Utility of standardized exercise oximetry to predict cardiopulmonary morbidity after lung resection.

Abstract

To evaluate if desaturation, measured by finger oximetry on standardized exercise, accurately predicts cardiopulmonary morbidity after pulmonary resection. A prospective observational clinical study was carried out on 81 consecutive lung carcinoma patients scheduled for pulmonary resection from February 1998 to March 1999. Finger oximetry was monitored during an incremental to exhaustion cycle exercise test. The presence or absence of desaturation (cut-off value 90%) during exercise was recorded. Other independent analyzed variables were: age of the patient (over 75th percentile), body-mass index (BMI) (over 75th percentile), presence of major cardiovascular co-morbidity, and calculated postoperative FEV1% (under 25th percentile) according to the number of resected segments (ppoFEV1%). Postoperative cardiopulmonary morbidity was the evaluated dependent outcome. Fisher's exact test and risk calculation on contingency tables were used for statistical analysis. A lobectomy was performed in 62 cases, a pneumonectomy was performed in 16 cases, and a segmentectomy was performed in three cases. The mean age of the patients was 63.6 years (SD 10.3, range 34-79 years, 75th percentile 72 years), the mean BMI was 25.9 (SD 4.9, range 16.9-38.1, 75th percentile 29.3), and the mean ppoFEV1% was 64.1 (SD 2016.1, range 29.5-98.6, 25th percentile 50.5). In 14 patients exercise desaturation was registered. Postoperative cardiopulmonary morbidity was presented in 32 cases (five deaths). No correlation was found between postoperative morbidity and any of the following variables: age of the patient, BMI, and co-morbidity. On univariate analysis only low ppoFEV1% (P<0.001) was associated with the outcome, so no multivariate analysis has been carried out. We have shown that desaturation during standardized exercise in this series adds no relevant information to predict postoperative cardiorespiratory morbidity after lung resection.