## Severity of chronic obstructive pulmonary disease and its relationship to lung cancer prognosis after surgical resection.

- Yasuo Sekine, Hidemi Suzuki, Yoshito Yamada, Eitetsu Koh, Ichiro Yoshino
- The Thoracic and cardiovascular surgeon
- 2013

- Published 2014 in Heart & lung : the journal of critical care

Pulmonary function tests (PFTs) are widely used in the clinical evaluation of patients with pulmonary and cardiovascular disease. The underlying premise is that lung volumes and air flow through the airways are correlated with measurements of body size and are influenced by age. The result is a set of equations that have lung volume and airway flow as dependent variables of height, age, sex, race and weight and their appropriate coefficients as independent predictors. Higher order exponential relationships are accommodated using logarithms.1 The assumption that measurements of body size determine lung size and airway flow is derived form basic biological observations that establish that organ size in mammals scales with the square of the length and thatmeasurements of body size determine metabolic rate, heat losses and the mechanical energy necessary for distribution of oxygen through the tissues.2e4 Reference equations that establish expected or predicted values for lung volume and for spirometry based on age and height derived from datasets of healthy individuals are widely used.5,6 Central to the interpretation of PFTs is what constitutes a normal value. The conventional approach has been to accept that healthy individuals have a variability equivalent to a fixed percentage above or below a normal value for which values that fall below this percentage are considered abnormal. Following this approach a percentage below 70% for FEV1/FVC is used to define airway obstruction and an FEV1 below 80% of predicted value defines abnormalities for spirometry, lung volumes, and diffusing capacity for carbon monoxide (DLCO).7 If this approach is to be valid, the variability of the observations in a sample of patients needs to be a fixed proportion of the actual value. If one takes into account the heterogeneity of biological processes in humans, the assumption that the variability of the values for PFTs is a fixed proportion across observations is debatable. In fact, clinical studies document sources of variation related to factors different form body size such as gender, ethnicity and geographical origin.8 An alternative approach considers the observations of PFTs as samples from a normally distributed population of PFT values. Making this assumption allows the use of statistical techniques to determine what is abnormal which in this case is defined as those observations that are outside the top and bottom five percentile of the normal curve. This approach has a number of attractive features. First, it is consistent with the observed distribution of samples in large databases of PFT data which indeed correspond to a normal curve. Second, standardized measures of variation can be

@article{Berrizbeitia2014TheLL,
title={The lower limit of normal in the evaluation of pulmonary function.},
author={Luis D Berrizbeitia},
journal={Heart & lung : the journal of critical care},
year={2014},
volume={43 4},
pages={267-8}
}