Work

Confirmatory cross-sectional validation of the Asthma Impairment and Risk Questionnaire

Year: 2023

Type: article

Abstract: Clinical ImplicationsAlthough asthma control is defined by impairment and risk, many control tools assess only symptoms. Confirmatory cross-sectional performance metrics demonstrate that the Asthma Impairment and Risk Questionnaire, a composite measure, is an accurate control assessment of all asthma severities for patients aged 12 years and older. Although asthma control is defined by impairment and risk, many control tools assess only symptoms. Confirmatory cross-sectional performance metrics demonstrate that the Asthma Impairment and Risk Questionnaire, a composite measure, is an accurate control assessment of all asthma severities for patients aged 12 years and older. Although asthma control is defined by symptom impairment and exacerbation risk,1National Asthma Education and Prevention ProgramThird Expert Panel on the Diagnosis and Management of Asthma. Expert panel report 3: guidelines for the diagnosis and management of asthma - full report. National Heart, Lung, and Blood Institute, Bethesda, Md2007Google Scholar many control questionnaires assess only impairment.2Nathan R.A. Sorkness C.A. Kosinski M. Schatz M. Li J.T. Marcus P. et al.Development of the Asthma Control Test: a survey for assessing asthma control.J Allergy Clin Immunol. 2004; 113: 59-65Abstract Full Text Full Text PDF PubMed Scopus (2156) Google Scholar,3Juniper E.F. O’Byrne P.M. Guyatt G.H. Ferrie P.J. King D.R. Development and validation of a questionnaire to measure asthma control.Eur Respir J. 1999; 14: 902-907Crossref PubMed Scopus (1911) Google Scholar The Asthma Impairment and Risk Questionnaire (AIRQ) is an equally weighted, 10-item, "yes"/"no" control tool that assesses symptoms over the prior 2 weeks and exacerbations over the prior 12 months. The AIRQ was cross-sectionally validated against a standard of the current Asthma Control Test (ACT) score plus prior 12-month, chart-documented exacerbations in a study of 442 patients with asthma of all severities who were aged 12 years and older.4Murphy K.R. Chipps B. Beuther D.A. Wise R.A. McCann W. Gilbert I. et al.Development of the Asthma Impairment and Risk Questionnaire (AIRQ): a composite control measure.J Allergy Clin Immunol Pract. 2020; 8: 2263-2274.e5Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar The 10 AIRQ items address symptoms, activity limitations, sleep, rescue medication use, social activities, exercise, difficulty controlling asthma, and exacerbations. Fewer "yes" responses indicate better asthma control (0-1 indicates well-controlled; 2-4, not well-controlled; and 5-10, very poorly controlled). The receiver operating characteristic (ROC) area under the curve (AUC) to identify well- versus not well- or very poorly controlled asthma was 0.94 and well-controlled or not well-controlled versus very poorly controlled asthma was 0.92. Cut points of 2 or more "yes" responses yielded a sensitivity of 0.90 to identify not well-controlled or very poorly controlled asthma, and 5 or more yielded a specificity of 0.96 to detect very poorly controlled disease. The objective of this analysis was to explore the consistency of these robust performance metrics by conducting a confirmatory validation of the previously determined AIRQ items and control cut points in differentiating current control levels in a separate patient cohort. The original AIRQ cross-sectional validation was an analysis of the first 442 patients enrolled from 12 specialty practices whose data were collected before August 9, 2019.4Murphy K.R. Chipps B. Beuther D.A. Wise R.A. McCann W. Gilbert I. et al.Development of the Asthma Impairment and Risk Questionnaire (AIRQ): a composite control measure.J Allergy Clin Immunol Pract. 2020; 8: 2263-2274.e5Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar The confirmatory convenience sample was composed of 400 patients enrolled after August 9, 2019, from a different selection of 12 specialty practices and one specialty-affiliated primary care site. All patients participated in a longitudinal study assessing the ability of AIRQ to predict future 12-month exacerbations.4Murphy K.R. Chipps B. Beuther D.A. Wise R.A. McCann W. Gilbert I. et al.Development of the Asthma Impairment and Risk Questionnaire (AIRQ): a composite control measure.J Allergy Clin Immunol Pract. 2020; 8: 2263-2274.e5Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar,5Chipps B.E. Murphy K.R. Wise R.A. McCann W.A. Beuther D.A. Reibman J. et al.Evaluating construct validity of the Asthma Impairment and Risk Questionnaire using a 3-month exacerbation recall.Ann Allergy Asthma Immunol. 2022; 128: 544-552.e3Abstract Full Text Full Text PDF PubMed Scopus (5) Google Scholar At enrollment, patients completed the AIRQ items assessed in the original cross-sectional study. Responses were evaluated against the same composite comparator for validation: baseline symptom control per ACT plus prior-year, chart-documented exacerbations.4Murphy K.R. Chipps B. Beuther D.A. Wise R.A. McCann W. Gilbert I. et al.Development of the Asthma Impairment and Risk Questionnaire (AIRQ): a composite control measure.J Allergy Clin Immunol Pract. 2020; 8: 2263-2274.e5Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar At study entry, each patient’s physician completed a chart abstraction form for clinical data and prior-year exacerbations.4Murphy K.R. Chipps B. Beuther D.A. Wise R.A. McCann W. Gilbert I. et al.Development of the Asthma Impairment and Risk Questionnaire (AIRQ): a composite control measure.J Allergy Clin Immunol Pract. 2020; 8: 2263-2274.e5Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar This study was conducted with informed consent or assent and institutional approval, and was consistent with recognized ethical standards. We used descriptive statistics to characterize patients in the confirmatory convenience group by sociodemographic and clinical features and compare them with patients in the cross-sectional validation group. Logistic regression analyses on the confirmatory sample data were run for the 10 cross-sectionally validated AIRQ items to examine reproducibility of AIRQ performance metrics for identifying current asthma control in two models. Model 1 evaluated well-controlled versus not well-controlled or very poorly controlled asthma, and model 2 evaluated well-controlled or not well-controlled versus very poorly controlled asthma. We compared the ROC curves of the cross-sectional and confirmatory sample analyses using χ2 test. The sensitivity and specificity of the cut points determined by the previous cross-sectional validation (two or more "yes" responses for not well-controlled or very poorly controlled vs well-controlled asthma, and five or more responses for very poorly controlled vs well-controlled or not well-controlled asthma) were assessed for the confirmatory sample. The confirmatory convenience sample included 400 patients; the mean (SD) age was 45.5 (18.7) years; 48 (12.0%) were adolescents (see Table E1 in this article’s Online Repository at www.jaci-inpractice.org). There were no differences between the confirmatory convenience and cross-sectional samples in symptom-based control or lung function at entry, as reflected by mean (SD) ACT score (18.6 [4.5] vs 18.4 [4.6]; P = .61), mean (SD) percentage of predicted prebronchodilator FEV1 (83.8% [17.4%] vs 85.3% [16.9%]; P = .20), and mean (SD) percentage of postbronchodilator FEV1 (88.0 [17.4%] vs 88.7% [17.0%]; P = .55), respectively. Baseline AIRQ control level distributions were also similar for patients in the confirmatory convenience (38.8% well-controlled, 35.5% not well-controlled, and 25.8% very poorly controlled) and cross-sectional (31.9% well-controlled, 38.9% not well-controlled, and 28.5% very poorly controlled) groups (P = .08). However, patients within the confirmatory convenience sample differed from those in the cross-sectional validation sample4Murphy K.R. Chipps B. Beuther D.A. Wise R.A. McCann W. Gilbert I. et al.Development of the Asthma Impairment and Risk Questionnaire (AIRQ): a composite control measure.J Allergy Clin Immunol Pract. 2020; 8: 2263-2274.e5Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar regarding sex (65.8% vs 72.6% female [P = .03]), race (19.8% vs 8.1% Black [P < .001] and 68.3% vs 84.4% White [P < .001]), and ethnicity (11.3% vs 2.9% Hispanic [P < .001]). In addition, fewer patients in the confirmatory convenience sample had an exacerbation in the past year compared with the cross-sectional validation sample (37.5% vs 44.6% [P = .04]), and the prescribed Global Initiative for Asthma (GINA)6Global Initiative for AsthmaGlobal strategy for asthma management and prevention.http://www.ginasthma.orgDate: 2018Date accessed: December 12, 2022Google Scholar step levels of asthma therapy were significantly different among groups (12.0% vs 6.6% GINA 1, 11.0% vs 14.7% GINA 2, 13.0% vs 11.3% GINA 3, and 64.0% vs 67.4% GINA 4/5 [P = .02]). Both models performed well, exceeding the model-fit criteria. For model 1 (well-controlled vs not well-controlled or very poorly controlled), there was no difference in ROC AUC for the confirmatory versus cross-sectional samples (0.91 vs 0.94 [P = .12]) (Figure 1). For model 2 (well-controlled or not well-controlled vs very poorly controlled), the ROC AUC was lower for the confirmatory versus cross-sectional samples (0.88 vs 0.92 [P = .03]) (Figure 2). However, the 10 AIRQ items still provided excellent discriminatory ability.7Mandrekar J.N. Receiver operating characteristic curve in diagnostic test assessment.J Thorac Oncol. 2010; 5: 1315-1316Abstract Full Text Full Text PDF PubMed Scopus (1885) Google ScholarFigure 2Asthma Impairment and Risk Questionnaire receiver operating characteristic (ROC) curves comparing confirmatory convenience sample (dashed line) and cross-sectional validation sample (solid lines) using model 2 (well- or not well-controlled vs very poorly controlled). AUC, area under the curve.View Large Image Figure ViewerDownload Hi-res image Download (PPT) For the confirmatory patient group, an AIRQ score cut point of two or more "yes" responses yielded a sensitivity of 0.85 to identify not well-controlled or very poorly controlled asthma, and a cut point of five or more yes responses yielded a specificity of 0.94 to detect very poorly controlled disease. These results were clinically similar to sensitivity and specificity performance metrics in the original AIRQ cross-sectional validation: 0.90 and 0.96, respectively. This confirmatory validation of AIRQ items and cut points demonstrates that the AIRQ is a rigorous composite measure that accurately differentiates asthma control among demographically and clinically diverse patients aged 12 years and older. As with the developmental validation studies for the ACT2Nathan R.A. Sorkness C.A. Kosinski M. Schatz M. Li J.T. Marcus P. et al.Development of the Asthma Control Test: a survey for assessing asthma control.J Allergy Clin Immunol. 2004; 113: 59-65Abstract Full Text Full Text PDF PubMed Scopus (2156) Google Scholar and the Asthma Control Questionnaire,8Olaguibel J.M. Quirce S. Juliá B. Fernández C. Fortuna A.M. Molina J. et al.Measurement of asthma control according to Global Initiative for Asthma guidelines: a comparison with the Asthma Control Questionnaire.Respir Res. 2012; 13: 50Crossref PubMed Scopus (79) Google Scholar almost all patients in this study were recruited from specialty sites. However, race and ethnicity profiles for the ACT and Asthma Control Questionnaire analyses were unspecified, and the validity of ACT control-level cut points in Black adolescents with asthma is uncertain.9Burbank A.J. Todoric K. Steele P. Rosen J. Zhou H. Frye M. et al.Age and African-American race impact the validity and reliability of the Asthma Control Test in persistent asthmatics.Respir Res. 2018; 19: 152Crossref Scopus (3) Google Scholar Despite the differences in sex between the original AIRQ cross-sectional validation and confirmatory study patient groups,7Mandrekar J.N. Receiver operating characteristic curve in diagnostic test assessment.J Thorac Oncol. 2010; 5: 1315-1316Abstract Full Text Full Text PDF PubMed Scopus (1885) Google Scholar the latter of which encompassed a more representative inclusion of Black patients and Hispanic patients, the performance metrics of AIRQ were similar, with outstanding or excellent differentiation based on ROC AUC values.7Mandrekar J.N. Receiver operating characteristic curve in diagnostic test assessment.J Thorac Oncol. 2010; 5: 1315-1316Abstract Full Text Full Text PDF PubMed Scopus (1885) Google Scholar The AIRQ provides a rigorously validated composite asthma control measure for patients aged 12 years and older that evaluates both control domains in a single instrument. Although the current study used the AIRQ with a 12-month recall for the risk assessment items, the validated follow-up AIRQ with a 3-month recall for the risk items may be useful in studies or clinical assessments that are more frequent than yearly.5Chipps B.E. Murphy K.R. Wise R.A. McCann W.A. Beuther D.A. Reibman J. et al.Evaluating construct validity of the Asthma Impairment and Risk Questionnaire using a 3-month exacerbation recall.Ann Allergy Asthma Immunol. 2022; 128: 544-552.e3Abstract Full Text Full Text PDF PubMed Scopus (5) Google Scholar Further analyses of AIRQ validity and effectiveness in clinical practice among patients in primary care and of varying socioeconomic, educational, racial, and ethnic backgrounds are ongoing. We thank Karin S. Coyne, PhD, of Evidera, for critical review of the manuscript; Melissa Ross, PhD, of Evidera, for project management of this study; and Ren Yu, MA, of Evidera, for statistical programming. Medical writing support was provided by Shravanthi Chidambaram, PhD, of Lumanity Communications Inc. (Yardley, Pa), which was in accordance with Good Publication Practice (GPP 2022) guidelines and funded by AstraZeneca (Wilmington, Del). The data that support the findings of this study are available from the corresponding author upon reasonable request. Table E1Sociodemographic and clinical characteristics of confirmatory convenience sample compared with cross-sectional validation sampleCharacteristicConfirmatory convenience sample (n = 400)Cross-sectional validation sampleE2Global Initiative for AsthmaGlobal strategy for asthma management and prevention.http://www.ginasthma.orgDate: 2018Date accessed: December 12, 2022Google Scholar (n = 442)PSociodemographic characteristics Mean age, y (SD)45.5 (18.7)44.0 (20.1).27 Adolescent, n (%)48 (12.0)77 (17.4).08 Female sex, n (%)263 (65.8)321 (72.6).03 Race, n (%)American Indian or Alaska Native3 (0.8)1 (0.2).53Asian8 (2.0)8 (1.8).95Black79 (19.8)36 (8.1)<.001White273 (68.3)373 (84.4)<.001Other∗Other race includes African American and Native Hawaiian/Pacific Islander; Asian and Native Hawaiian/Pacific Islander; biracial: African American and Caucasian; French, German, English, and some American Indian; Hispanic; Mexican; White, and African American (n = 4); White and American Indian (n = 2); White and American Indian/Alaska Native (n = 2); White and Asian (n = 2); White, American Indian/Alaska Native, and Spanish; and White, Asian, and American Indian; and White, Spanish, and other.30 (7.5)20 (4.5).19 Ethnicity, n (%)Hispanic45 (11.3)13 (2.9)<.001Clinical characteristics Baseline ACT score (mean [SD])18.6 (4.5)18.4 (4.6).61 Baseline ACT category, n (%).72Well-controlled196 (49.0)206 (46.6)Not well-controlled105 (26.3)111 (25.1)Very poorly controlled96 (24.0)121 (27.4)Missing3 (0.8)4 (0.9)≥1 exacerbation in prior year, n (%)150 (37.5)197 (44.6).04FEV1 (% predicted) prebronchodilator (mean [SD])83.8 (17.4)85.3 (16.9).20FEV1 (% predicted) postbronchodilator (mean [SD])88.0 (17.4)88.7 (17.0).55GINA step therapy, n (%)†As per GINA 2018 report.E2.02 GINA 148 (12.0)29 (6.6) GINA 244 (11.0)65 (14.7) GINA 352 (13.0)50 (11.3) GINA 4/5256 (64.0)298 (67.4)Biologics, n (%)43 (10.8)41 (9.3).48Asthma Impairment and Risk Questionnaire control group, n (%).08 Well-controlled155 (38.8)141 (31.9) Not well-controlled142 (35.5)172 (38.9) Very poorly controlled103 (25.8)126 (28.5) Missing03 (0.7)ACT, Asthma Control Test; GINA, Global Initiative for Asthma.∗ Other race includes African American and Native Hawaiian/Pacific Islander; Asian and Native Hawaiian/Pacific Islander; biracial: African American and Caucasian; French, German, English, and some American Indian; Hispanic; Mexican; White, and African American (n = 4); White and American Indian (n = 2); White and American Indian/Alaska Native (n = 2); White and Asian (n = 2); White, American Indian/Alaska Native, and Spanish; and White, Asian, and American Indian; and White, Spanish, and other.† As per GINA 2018 report.E2Global Initiative for AsthmaGlobal strategy for asthma management and prevention.http://www.ginasthma.orgDate: 2018Date accessed: December 12, 2022Google Scholar Open table in a new tab ACT, Asthma Control Test; GINA, Global Initiative for Asthma.

Source: The Journal of Allergy and Clinical Immunology In Practice

DOI: 10.1016/j.jaip.2023.07.018