lung sizes and (2) those differences are mitigated by normalization of MP to static respiratory system compliance (Crs) as proxy for functional lung size.
We retrospectively included mechanically ventilated, adult patients from intensive care units at Beth Israel Deaconess Medical Center, Boston, USA. During the first 24 h of controlled mechanical ventilation, MP [1, 4, 5] was calculated as:
Normalized MP (MPnorm) was defined as MPnorm = MP/Crs [2]. The modification of the association of MP and MPnorm with 28-day mortality by self or proxy-reported patient sex was assessed using interaction terms and logistic regression, with a comprehensive confounder model (Electronic Supplemental Material, ESM).
Among 19,806 (37.9% female) patients (ESM), females had a 12.1% (95%CI 10.7–13.4) lower MP compared to males. By contrast, MPnorm was 10.4% (95%CI 8.4–13.5) higher in females.
In adjusted analysis, a higher MP was associated with increased odds of mortality (aOR: 1.44 per 5 J/min increase, 95%CI 1.39–1.49, p < 0.001). This association was modified by patient sex (p-for-interaction = 0.018), resulting in a 52.8% larger risk increase per J/min increase in female compared to male patients (Fig. 1A). At a previously suggested harmful threshold of 17 J/min (5), mortality risk was an absolute 8.2% (95%CI 7.0–9.4, p < 0.001) higher in females compared to males.
The normalization of MP mitigated differential effects (p-for-interaction = 0.23, Fig. 1B). The effect modification of MP by Crs (Fig. 1C) suggested that differential effect sizes were not due to sex per se but rather differences in functional lung size.
The mediation analysis revealed that higher MPnorm delivered to female patients partially mediated increased mortality in females (21.4% mediation).
The effects of driving pressure, as a VILI marker scaled to Crs, were comparable among males and females (p = 0.14, ESM).
This is the first study to report differential effects of MP on mortality, depending on functional lung size. In females compared to males, similar MP is associated with higher risks of mortality due to smaller functional lung sizes. When thresholds of MP are to be established, normalization of MP to functional lung size (MPnorm) should be done to mitigate these differences. The mediation analysis revealed that higher mortality among female patients can be partially explained by higher levels of MPnorm. This highlights the utility of MPnorm as potential target to mitigate disparities and VILI, which should be investigated prospectively.
