The pulmonary vascular permeability index (PVPI) is a quotient of the extravascular lung water (EVLW) and the pulmonary blood volume (PBV). In acute respiratory distress syndrome (ARDS), the alveolar-capillary membrane integrity is disrupted. The result is a disproportionate increase of EVLW compared to the PBV and, hence, an increase in PVPI. Thus, PVPI has repetitively been discussed to extend the definition of ARDS. Besides sex, the influence of other anthropometric variables on PVPI has not been studied so far. However, since it is known that EVLW depends on body height and sex, we hypothesize that PVPI depends on anthropometric variables as well. This prospective single-center observational study included 1533 TPTD measurements of 251 non-critically ill patients (50.6% men) undergoing elective neuro-, thoracic, or abdominal surgery at the Munich Clinic Bogenhausen of the Technical University of Munich. Multivariate regressions were used to measure the influence of sex, age, and body height on PVPI. In all patients, PVPI was significantly higher in women (P < 0.001), with 34.4% having a PVPI > 2 compared to 15.9% of men. Mean PVPI significantly decreased with height (P < 0.001) and age (P < 0.001). Multivariate regressions allowed the calculation of mean reference surfaces. The 95th percentile surface for PVPI was > 3 for small and young women and well above 2 for all but tall and elderly men. In patients who underwent (lung reduction) thoracic surgery, the PVPI before and after surgery did not differ significantly (P = 0.531), and post-surgical PVPI did not correlate with the amount of lung resected (P = 0.536). Hence, we conclude that PVPI may be independent of the extent of lung volume reduction. However, PVPI is heavily dependent on sex, age, and body height. Anthropometric variables thus have a significant impact on the likelihood of misclassified abnormal PVPI. This warrants further studies since an increased PVPI, e.g. in the context of an ARDS, may be overlooked if anthropometric variables are not considered. We suggest reference surfaces based on the 95th-percentile corrected for sex, age, and height as a novel approach to normalize PVPI.
Read full abstract