INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: preschool; spirometry; computer-animated; lung function

Respiratory disorders are very common in preschool children (1), yet a "diagnostic gap" remains in this age group with respect to objective testing of pulmonary function. Spirometry has been recognized as a hallmark procedure in the evaluation of lung disease for more than a century. The results of an FVC maneuver help to quantify functional compromise. Disease course and response to therapy can be evaluated by repeated testing under standardized conditions (4). Technical recommendations and normal data have been published for infants and for children of school age (4). However, present knowledge about lung function and growth in the 3 to 6 yr age group is scarce and mainly derived from indirect methods or extrapolation of spirometric data from older children and infants (10). Both the clinical management of lung disease and the evaluation of treatment efficacy in this age group are based almost entirely on clinical and radiologic information with inherently limited sensitivity. A majority of preschool children will not produce adequate FVC loops upon verbal instruction. Their understanding of the maneuver, ability to differentiate inspiration from expiration, and attention span are naturally limited. Therefore, several manufacturers have integrated computer-incentive games into commercially available spirometry software. These teach the subject to blow out candles, move clouds, pop balloons with arrows, and perform other operations, and concentrate exclusively on forced expiratory flow. However, if both the initial inhalation to TLC and the complete exhalation to RV are disregarded, the FVC maneuver may be partial, and spirometric values other than peak flow may be unreliable.

In the study reported here, we wished to assess whether a new computer-animated interactive system (SpiroGame; Zan Messgeraete Gmbh, Oberthulba, Germany) could encourage participation and train young children to provide valid and reproducible spirometric test results.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

One hundred and twelve children aged 3 to 6 yr (Table 1) participated in the study. We contacted kindergarten facilities in the vicinity and attempted to include all of the attending children in a cross-sectional design. Recruitment was not based on specific characteristics of either children or parents. The children enrolled in the study were considered naive to spirometry. A technician who was moderately experienced with lung function testing in children, and had not previously met any of the children, conducted all of the tests. Measurements were performed in the children's kindergartens. Parents gave written informed consent and answered a questionnaire about the medical history of the child and the child's family that specifically addressed respiratory problems. Symptoms and signs of airway disease were recorded on the day of testing. No child had previously performed pulmonary function tests.

Instrumentation

The SpiroGame (patent pending) includes a small pneumotachograph (ZAN100; ZAN Messgeraete GmbH, Oberthulba, Germany). The pneumotachograph, pressure transducer, and A/D converter are bonded inside a hand-held breathing apparatus with the design of an ice-cream cone (Figure 1). The total breathing apparatus dead space was 20 ml. Physical characteristics were evaluated at the Institute of Fluid Power Transmission and Control (IFAS) of the Technical University of Aachen, and met American Thoracic Society (ATS) guidelines (4). Data acquisition, processing, calibration, and data processing by computer software were done on a personal computer connected to the pneumotachograph. The software consists of two games: The first game interactively teaches the child who is undergoing testing to differentiate between inhalation and exhalation by simulating a caterpillar crawling to an apple over a period of 30 s of tidal breathing. Tidal volume (V_t) and breathing rate targets are preset at 7 to 10 ml/kg and 20 ± 7 (mean ± SD) breaths/min. The second game interactively teaches an FVC maneuver by simulating a bee flying from flower to flower and over a fence. Breathing targets are initial tidal breathing, followed by IC, peak expiratory flow (PEF), and FVC. The targets are calculated form extrapolated predicted values for height (13). The game has several levels of difficulty, and upon completion of a maneuver a new picture rewards the child for success.

View larger version (67K): [in this window] [in a new window]   Figure 1.   Illustration of the SpiroGame breathing apparatus, designed as an ice-cream cone. The pneumotachograph is embedded in the "ice-cream" part of the apparatus, and the pressure transducer in the waffle shaped "cone."

For comparison, we used the MasterLab system (E. Jaeger, Inc., Hoechberg, Germany). The software incentive with this system simulates the blowing out of candles (candle-blowing), and the target flow rate is calculated according to predicted PEF. The spirometer contains an adult-size pneumotachometer (Fleisch pneumotaphygraph #3) with a dead space of 35 ml. Comparability of the two spirometers used in the study was assessed in a preliminary study through mimicry of 20 FVC maneuvers with a 1-L calibration syringe (electronic correction for body temperature, atmospheric pressure, and saturation [BTPS] was bypassed). Both systems yielded reproducible although slightly different results: the mean FVC with the ZAN PT pneumotachograph was 1.00 ± 0.011 and that with the Jaeger pneumotachograph was 1.02 ± 0.004 L (p < 0.001). Subjects performed the tidal breathing game and were then randomized to perform FVC maneuvers with either the SpiroGame followed by the Master Lab candle-blowing game, or vice versa.

Cooperation with the procedure was scored by the examiner as follows: 0 = none, 1 = partial, 2 = sufficient, 3 = good, and 4 = very good repetitive effort. Training time, defined as the time for learning to perform correct spirometry, included both the instructions and the performance. Instructions were continued until it was felt that the child had achieved its maximal possible performance. Training time was measured. All loops were reevaluated by each author separately, and were excluded if their shape indicated poor effort, coughing, or glottis closure. Incomplete expiration, as defined by an abrupt decrease of flow to zero from values > 300 ml/s (expiratory breaking point), was also a cause for exclusion. The best loop was defined as the effort with the highest FVC + FEV₁ values (or FVC + FEV_0.5 values in the absence of FEV₁). The ethics committee of Aachen University Hospital approved the protocol for the survey.

Statistics

Student's paired and unpaired t tests were used for comparison of the SpiroGame and candle-blowing systems, and a value of p < 0.05 was considered significant. Reproducibility was evaluated through the differences between the two best FVC maneuvers, and the intrasubject coefficient of variation (CV) was calculated. FVC, FEV₁, and PEF values of healthy children using the SpiroGame were compared with extrapolated predicted values by using a paired t test and exponential regression correlation. Differences between the two spirometric systems examined in the study, using mean measurements of the same variable, were tested with a Bland and Altman analysis (14).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The overall cooperation score was good with both the SpiroGame and candle-blowing methods, with a score of 2.7 ± 1.0 (mean ± SD) for the SpiroGame and 2.4 ± 1.2 for the candle-blowing method. Of the 112 children included, 10 could not be motivated to perform spirometry. Of these, six children (five 3-yr-old and one 4-yr-old child) declined further testing after determination of their height and weight. Another four children did not continue after the tidal breathing game. The remaining 102 children agreed to attempt an FVC maneuver and were included in the comparative analysis of the two spirometric methods. Table 2 compares both systems for cooperation (n = 112), percentage of success, and training time (n = 102).

Of the 102 participants who attempted FVC maneuvers, 69 were free of acute or chronic respiratory disease. Twenty-one suffered from rhinitis and/or cough on the day of testing. The other 12 suffered from various lung diseases: consisting of recurrent bronchitis or wheezing in five, neonatal chronic lung disease in two, and physician-diagnosed bronchial asthma and allergies in five.

Successful spirometry was achieved by 69.6% of the children with the SpiroGame method and by 47.1% with the candle-blowing method (n = 102, p = 0.002). Neither success rate nor absolute values were significantly affected by the order of measurements. Young children were more likely to entirely refuse a measurement, which explains the paradoxical finding of higher success rates in the age group of 3- and 4-yr-old subjects. Among those agreeing to try spirometry, (n = 102), success rate and training time were independent of age.

The main reason for failure with both systems was poor effort due to lack of comprehension or coordination, rather than poor motivation. Incomplete expiration was common with the candle-blowing method (41 of 102 children), but rare with the SpiroGame (six of 102 children). With both systems, children performed up to five successful loops each, with no significant age effect but with much variability between children. The number of successful loops per child was 1.5 ± 1.5 for the SpiroGame and 1.0 ± 1.3 for the candle-blowing method (p = 0.064). Teaching of the tidal breathing game took 6.4 ± 1.7 min. Mean training time for the FVC maneuver was 6.9 ± 1.6 min for the SpiroGame and 4.4 ± 0.8 min for the candle-blowing method (p < 0.001).

Comparison of the spirometric parameters for both systems, and intrasubject CV for each method are shown in Table 3. Most remarkably, expiration for longer than 1 s, producing values for FEV₁, was achieved by 55 of 69 (79%) subjects with the SpiroGame but by only by two of 48 subjects (4%) with the candle-blowing method, rendering the results of the two methods incomparable. FEV_0.5 was produced by all 69 (100%) children with the SpiroGame and by 46 (96%) children with the candle-blowing method. For children who produced at least one adequate loop with both methods (n = 36), FEV_0.5 values were significantly lower with the SpiroGame than with candle-blowing (0.81 ± 0.16 L and 0.94 ± 0.16 L, respectively; p < 0.001). The maneuver for PEF was performed adequately with the SpiroGame only when the whole maneuver was performed correctly. The PEF maneuver was performed correctly with candle-blowing by 90 of 102 children, regardless of the adequacy of the entire FVC maneuver. PEF values were significantly lower with the SpiroGame than with candle-blowing method (2.45 ± 0.48 L/s and 2.54 ± 0.47 L/s, respectively, p < 0.02). No significant difference was found for expiratory flow at 50% of FVC (MEF₅₀). A representative FVC maneuver is shown for each method in Figure 2. Actual FVC values with the SpiroGame were slightly higher than with the candle-blowing method (1.07 ± 0.37 L and 1.06 ± 0.44 L, respectively), although this trend did not reach significance. However, the 2% higher volume reading in the calibration syringe trial that was found with the candle-blowing pneumotachograph than with the SpiroGame pneumotachograph might make this difference significant.

View larger version (10K): [in this window] [in a new window]   Figure 2.   Representative forced flow-volume curves (actual: solid line) and extrapolated predicted: broken line) as generated: (A) by a 3-yr-old child using SpiroGame, who continued an expiration effort for longer than 1 s; and (B) by a 4-yr-old child using the candle-blowing system, for whom a premature expiration break can be seen at 500 L/s.

The intrasubject CV was below 5% for most parameters. The CV was slightly better with the SpiroGame for FVC and FEV_0.5, whereas the candle-blowing method yielded peak flow parameters with higher reproducibility. The graphic analysis done according to the method of Bland and Altman did not disclose a systematic difference between the two systems for the measurement of FVC (presented in Figure 3), FEV_0.5, or PEF.

View larger version (18K): [in this window] [in a new window]   Figure 3.   Analysis of the difference in FVC values with the SpiroGame and candle-blowing spirometry systems, as compared with mean FVC values in a Bland and Altman analysis (14). Dotted lines represent 95% coefficient of variation values.

Of the 69 symptom-free children in the study, 46 performed an adequate FVC maneuver with the SpiroGame method. However, spirometry was strongly suggestive of obstructive airways disease in seven of these children, although no specific history had been reported. Results of spirometry for the remaining 39 healthy children were compared with extrapolated predicted values according to height (Figures 4a to 4c) (13). FVC values were below the extrapolated predicted values (1.25 ± 0.39 L and 1.34 ± 0.37 L, respectively; p < 0.001). There was no difference between actual FEV₁ and extrapolated predicted values for this measure (1.22 ± 0.37 L and 1.20 ± 0.32 L, respectively), nor in actual and extrapolated predicted PEF values (2.76 ± 0.83 L/s and 2.82 ± 0.70 L/ s, respectively). MEF₅₀ values were significantly higher than extrapolated predicted values (2.02 ± 0.51 L/s and 1.85 ± 0.40 L/s, respectively, p < 0.001), as were values for IC preceding forced expiration (1.18 ± 0.34 L and 1.04 ± 0.25 L, respectively; p < 0.001).

View larger version (23K): [in this window] [in a new window]   Figure 4.   Spirometry values in relation to height. Spirometry values for 39 healthy children who performed acceptable FVC maneuvers with SpiroGame (open circles) as compared with extrapolated predicted values (solid lines), plotted against height (13). Top: FVC. Middle: FEV1. Bottom: peak expiratory flow (PEF). Circles, actual values; line, extrapolated prediction.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We developed and assessed SpiroGame, a system that facilitates the performance of spirometry in very young children. This new method is intended to overcome the problems that arise when traditional spirometry is applied to young children, such as inadequate hardware design, limited cooperation and comprehension of technique, and quality control. Traditional spirometric devices are large and bulky, with dead space sometimes exceeding 3 ml/kg and a pneumotachograph insensitive in the lower flow range, leading to unacceptable inaccuracies. Games as a visual incentive for lung function tests are not new, but generally do not differentiate between the separate steps of the FVC maneuver. Most children under 6 yr of age tend to breathe rapid, repeated tidal volumes and are difficult to train in the performance of a single, large forced expiratory maneuver, leading to inaccurate maximal flow rates and volumes. Incomplete initial inspiration to TLC, submaximal effort on expiration, or incomplete expiration with a break before reaching RV greatly affects the results.

The ATS and the British guidelines (4, 6) provide strict quality control criteria for spirometric testing. These standards are essential for adults and may be suitable for school children, but children under the age of 6 yr have rarely been able to meet the requirements for length of expiration of reproducibility index (15). As a result of these shortcomings, and occasionally because of the small true VC of preschool children, volume may be expired forcefully in less than 1 s (15). Our findings strengthen the proposal that FEV_0.5 may be an alternative measure in these cases (16).

In the present study, we sought to overcome these problems by utilizing a small, user-friendly, ice cream cone-shaped pneumotachograph and an interactive visual incentive. Performance with SpiroGame was compared with that achieved with the flow-targeted candle-blowing game. With both systems, the influence of a patient and skilled examiner cannot be overestimated. Similarly, the familiar environment, rather than the setting of a clinic laboratory, enhanced the children's performance in our study. The initial game, teaching tidal breathing, probably also contributed to a higher rate of success in FVC maneuvers with both systems.

Acceptance by subjects was high for both systems. The duration of training with SpiroGame was similar to that described for school children and adults performing spirometry, and much shorter than previously described for younger children. Although training time was somewhat longer with the SpiroGame than with the candle-blowing method, it resulted in a markedly greater success rate (Table 2).

Definition of criteria for the acceptability of results of spirometry remains a crucial issue in young children. Although new "end of test" criteria have been suggested (17), no general consensus exists for preschool children. The cutoff values that we used are arbitrary and serve only as a suggestion. Clinicians and lung physiologists may differ in opinion about the acceptability of these criteria. Our decision to regard a single adequate loop as an acceptable test result, and to not implement a criterion of reproducibility, can also be debated. We felt that meaningful information would have been discarded if standard criteria for older children and adults had been applied. Nevertheless, a substantial proportion of children succeeded in producing data of a quality and reproducibility vastly superior to those found in earlier observations. Successful spirometric measurements in an emergency room study were reported in 0% to 17% of cases for children 2 to 3 yr old and 43% for children 5 to 9 yr old (18). In another study, two of five 5-yr-old children and 11 of 17 6-yr-old children performed an acceptable FVC maneuver (19, 20). Wagner and colleagues (21) reported a 61% success rate in children 5 to 7 yr old, but the children were not naive to pulmonary function testing. Given the relevance of objective data for short- and long-term clinical management of asthma, it may be worthwhile to train young asthmatic children in spirometry twice daily during hospitalization, or to begin repeated training sessions from the age of 2.5 yr, as suggested earlier (19, 20, 22).

Our study was not intended to correlate disease and lung function or to establish tables of normal spirometric values for preschool children. Therefore, we used a nonvalidated parental report concerning prior symptoms, without a doctor's diagnosis or independent assessment. Differences between actual and extrapolated predicted values are of interest, but need to be validated in a large group of normal subjects. The anatomy of the airway in this age group may render extrapolation from older children inaccurate. Nevertheless, the overall similarity between SpiroGame measurements and extrapolated predicted values supports the general validity of this system.

Alternative methods of evaluating lung function in preschool children include the assessment of forced partial flow-volume curves (23, 24). In infants, a decrease in maximal flow at FRC (VmaxFRC) was identified as indicative of airway obstruction, but this parameter depends on the reproducibility of FRC and the presence of flow limitation, which have frequently been questioned (7). Furthermore, pharmacologic sedation and relaxation of the respiratory muscles are required for the examination. Reference values for VmaxFRC in preschool children are yet to be established. Tidal breathing measurements have been used to assess airway resistance in epidemiologic studies, but may be too nonspecific or insensitive for clinical use (7). Other methods applicable during quiet tidal breathing are the forced oscillation and the interrupter techniques (25, 26). Neither has so far been widely used in preschool children.

Unlike other incentive spirometry games on the market, the SpiroGame system teaches a full FVC maneuver in a step-by-step manner. We found that a significantly higher proportion of children succeeded in producing at least one adequate loop with this system. FEV₁ as a customary target variable was achieved by 80% of the children in our study in the form of an acceptable SpiroGame measurement, and by 54% of all children attempting SpiroGame. In contrast, the candle incentive led to slightly but significantly higher peak flow values but poor results for FEV₁, accompanied by premature expiratory breaks. This might be expected in a peak flow-targeted game. Although there was no significant difference in FVC for successful maneuvers with either game, FVC was significantly lower with candle-blowing than with SpiroGame when maneuvers with evidence of expiratory breaks were included.

Integration of all steps with SpiroGame into a full FVC maneuver greatly reduces inadequate loops by reducing the rate of incomplete expiration and premature expiratory breaks, which was the main problem observed with the candle-blowing game. In addition, SpiroGame probably enhances inhalation to full IC by using a separate visual incentive and thus producing IC values exceeding extrapolated predicted values. This prior maximal inhalation may have contributed to success in achieving FEV₁ measurements with SpiroGame (27, 28). Adequacy of prior inspiration to lung capacity could not be assessed with the candle-blowing game. Success with SpiroGame was achieved at the expense of a moderate increase in testing time, without exceeding the usual requirement for school children.

We conclude that computer-animated interactive programs, together with specially designed hardware, as in the SpiroGame system, facilitate successful, high-quality spirometric testing in preschool children. The concept that few preschool children can perform FVC maneuvers may require revision. With the SpiroGame system, children as young as 3 to 6 yr can learn to reliably and reproducibly perform a complete spirometric test within a reasonable time. SpiroGame may thus be applicable for clinical use in pediatric pulmonary function laboratories.