Value of the average basal daily walked distance measured using a pedometer to predict maximum oxygen consumption per minute in patients undergoing lung resection.

OBJECTIVES

Maximum oxygen consumption per min (VO(₂max)) is currently considered the most accurate test for the preoperative risk assessment in patients scheduled for pulmonary resection. Due to its high-technology requirements and cost, VO(₂max) is performed less frequently than is desired. The objective of this investigation is to determine if the measurement of the basal daily ambulatory activity of the patients, with a pedometer, can be used to predict VO(₂max) values.

METHODS

This is a prospective study on 38 patients referred for scheduled lobectomy or pneumonectomy. Daily basal preoperative activity of the patients was measured 3 weeks before surgery by means of an OMROM HJ-72OIT-E2 pedometer. Before surgery, VO(₂max) (dependent variable) was calculated using a Master Screen CPX module of Jaeger-Vyasis-Healthcare. The following independent variables were studied: age, sex, preoperative forced expiratory volume in 1s percentage (FEV1%) and carbon monoxide diffusing capacity percentage (DLCO%), mean number of steps per day (aerobic and non-aerobic), mean daily time of aerobic activity (in min) and mean daily walked distance (in km). Two linear regression models with bootstrap robust estimation of the standard error of the coefficients were adjusted and the estimated values of VO(₂max) were kept as a new variable for comparison. To avoid collinearity problems, only one of the pedometer records entered the regression model.

RESULTS

Data of the series (mean ± SD): age 62.8 ± 10.14 years; FEV1% 90.1 ± 21.8; DLCO% 82.8 ± 20.1. After collinearity analysis, mean daily walked distance was chosen as the most representative variable. In the first regression model, 'Distance' (p = 0.000) was highly correlated to the dependent variable (adjusted R²: 0.812). The second model improved the predictive value of the first one adding DLCO% to the model. In this model, DLCO% (p = 0.000) and 'Distance' (p = 0.002) were correlated to the dependent variable. The adjusted R² of the second lineal model was 0.935.

CONCLUSION

These preliminary data show that a combination of the measured daily ambulatory activity using a pedometer, especially the mean daily walked distance in km, and the DLCO% of the patient could predict the VO(₂max) value. Larger data series are needed for conclusive results.