Evaluation of the questionnaires’ validity in assessing the severity of idiopathic pulmonary fibrosis in correlation with high-resolution computed tomography, lung diffusion, and cardiopulmonary exercise tests

Introduction Health-related quality-of-life questionnaires need to be incorporated into the evaluations of idiopathic pulmonary fibrosis (IPF) patients to assess their influence. Aim The aim of the study was to evaluate the validity of generic and specific questionnaires in assessing the severity of IPF in correlation with high-resolution computed tomography (HRCT), diffusion lung capacity for carbon monoxide (DLCO), and cardiopulmonary exercise testing (CPET). Patients and methods Forty stable IPF patients were prospectively recruited and categorized on the basis of spirometry, DLCO, HRCT, and CPET. The results were correlated with a generic International Physical Activity Questionnaire (IPAQ) and a Specific Saint George Respiratory Questionnaire (SGRQ). Results IPF patients showed restrictive pattern with impairment of diffusion capacity (forced vital capacity (FVC)=56±14.8% and DLCO=48.5±20% of predicted value) with a total semiquantitative scoring of HRCT 16.6±8. The mean total score of the SGRQ questionnaire for all studied cases was 56.5+21 and categorical scoring of IPAQ showed that 45, 42.5, and 12.5% of patients were in moderate, severe, and mild categories, respectively. There was a negative correlation between the total score of SGRQ and VO2max (ml/kg/min) (maximum oxygen consumption) (r=−0.35) and VE’ (l/min) (minute ventilation) (r=−0.39) on CPET, as well as with DLCO (r=−0.53), and a positive correlation with HRCT score (r=0.63). There was a highly significant correlation between IPAQ and VO2max (χ2=28), VE’ (χ2=14.8) and desaturation percentage variables of CPET, DLCO (r=0.61), and HRCT score (r=−0.68). Conclusion Correlations between physiological parameters including DLCO and CPET, radiological parameters in the form of HRCT, and health-related quality-of-life assessment using SGRQ and IPAQ were strong and it was possible to distinguish IPF patients with severely impaired lung functions.


Introduction
Idiopathic pulmonary fibrosis (IPF) is defined as a specific form of chronic fibrosing interstitial pneumonia limited to the lung and associated with histological appearance of usual interstitial pneumonia on surgical lung biopsy [1]. High-resolution computed tomography (HRCT) of the chest has changed the diagnostic evaluation and commonly shows patchy, predominantly peripheral, subpleural, bibasal reticular abnormalities [2]. Pulmonary function tests are used to quantify disease severity, monitor disease progression, and identify variables most strongly predictive of mortality.
Health-related quality-of-life (HRQL) questionnaires have been applied to quantify average changes in health and the effects on the patient's daily life in several types of chronic lung diseases. Two kinds have been used: generic and disease specific. Both can be used in IPF [3].

Aim
The aim of this study was to evaluate the validity of the generic International Physical Activity Questionnaire (IPAQ) and Specific Saint George Respiratory Questionnaire (SGRQ) in assessing the severity of IPF in correlation with HRCT, diffusion lung capacity for carbon monoxide (DL CO ), and cardiopulmonary exercise testing (CPET).

Patients and methods
This prospective study included 40 consecutive stable IPF patients on steroid dose 10 mg or less who were diagnosed according to radioclinical criteria mentioned in American Thoracic Society (ATS) 2011 [4]. The patients were selected from either the outpatient chest clinic or inpatient wards of Ain Shams University hospitals during the period from August 2011 to August 2012. Smokers or ex-smokers, patients with other interstitial lung diseases (ILDs), patients with comorbid disease that precluded exercise (such as severe orthopedic or neurological deficits or history of syncope on exertion or unstable cardiac disease), and patients with cardiac, renal, hepatic diseases, or diabetes mellitus were excluded. All patients gave their formal consent. The protocol was approved the Ethical commiitte of the Ain Shams University.
All patients were subjected to the examinations given below.
High-resolution computed tomography of the chest HRCT of the chest was performed on a General Electric multislice 4 multidetector (General Electric) with a scanning time of 5 s so that scans with 1-2 mm collimation were obtained at 10 mm intervals from the lung apices to the bases at full inspiration with breathholding after deep inspiration and the patient supine. A semiquantitative scoring system (Table 1) was then used to evaluate the correlation between HRCT findings and the clinical severity of disease [6].
Cardiopulmonary exercise testing CPET was performed using a standardized protocol in accordance with the ATS/American College of Chest Physicians statement [7]. All patients underwent a symptom-limited CPET with an electromagnetically braked cycle ergometer (Ergometrics 900; Erich Jaeger GmbH) using a ramp protocol. Cardiopulmonary data were collected and analyzed with an exercise metabolic unit (Master Screen-CPX; Viasys, Germany). The following parameters were recorded: heart rate (HR), minute ventilation (VE), maximum oxygen consumption (VO 2 max), VO 2 max/kg, ventilatory equivalent for oxygen (VE/VO 2 ) and for carbon dioxide (VE/VCO 2 ), anaerobic threshold (AT), VE/ VCO 2 at AT, respiratory rate, oxygen pulse (VO 2 / HR), breathing reserve, and HR reserve; these parameters were measured at the highest comparable workload for each patient. The ratio of total dead space to tidal volume (V d /V t ) was also computed.
Quality-of-life assessment by two different questionnaires (Saint George Respiratory questionnaire and International Physical Activity Questionnaire) The Arabic version of the SGRQ (respiratory-specific questionnaire) [9] was used, which consists of three components: symptoms, which measure respiratory symptoms; activities, which measure impairment of mobility or physical activity; and impact, which measures the psychosocial impact of disease. Scores for these components and the summary score are on a 100-point scale calculated by an Excel scoring calculator. Higher scores correspond to worse HRQL. The Arabic version of the IPAQ (nonspecific questionnaires) [10] was used, which consists of four generic items (leisure time physical activity, domestic and gardening activities, and work-related physical activity and transport-related physical activity). The questionnaire provided separate scores on walking and on moderate and vigorous activities; categorical scoring was then done (category 1 or mild, category 2 or moderate, and category 3 or severe). Both questionnaires were administered and scored according to the instruction manual before the execution of the CPET and lung function tests.

Statistical analysis
Data were analyzed and tabulated by an IBM computer using statistical program for the social sciences, version 12 (SPSS; SPSS Inc., Chicago, Illinois, USA). Quantitative variables were presented as mean, SD, and range; qualitative variables were presented as number and percentage; the χ 2 -test was used to compare qualitative variables; the Spearman correlation was used to rank variables positively or inversely versus each other in the same group; receiver operator characteristic curve was used to find out the best cutoff value of a certain predictor and its validity parameters (sensitivity, specificity, positive predictive value, negative predictive value, and accuracy). P-value was considered insignificant if more than 0.05 and significant if up to 0.05.

Results
The population studied consisted of 40 patients. Totally, 20 patients were male and 20 were female, with a mean age of 54±14 (range: 34-81) years. Every study participant underwent spirometry, single-breath carbon monoxide diffusing capacity evaluation, and HRCT of the chest. The mean value of FVC% was 56±14.8. The mean value of DL CO % of predicted was 48.5±20. On the basis of the severity of diffusion impairment, 14 (35%) patients were graded mild (grade B), 15 (37.5%) patients were graded moderate (grade C), and the remaining 11 (27.5%) patients were graded severe (grade D).
The semiquantitative scoring of HRCT revealed a mean total score of 16.6±8 (range: 6-38) for all studied cases. The mean scores of ground-glass opacity, reticular fibrosis, mixed ground-glass opacity, and reticular fibrosis and honeycombing were 0.7±1.5, 7.9±4, 2.5±2, and 5.5±4, respectively. The correlation between CPET and SGRQ is shown in Table 3. There was an inverse correlation between the total score of SGRQ and both VO 2 max (Fig. 1) and VE' on the Spearman correlation test. There was an inverse correlation between VO 2 max and both symptom and activity domains of SGRQ. There was a significant inverse correlation between AT and the symptom domain of SGRQ. Further, there was an inverse correlation between VE' and the impact domain of SGRQ. The correlation between CPET AT, anaerobic threshold; CPET, cardiopulmonary exercise testing; VE', minute ventilation; VO 2 max, maximum oxygen consumption. a Grading was according to reference [8].
and IPAQ is presented in Table 4. There was a highly significant association between IPAQ and VO 2 max, VE', and the desaturation percentage variables of CPET.
The correlation between HRCT and SGRQ is illustrated in Table 5. There was a positive correlation between total score of SGRQ and both honeycombing and reticular fibrosis patterns of HRCT, whereas there was an inverse correlation between the total score of SGRQ and mixed ground-glass opacification/reticular fibrosis pattern of HRCT on Spearman's correlation test. IPAQ showed an inverse correlation versus honeycombing and reticular fibrosis patterns of HRCT and positive correlations with mixed ground-glass opacification/ reticular fibrosis and ground-glass opacification patterns of HRCT on the Spearman correlation test (Table 6).
There was an inverse correlation between DL CO and the total score of SGRQ and most domains of SGRQ and a positive correlation between DL CO and IPAQ on the Spearman correlation test (Tables 7  and 8).
Finally, SGRQ and IPAQ were considered better positive than negative in the prediction of severity of IPF, with an overall sensitivity of 81 and 96%, an overall specificity of 60 and 40%, and an overall accuracy of 65 and 72.5%, respectively (Tables 9 and  10, Fig. 2).  Inverse correlation between total score of the Saint George Respiratory Questionnaire and maximum oxygen consumption  Saint George Respiratory Questionnaire as a better positive than negative predictor of severity

Discussion
Our results showed that generic (IPAQ) and specific (SGRQ) questionnaires had a good relationship with HRCT, DL CO , and CPET.
Incorporation of HRQL questionnaires into the routine evaluations of IPF patients seems beneficial, as they assess several dimensions (such as the influence of disease on physical, emotional, and social functioning) that are not estimated by traditional methods of clinical assessment [11]. Generic (IPAQ) and specific (SGRQ) questionnaires were used to test their validity in assessing the severity of IPF in correlation with HRCT, DL CO , and CPET. Although both questionnaires are self-administered instruments, we preferred to do the interview ourselves in this study. We made this choice because the patient population seen in our clinic or department has low cultural and social status, and a substantial number of patients and controls had difficulty reading.
The face and content validity of the SGRQ in ILD has been previously tested, and while some authors raised some concerns about its face validity, especially as far as the questions included in the 'symptoms' domain (i.e. those more specific for COPD-related manifestations) were concerned, others had the instrument rate by ILD patients for relevance, obtaining a highly positive feedback [3]. The IPAQ was developed to facilitate surveillance of physical activity based on a global      standard and has become the most globally used instrument to assess the quality of life. It has suggested that nonspecific disease questionnaires are sensitive and can be applied to IPF patients [12,13].
This study subjected all patients to spirometry, singlebreath carbon monoxide diffusing capacity evaluation, and HRCT of the chest. Our results revealed impaired mean FVC% of predicted (56±14.8) and mean DL CO % of predicted (48.5±20), together with peculiar HRCT findings, all of which were consistent with IPF. It is becoming increasingly accepted that a highly suggestive clinical presentation, including typical HRCT scan findings, can be used in the absence of a lung biopsy specimen to make a likely diagnosis of IPF [14]. Blivet et al. [15] agreed with our results, as they showed a restrictive ventilatory impairment in IPF with a mean FVC of 55% and impaired DL CO of 55% of predicted.
The correlations between physiological parameters including DL CO and CPET, radiological parameters in the form of HRCT; and quality-of-life assessment methods using SGRQ and IPAQ questionnaires were highly significant and were able to identify patients with severely impaired lung functions in our study.
Peng et al. [16] also showed a significant correlation between total CT scores and each component of the SGRQ, and changes in ground-glass opacity on CT were also correlated with changes in each SGRQ domain, similar to our study results.
CPET as a noninvasive tool is being increasingly used in a wide spectrum of clinical applications on the basis of the assumption that it provides a global assessment of the integrative submaximal and peak exercise responses that are not adequately reflected through the measurement of individual organ system function thus providing fruitful information for clinical decision making [7].
CPET with gas exchange parameters including ventilatory equivalent for O 2 (VE'/VO 2 ) and ventilatory equivalent for CO 2 (VE'/VCO 2 ) was used to calculate an AT, which had the potential of noninvasively grading the severity of exercise limitation in patients along with other ventilatory parameters like VE' and metabolic parameters like VO 2 max. O 2 pulse acted as a guide for quantifying the pulmonary vascular blood flow and for assessing future responses to therapy even before overt right ventricular failure and pulmonary hypertension became evident at rest with a condition of having normal EF as well as increased levels of V d /V t [17].
Correlating exercise capacity [using the 6 min walking test (SMWT)] and other parameters to the quality-of-life questionnaires was done in many studies, all of which showed significance [3,18]. Hence we tried to find a correlation using CPET by means of an electronically braked cycle ergometer using the incremental work rate, and after the completion of the primary assessment for the patients enrolled in the study the two previously mentioned questionnaires were compared with our results, which showed that the IPAQ had a better validity than the SGRQ as regards desaturations occurring with exercise and its ability to discriminate between patients with more severely impaired exercise function and dyspnea as a reason for stopping exercise. As for VO 2 max, there was a significantly inverse correlation to SGRQ, especially activity and symptom domains, which signified that the worse the VO 2 max, the higher the SGRQ score; in contrast, IPAQ showed highly significant direct correlation. Although the prevalence of leg discomfort as the limiting factor in IPF is unknown, it is likely to be significant. One-third of our patients stopped exercise during CPET because of leg discomfort and not because of mechanical constraint.
Nogueira et al. [19] detected a correlation between quality of life and CPET VO 2 max and AT in cardiac failure cases by analyzing both a generic Medical Outcomes Short Form 36 (SF-36) questionnaire and a specific Minnesota Living with Heart Failure questionnaire. They observed that the Minnesota Living with Heart Failure questionnaire showed a mild to moderate correlation with VO 2 max and AT. Their results matched ours in cases of IPF regarding significance of correlations between VO 2 max, AT, and HRQL questionnaires [19].
Chou et al. [20] agreed with our results regarding the peak oxygen consumption in patients with rheumatoid arthritis-ILD, which correlated with the results of physical wellness, mental wellness, and environmental health with P values of less than 0.001, 0.01, and 0.04, respectively.
Beretta et al. [18] showed a good correlation of questionnaires with the SMWT and HRCT scores and other validated measures for perceived breathlessness and FVC. The strongest correlations were found between the SMWT and SGRQ scores, with an almost linear association with the activity score. The study observed that the SGRQ, although not specifically designed for ILD, proved to be a valid respiratory-specific questionnaire for the evaluation of HRQL, strongly correlating with standardized tools [18].
The process of validating a questionnaire designed to measure impaired health is multifactorial. Evidence for the validity of such instruments is built up from a large number of tests of the relationship between the questionnaire and relevant measures of disease activity and its effects on the patient's health and well-being. We have shown that the SGRQ and IPAQ correlated with a number of measures of disease severity and activity relevant to patients with IPF. The pattern of these correlations with the components of the SGRQ and IPAQ suggested that these components were addressing relatively specific areas of impaired health in this population. The results of this direct comparison of SGRQ and IPAQ have confirmed their feasibility for use in patients with more severe IPF. The validity of both quality-of-life questionnaires was supported by the correlations between their total scores and their domain scores. Almost all cross-sectional and longitudinal correlations were highly statistically significant. The overall patterns of correlations were consistent with expectations. Lee et al. [21], studying the validity of IPAQ, showed fairly weak evidence to support the use of the IPAQ as either a relative, or as an accurate and absolute, measure of physical activity, although its proven reliability shows it can be used with care in repeated-measures studies, although the true magnitude of the change over time, if any, may not be accurate. This difference between our results and their results may be accounted for by the lack of demographic characteristics in this study, including place of study, targeted population, sample size, male-female ratio, and age, and seemed to be related to differences in validity between the IPAQ and the criterion measure. Our results concurred with those of Beretta et al. [22], who showed that SGRQ, although not specifically designed for ILD, proved to be a valid respiratory-specific questionnaire for the evaluation of HRQL in those patients, strongly correlating with standardized tools to measure the patient's lung involvement, its physical tolerance to exertion or perceived breathlessness, and providing additional information to traditional measures of lung involvement. Zimmermann et al. [3] showed that a specific rather than a generic questionnaire is a more appropriated and specific instrument to evaluate HRQL in IPF patients and that goes with our results. They also noticed that total SGRQ domains presented a significant correlation with FVC and DL CO as well as with SMWT and rest dyspnea index. Good correlations were also observed between lung function, exercise capacity, and rest dyspnea index and all SGRQ domains, except the symptom domain [3]. Swigris et al. [23] agreed that SGRQ was a sensitive instrument for evaluating HRQL in patients with ILD. Patients in their study reported substantially impaired HRQL, especially in domains that measured physical health and level of independence [23]. Tzanakis et al. [24] suggested that dyspnea scales and the SGRQ are sensitive tools for assessing HRQL in patients with IPF. Chang and colleagues also showed that generic questionnaires and specific ones are good tools for measuring HRQOL or health status in patients with ILD [13,25].
In the light of our study results, it was concluded that the correlations between physiological parameters including DL CO and CPET, radiological parameters in the form of HRCT, and quality-of-life assessment methods using SGRQ and IPAQ were highly significant and can discriminate IPF patients with severely impaired lung functions.

Financial support and sponsorship
Nil.

Conflicts of interest
There are no conflicts of interest.