Published ahead of print on February 13, 2003, doi:10.1164/rccm.200204-339OC
© 2003 American Thoracic Society Sensitivity of Spirometric Measurements to Detect Airway Obstruction in InfantsDepartment of Pediatric Pulmonology and Critical Care, Indiana University School of Medicine, James Whitcomb Riley Hospital for Children, Indianapolis, Indiana Correspondence and requests for reprints should be addressed to Robert S. Tepper, M.D., Ph.D., Department of Pediatrics Indiana University School of Medicine, James Whitcomb Riley Hospital for Children, 702 Barnhill Drive, Room 4270, Indianapolis, Indiana 46202–5225. E-mail: rtepper{at}iupui.edu
We evaluated the ability of forced expiratory flow volume curves from raised lung volumes to assess airway function among infants with differing severities of respiratory symptoms. Group 1 (n = 33) had previous respiratory symptoms but were currently asymptomatic; group 2 (n = 36) was symptomatic at the time of evaluation. As a control group, we used our previously published sample of 155 healthy infants. Flow volume curves were quantified by FVC, FEF50, FEF75, FEF25–75, FEV0.5, and FEV0.5/FVC, which were expressed as Z scores. All variables except FVC had Z scores that were significantly less than zero and distinguished groups 1 and 2 with progressively lower Z scores. The mean Z scores of the flow variables (FEF50%, FEF75%, and FEF25–75%) were more negative than the Z scores for the timed expired volumes (FEV0.5 or FEV0.5/FVC) for both groups. In general, measures of flow identified a greater number of infants with abnormal lung function than measures of timed volume; FEF50 had the highest performance in detecting abnormal lung function. In summary, forced expiratory maneuvers obtained by the raised volume rapid compression technique can discriminate among groups of infants with differing severity of respiratory symptoms, and measures of forced expiratory flows were better than timed expiratory volume in detecting abnormal airway function.
Key Words: infants • spirometry • forced expiratory flows Respiratory diseases during infancy produce predominantly airway obstruction; however, clinical examination alone may not provide a sensitive and objective quantification of the degree of airway obstruction, particularly for infants who are not acutely ill. Forced expiratory maneuvers are routinely used to quantify airway obstruction in older children and adults with respiratory disease. In contrast to measures of resistance, which are greatly influenced by the extrathoracic airway, forced expiratory flows are primarily determined by the intrathoracic airway. We have previously demonstrated that flow limitation can be achieved during forced expiratory maneuvers in infants when initiated from near total lung capacity, and we have recently described reference values for this age group (1–3). Several studies have used the raised volume rapid thoracic compression technique to assess lung function in infants; however, the variables used to quantify the forced expiratory flow volume curves differ among studies, and the relative ability of different variables to discriminate airway obstruction has not been completely evaluated (3, 4–11). In older children and adults, the most frequently used parameter to quantify airway obstruction is the FEV1 and its ratio to the FVC. FEV1 is a robust measure of airway obstruction in children and adults because it has a smaller intrasubject and intersubject variability than measures of FEF50 and FEF75. However, the time constant for lung emptying during forced expiration is significantly shorter for infants than children and adults. The rate constant for forced expiration is greater in infants than adults, and the forced expiratory maneuver is usually complete in less than on second in healthy infants (12). Therefore, short timed expiratory volumes, such as FEV0.5, have been used as a measure of airway obstruction for infants (3, 6, 7, 9, 13, 14). The aim of this study was to demonstrate the ability of forced expiratory flow volume curves from raised lung volumes to discriminate among infants with differing severities of respiratory symptoms and to compare the ability of variables used to quantify the flow volume curve to detect airway obstruction. Some of the results of these studies have been previously reported in the form of an abstract (15).
Subjects Primary care physicians referred patients to James Whitcomb Riley Hospital for Children for evaluation by a pediatric pulmonologist, who ordered spirometric measurements. From the clinical history and physical examination obtained by the pediatric pulmonologist, each infant was classified as follows: group 1 had previous respiratory symptoms but was asymptomatic on date of evaluation; group 2 was symptomatic with current respiratory symptoms such as coughing, rhonchi, or wheezing on date of evaluation. As a control group, we used our data generated from 155 healthy infants. Approval for this study was obtained from our institutional review board.
Spirometry
Analysis Unpaired t tests were used to assess differences among the infants within the respiratory groups. The mean values for all pulmonary function variables were calculated on Z scores and were compared among groups using two-way analysis of variance. Post hoc comparisons were based on the Tukey test. A p < 0.05 was considered statistically significant. The performance of a lung function variable to detect abnormally decreased airway function was assessed by calculating the sensitivity and specificity at the fifth percentile of the reference population, which corresponds to a Z score of -1.645. In addition, the performance of these variables to detect abnormal values was assessed by analysis of receiver operator characteristic curves, which were constructed in steps of 0.5 Z score between -2.5 and 2.5. The area under the receiver operator characteristic curve represented the discriminative capacity of that particular parameter.
Spirometric measurements from 69 consecutive infants referred to the infant pulmonary function laboratory between September 1997 and July 1998 were analyzed. There were 33 asymptomatic patients on the test date (group 1) and 36 symptomatic patients in group 2 (Table 1) . Patients were between 14 and 144 weeks of age. There were no significant differences among the groups with respect to sex or body length; however, group 1 was younger. The ages and body lengths of the patients were all within the range of the 155 healthy infants that we previously evaluated to generate the reference data used to calculate Z scores for the spirometric variables.
All of the pulmonary function variables in group 1 except FVC and FEV0.5 had Z scores that were significantly less than zero. In addition, the variables tracked clinical symptoms; progressing from asymptomatic to symptomatic was associated with lower Z scores. Flow variables (FEF50, FEF75, and FEF25–75) and timed volume variables (FEV0.5, FEV0.5/FVC) were able to distinguish groups 1 and 2; only FVC was not significantly different among groups. The mean Z scores of the flow variables (FEF50%, FEF75%, and FEF25–75%) were more reduced (more negative) than the Z scores for the timed expired volumes (FEV0.5 or FEV0.5/FVC) for both groups (Figure 1) . Infants in groups 1 had significantly lower Z scores for FEF50 than FEV0.5 and FEV0.5/FVC. FEF75 and FEF25–75 were significantly lower than FEV0.5, but the difference between these variables and FEV0.5/FVC did not reach statistical significance. The infants in group 2 had significantly lower Z scores for FEF50, FEF75, and FEF25–75 when compared with FEV0.5 and FEV0.5/FVC.
The numbers of subjects in each group with pulmonary function variables below the fifth percentile are summarized in Table 2 . For the reference population, between 5% and 7% of the subjects had lung function values below fifth percentile in all studied variables; this was expected, as the values were normally distributed. For the infants in the different respiratory groups, the pulmonary function variables demonstrated different performances in detecting abnormal lung function. Most of infants had FVC values in the normal range, even those infants with respiratory symptoms (group 2). In general, measures of flow (FEF50%, FEF75%, and FEF25–75%) identified a greater number of infants with abnormal lung function than measures of volume (FEV0.5, FVC, and FEV0.5/FVC) in both groups. FEF50% identified the greatest number of infants with abnormal lung function (Table 2). FEF75% and FEF25–75% had similar performances to FEF50; however, FEV0.5 and FEV0.5/FVC had lower performances in detecting airway obstruction. In addition, the differences in sensitivity between flow variables and timed volumes were more noticeable in those infants who were asymptomatic when evaluated (group 1).
The sensitivities and specificities for the different pulmonary function variables to detect "abnormal lung function" in the control group and combined respiratory groups are summarized in Table 3 . Using a cut-off of Z = -1.645, the specificities were 93% for all variables, and the sensitivities ranged from 12 to 51%. The pulmonary function parameter with the greatest area under the receiver operator characteristic curve was FEF50%, followed by FEF75% and FEF25–75%, whereas FEV0.5, FVC, and FEV0.5/FVC demonstrated lower discriminative capacity.
Our study demonstrated that several variables used to quantify the forced expiratory flow volume curves obtained with the raised volume rapid thoracic compression technique successfully discriminated infants with different degrees of respiratory symptoms. We found that there was a worsening of airway function as the groups progressed from asymptomatic to symptomatic. We also demonstrated that measures of volume-referenced flows had greater sensitivity than a timed expiratory volume in detecting infants with abnormal lung function. The group of infants with a history of respiratory symptoms but asymptomatic at the time of evaluation had decreased airways function compared with our reference population. This finding demonstrates that in the assessment of asymptomatic infants, measurement of forced expiratory flow volume curves from raised lung volumes can offer additional and quantitative information to our clinical examination. Our findings are consistent with previous reports of lower airway function in healthy infants exposed to cigarette smoking, as well as lower airway function in asymptomatic infants with cystic fibrosis (3, 5, 9, 10, 16). In contrast to the asymptomatic group 1, those infants with respiratory symptoms (group 2) had mean Z scores for forced expiratory flows between -2 and -3 and approximately 50% of these infants had abnormal values. We found that the impairment in airway function was more evident with volume-referenced flows (FEF50, FEF75, and FEF25–75) than timed volume variables (FEV0.5 and FEV0.5/FVC). Although the Z scores were significantly below zero for all variables other than FVC, the volume referenced flow variables had lower, more negative mean values than the timed expiratory volumes (Figure 1). In addition, a smaller number of infants with abnormal lung function were detected with timed expiratory volumes, particularly for patients in group 1, those infants who were asymptomatic at time of testing. The overall performance of the variables assessed by receiver operator characteristic curves was also lower for FEV0.5 than for volume-referenced flows. These results suggest that flows have a better discriminative power than timed volumes, particularly in asymptomatic patients. Timed expiratory volume, such as FEV1, is the preferred parameter to assess airway function in older children and adults because of the lower intratest and intersubject variability than forced expiratory flows. Although FEV0.5 also has a lower intratest and intersubject variability than forced expiratory flows in infants, FEV0.5 does not appear to be a better parameter than forced expiratory flow to assess airway obstruction in infants (3, 8). The difference between infants and adult subjects in the utility of flows versus timed volumes to detect airway obstruction may reflect the progressive increase in the time constant for emptying of the lung with progressive growth from infancy to preschool years to adulthood (3, 12, 17). This results from results from the more rapid increase in lung volume than airway size early in life. In contrast to healthy adults whom empty approximately 80% of FVC in 1 second (FEV1/FVC = 80%), almost all infants and many preschool children empty their entire FVC during the first 1 second (FEV1/FVC = 1). As the time constant for emptying of the lung increases during infancy, the ratio of a timed volume to FVC (FEVt/FVC) will not remain constant during this period of rapid lung growth. Therefore, there is not a single ratio for any FEVt/FVC that can be used through this age range. In this study, our reference data were generated in our laboratory, and the reference values were expressed as Z scores, which adjusted for growth without assuming a constant ratio for FEV0.5/FVC. Using a shorter timed volume than FEV0.5 would potentially improve the performance of this variable in the youngest infants but at a cost of lowering its performance in older infants. On the other hand, longer timed volumes would approach FVC in many infants, reducing its capability to detect obstruction (11). In summary, variables measured from forced expiratory maneuvers obtained by the raised volume rapid compression technique can detect airway obstruction in asymptomatic infants with a history of previous respiratory symptoms, as well as discriminate a group of infants with respiratory symptoms. Among the different variables that we used to quantify the forced expiratory flow volume curve, forced expiratory flows were better than timed expiratory volume in detecting abnormal airway function. Received in original form April 17, 2002; accepted in final form February 6, 2003
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
