Determining the Best Tool Comparable with Global Initiative for Asthma Criteria for Assessing Pediatric Asthma Control

Abstract

Background: Guidelines such as Global Initiative for Asthma (GINA) recommend disease control as the mainstay of asthma management.

Objective: To investigate which measure of asthma control best correlates with the GINA criteria for determining asthma control in children.

Methods: Child asthma-patients at a tertiary hospital were enrolled in the study after evaluation of response to treatment. Asthma control test (ACT)/pediatric asthma control test (PACT), Pediatric Asthma Quality of Life Questionnaire (PAQLQ), fractional exhaled nitric oxide (FeNO), and lung function parameters were evaluated. Patients were examined by asthma specialists and control status was evaluated based on GINA.

Results: The median age (interquartile range) of patients was 10.7 (8.4–12.9) years, 57.9% of patients were boys. Of 228 children, 84.2%, 9.6%, and 6.1% displayed “well-controlled”, “partially controlled”, and “uncontrolled” asthma, respectively, according to GINA. The patients with “partially controlled” and “uncontrolled” asthma were grouped as “not well-controlled.” The cutoff levels were 22, 21, and 5.9 for PACT, ACT, and PAQLQ, respectively, for determining “well-controlled” asthma (P < 0.001). With these cutoff values, ACT exhibited higher comparability with GINA than PACT and PAQLQ (κ = 0.473, 0.221, and 0.150, respectively, P < 0.001). PAQLQ had higher agreement with GINA criteria in children ≥12 years old (κ = 0.326, P < 0.001 and κ = 0.151, P = 0.014, respectively). Correctly classified patients with PACT, ACT, and PALQLQ based on GINA with these cutoff levels were 93 (64.1%), 63 (75.9%), and 139 (62.9%), respectively. FeNO and lung function parameters were unsuccessful at revealing control status according to GINA.

Conclusion: ACT is better than PACT for comparability with GINA. Better correlation of PAQLQ and ACT and better comparability of PAQLQ and GINA were evident in older children.

Introduction

Asthma is one of the most common chronic diseases in childhood.¹ It is a heterogeneous inflammatory disease of the airways which is characterized by recurrent wheezing episodes, cough, dyspnea, and chest tightness during daily activity and sleep.¹ Childhood asthma has a wide variation in age of onset, type, and frequency of respiratory symptoms, exacerbation rates, triggers, lung function parameters, comorbidities, and underlying inflammatory patterns. Therefore, an individualized approach may be needed to improve asthma outcome and reduce future risk of exacerbations, decline in lung function parameters, and side effects of therapy.²

Since the 2000s, international asthma guidelines, such as the Global Initiative for Asthma (GINA)¹ and the National Asthma Education and Prevention Program,³ recommend achieving “control” in asthma, defined as having achieved effective management of symptoms, nocturnal awakening, reliever use, lung function parameters, and activities of daily living. The level of asthma control is the extent to which the manifestations of asthma can be observed in the patient, or have been reduced or removed by controller treatment recommended by GINA.¹ Despite access to controller therapy, a considerable proportion of children obtain suboptimal asthma control.⁴ Despite published treatment guidelines, numerous children suffer from sleep disturbances, exercise intolerance, and school absenteeism and require unscheduled health care visits and even hospitalization.⁵ Hence, there is a need for validated, noninvasive instruments to easily assess asthma control in children, which should correlate with asthma control as defined by respected asthma guidelines.

Several validated instruments have been developed to guide asthma specialists in determining the degree of asthma control and making modifications to the therapeutic plan. The asthma control test (ACT) and the pediatric asthma control test (PACT) are 2 instruments for assessing symptom control which are recommended by GINA.¹ Both are validated and reliable tests used worldwide. Nevertheless, the comparability of ACT and PACT, including their respective cutoff values, can be difficult to assess. Both instruments reflect the patient's and/or caregiver's perception of asthma and may differ to the symptom-based control suggested by GINA.^6–10

Although lung function parameters do not correlate strongly with asthma symptoms in children, forced expiratory volume in 1 s (FEV₁) is recommended by GINA to be used during asthma follow-up visits, as FEV₁ has been found to correlate with both asthma control and future exacerbations.¹

Fractional exhaled nitric oxide (FeNO), a marker of airway inflammation, is noninvasive, reproducible, and relatively easy to detect in patients.¹¹ Elevated FeNO levels during loss of asthma control and decreased FeNO after treatment with corticosteroids imply that FeNO may be useful not only in predicting asthma exacerbations but also in monitoring the response to treatment.¹² This led many researchers to suggest that FeNO could be utilized as a tool for the assessment of asthma control; however, this method needs further clarification, particularly in children.¹¹

The Pediatric Asthma Quality of Life Questionnaire (PAQLQ) was developed as a self-administered test by Juniper et al. in 1996.¹³ It comprises 3 domains, symptoms, emotional functions and activity limitation. Using the PAQLQ, it can be shown that children and adolescents with poor asthma control will have more frequent symptoms, medication use, and activity limitation resulting in health-related reduction in quality of life.^14,15

Most studies manifested diverse results with different tools, because they were performed in numerous countries and care settings, with a range of study plans and methods.^6–9,16 Nevertheless, clinicians working with children need to know which tools can be used to reliably assess asthma control in real-world settings. Based on this requirement, we planned this study to evaluate comparability of the 5 commonly used tools for assessing asthma control (ACT, PACT, PAQLQ, FeNO, and lung function parameters) with the GINA criteria.

Materials and Methods

Study population

Between January 2012 and January 2014, asthma patients, aged 6–18 years old, were invited to participate in the study with follow-up appointments to be carried out at the Pediatric Allergy Department (outpatient clinic) of a tertiary research and training hospital. Study inclusion criteria necessitated physician diagnosed asthma based on those criteria recommended by the GINA guidelines and the U.S. National Asthma Education and Prevention Program.^1,3 Additional prerequisites were adherence to prescribed asthma treatment and the ability to perform accurate FeNO and spirometry measurements according to the guidelines. Patients who did not fulfill all inclusion criteria or who had comorbid diseases other than allergic diseases were excluded. Patients who underwent acute asthmatic attacks and respiratory infections, which were unfavorable for FeNO and spirometry tests, were also excluded from the study. The study was approved by the Institute's Committee on Human Research (12.20.2011-033). Parents/carers of patients were asked to submit written informed consent.

Study design

Before study recruitment, patients were evaluated for adherence to treatment and inhalation technique. Both patients and carers were requested to demonstrate their use of inhaler therapy. They were also given recommendations about regular use of controller therapy and correct inhalation technique. This initial evaluation also included form completion for ACT/PACT, PAQLQ, and measurement of FeNO, and lung function parameters. Patients who completed this initial visit successfully were enrolled in the study. The period for stable antiasthmatic therapy before enrolling the subjects into the study was 4 weeks. A diary was given to each patient to note symptoms and the controller/rescue treatment that they received daily. Patients who missed controller therapy for >2 days a week were considered to be noncompliant. Study participants were invited to attend the main visit 4 weeks later.

During the main visit, children/carers completed ACT or PACT according to their age group and PAQLQ. The second step was composed of FeNO measurement and spirometric evaluation, respectively. Patients were then evaluated by the same pediatric asthma physicians (blinded experiment with respect to PACT/ACT scores and FeNO results). Physicians determined the “control status” of patients according to the GINA guidelines via interviews with the patients/carers in concurrence with reference to the completed asthma diaries. Descriptive terms were elucidated by carers and the medical records were also used for reference.

Assessment of asthma control

As recommended by the GINA guidelines,¹ patients who displayed no daily asthma symptoms, no night-time awakening, did not require reliever therapy and suffered no activity limitation due to asthma over the previous 4 week period were classified in the category of “well-controlled asthma.” Patients displaying 1–2 of these 4 described criteria at least twice per week were classified “partly controlled asthma.” Patients who exhibited 3–4 of the above criteria were grouped as “uncontrolled asthma.” “Partly controlled” and “uncontrolled” groups were collectively defined as “not well-controlled.” Therefore, the analyses were performed for the groups labeled “well-controlled” and “not well-controlled” asthma.⁶

Skin prick test

Skin prick tests (SPT) were performed on all study groups, on the volar aspect of the forearm. Common inhaled allergens (house dust mites, cockroach, animal danders, fungi, mixed grass and tree pollens) were tested. All SPTs were performed using commercial extracts (Laboratoire Stallergenes, France). Histamine (10 mg/mL) was used as positive control. Reaction was evaluated 15–20 min after allergens were applied. Test results were defined as positive if the wheal diameter of the test area was at least 3 mm greater than that of the negative control.¹⁷ Positive SPT was evaluated as the presence of atopy.

Pediatric asthma control test and asthma control test

Participants younger than 12 years old were given PACT. Validated versions of the tests in the Turkish language were applied.^18,19 PACT is composed of 2 sections: 4 questions for the patient on perception of asthma control, limitation of activities, coughing and awakening at night (4 response options), and 3 questions for the carers specific to daytime complaints, daytime wheezing, and awakening at night (6 response options), all answers were given for the previous 4-week period.²⁰ The total sum of scores generated the final score. The maximum total score for PACT is 27, which reflects total control in asthma. ACT composed of 5 questions (limitation of activities, shortness of breath, awakening at night due to asthma symptoms, rescue therapy need, and rating asthma control over the last 4 weeks, with 5 response options) answered only by patients older than 12 years.²¹ The maximum total score for ACT is 25 demonstrating total control in asthma. Minimum total scores of ACT and PACT is 5 and 0, respectively, which reflects poorest asthma control. A score of ≤19 for both tests indicates uncontrolled asthma.^20,21

Pediatric Asthma Quality of Life Questionnaire

The PAQLQ is a self-administered test developed by Juniper et al. and translated into many languages.¹³ The validated Turkish language version of PAQLQ was used.²² It is a disease-specific quality of life questionnaire consisting of 23 items and 3 domains. These domains are symptoms (10 items), emotional functions (8 items), and activity limitations (5 items). Each of the items have 7-point response Likert type scales, ranging from 1 (severe impairment) to 7 (no impairment). The overall PAQLQ score is obtained by the arithmetic mean of all items, while the individual domain scores are the means of the items in those domains. The maximum score is 7 (the best quality-of-life status), while the minimum score is 1 (the worst quality-of-life status).

FeNO measurement

FeNO was measured online by using the single-breath technique with a portable analyzer (NIOX-MINO; Aerocrine, Stockholm, Sweden) according to guidelines and expressed as parts per billion.^12,23 Measurements were performed before lung function tests. After inhalation of nitric oxide-free air for total lung capacity, the measurement was performed while the patient exhaled at a constant flow rate of 50 mL/s. The American Thoracic Society guidelines recommend that FeNO measurements of asthmatic children should be evaluated with cutoff levels of <20 and >35 ppb. FeNO levels <20 ppb denote reduced eosinophilic airway inflammation and predict lower corticosteroid response, while >35 ppb designates inversely more eosinophilic inflammation and, as a consequence, better response to corticosteroid therapy.¹²

Lung function tests

All patients underwent lung function tests by using Spirolab II system (Medical International Research, Rome, Italy). Tests were performed by a trained operator with equipment calibrated daily. Measured parameters were Prebronchodilator FEV₁, Forced vital capacity (FVC), FEV₁ to FVC ratio (FEV₁/FVC), and forced expiratory flow between 25% and 75% of vital capacity (FEF_25%–75%). Results were reported as percentage of the predicted value according to references.²⁴

Statistical analysis

Results were expressed as median and interquartile range (IQR) or proportions as required. While the data set was not normally distributed, comparisons of continuous variables (PACT, ACT, PAQLQ, FeNO, FEV₁, FVC, FEV₁/FVC, and FEF_25%–75%) were performed using the Mann–Whitney U test. Percentages of these variables were compared using the chi-square test. Correlations between tests were investigated by the Spearman correlation test. Diagnostic value of PACT, ACT, and PAQLQ to detect “well-controlled” asthma as defined by GINA was determined by calculation of area under the curve (AUC) of receiver operating characteristic (ROC) curves. Comparability of performed tests with GINA criteria were assessed by the Kappa test (κ). Statistical significance was defined as P < 0.05. Analyses were performed using SPSS statistical software, version 20.0 (SPSS, Inc., Chicago, IL).

Results

A total of 240 children with asthma were enrolled in the study. Twelve patients were excluded due to the following reasons: 3 patients did not attend the main visit, 5 patients suffered acute respiratory tract infections likely to affect the FeNO measurement, and 4 patients were not adherent to their therapy. Three of 5 children with acute respiratory tract infections had also acute asthma attack and they were excluded from the study because of being unappropriate for FeNO test. A total of 228 patients [median age (IQR) 10.7 (8.4–12.9) years] completed the study, of whom 132 (57.9%) were boys (Table 1). Study participants performed 145 PACT, 83 ACT, and 221 PAQLQ tests (PAQLQ tests for 7 patients were excluded due to incomplete tests). Asthma follow-up duration was similar for older (≥12 years) and younger patients (<12 years) (P = 0.44). All patients participating in the study underwent FeNO and lung function tests.

Table 1.

Characteristics of the Study Population

	Study population (n = 228)
Male, % (n)	57.9 (132)
Age (year), median (IQR)	10.7 (8.4–12.9)
Duration of follow up (year), median (IQR)	2.4 (1.1–4.9)
Tobacco exposure, % (n)	57.9 (132)
Aeroallergen sensitization	48.2 (110)
Asthma therapy, % (n)
Only as needed SABA	25.0 (57)
ICS use	68.4 (156)
ICS monotherapy	52.2 (119)
LTRA monotherapy	6.6 (15)
ICS+LTRA	8.3 (19)
ICS+LABA	7.5 (17)
ICS+LABA+LTRA	0.4 (1)
Asthma severity, % (n)
Intermittent	25.9 (59)
Mild persistent	51.3 (117)
Moderate persistent	15.4 (35)
Severe persistent	7.5 (17)
Allergic rhinitis, % (n)	41.9 (95)
FeNO (ppb)	19 (15–27)
PACT	22 (18–25)
ACT	22 (18–24)
PAQLQ, median (IQR)	6.0 (4.9–6.6)
Symptoms	5.2 (4.2–6.4)
Emotional functions	6.0 (5.1–6.6)
Activity limitation	6.2 (5.2–6.9)
Spirometry, median (IQR)
FEV1 (% predicted)	93 (87–103)
FEV1/FVC (ratio)	98 (93–105)
FEF25–75 (% predicted)	90 (79–112)

ACT/PACT, asthma control test/pediatric asthma control test; FEF_25–75, forced expiratory flow at 25% to 75% of forced vital capacity; FeNO, fractional exhaled nitric oxide; FEV₁, forced expiratory volume in 1 s; FEV₁/FVC, the ratio of “forced expiratory volume in 1 s” to “forced vital capacity”; ICS, inhaled corticosteroid; IQR, interquartile range; LABA, long-acting beta-2 agonist; LTRA, leukotriene receptor antagonist; PAQLQ, Pediatric Asthma Quality of Life Questionnaire; SABA, short acting beta-2 agonist.

Of the study population, 171 (75.0%) were using controller therapy. Of those on controller therapy, 119 (52.2%), 15 (6.6%),19 (8.3%), and 17 (7.5%) were using inhaled corticosteroid (ICS) monotherapy, leukotriene receptor antagonist (LTRA) monotherapy, ICS+LTRA, and ICS+long acting beta-2 agonist (LABA) combined therapy, respectively. Only 1 patient was undergoing ICS+LTRA+LABA combined therapy (Table 1). No differences were found between groups for the types of medications used (Table 2).

Table 2.

Comparison of Characteristics of the Patients with Well-Controlled and Not Well-Controlled Asthma

Characteristics	Well-controlled asthma (n = 192)	Not well-controlled asthma (n = 36)	P
Gender
Male, %*(n)	61.4 (118)	38.9 (14)	0.012
Female, %*(n)	38.5 (74)	61.1 (22)
Age (year), median (IQR)	10.5 (8.3–12.5)	12.2 (9.6–14.4)	0.005
Duration of follow-up (year), median (IQR)	2.1 (1.0–5.1)	3.3 (1.5–6.4)	0.180
Tobacco exposure, %*(n)	58.8 (113)	52.7 (19)	0.498
Aeroallergen sensitization	51.6 (99)	30.6 (11)	0.021
Asthma therapy, %*(n)
Only as needed SABA	26.6 (51)	13.9 (5)	0.059
ICS use	66.1 (127)	72.2 (26)
ICS monotherapy	51.6 (99)	55.6 (20)
LTRA monotherapy	7.3 (14)	11.1 (4)
ICS+LTRA	8.3 (16)	2.8 (1)
ICS+LABA	6.3 (12)	13.9 (5)
ICS+LABA+LTRA	0.0 (0)	2.8 (1)
Asthmas severity, %*(n)
Intermittent	27.6 (53)	16.7 (6)	0.167
Mild persistent	49.5 (95)	61.1 (22)
Moderate persistent	17.2 (33)	5.6 (2)
Severe persistent	57.3 (11)	16.7 (6)
Allergic rhinitis, %*(n)	42.2 (81)	38.9 (14)	0.695
FeNO (ppb)	19 (14–28)	19 (16–26)	0.654
PACT	23 (19–26)	18 (13.5–21)	<0.001
ACT	22.5 (19.25–24.75)	15 (14–19)	<0.001
PAQLQ, median (IQR)	6.1 (5.2–6.6)	4.8 (3.9–5.6)	<0.001
Symptoms	5.4 (4.6–6.6)	3.8 (2.8–5.2)	<0.001
Emotional functions	6.2 (5.4–6.7)	4.6 (3.8–5.4)	<0.001
Activity limitation	6.4 (5.2–7.0)	5.5 (4.5–6.2)	0.001
Spirometry, median (IQR)
FEV1 (% predicted)	93 (87–103)	94 (83–103)	0.757
FEV1/FVC (ratio)	0.99 (0.93–1.05)	0.97 (0.94–1.05)	0.724
FEF25–75 (% predicted)	92 (79–112)	88 (79–106)	0.485

Statistically significant results were written in bold.

^* Percent of columns.

ACT/pACT, asthma control test/pediatric asthma control test; FEF_25–75, forced expiratory flow at 25% to 75% of forced vital capacity; FeNO, fractional exhaled nitric oxide; FEV₁, forced expiratory volume in 1 s; FEV₁/FVC, the ratio of “forced expiratory volume in 1 s” to “forced vital capacity”; ICS, inhaled corticosteroid; IQR, interquartile range; LABA, long-acting beta-2 agonist; LTRA, leukotriene receptor antagonist; PAQLQ, Pediatric Asthma Quality of Life Questionnaire; SABA, short acting beta-2 agonist.

Moderate and severe asthmatics were 52 (22.8%) of all subjects and 37 (16.2%) patients were on combined therapy with at least 2 controller therapy. The remaining 15 (6.6%) patients were on controller therapy with medium dose ICS according to age recommended by GINA. The median (IQR) of budesonide equivalent ICS for the study population was 400 (0–400) mcg/day. The not well-controlled asthma patients were on significantly higher ICS daily dosage compared to the controlled asthmatics [budesonide equivalent daily ICS dosage median (IQR): 400 (100–400) vs. 400 (0–400) mcg/day, P = 0.024].

Characteristics of the total study population are presented in Table 1. Total PAQLQ and symptoms domain were significantly higher for participants <12 years old (n = 145) than those ≥12 years old (n = 83) [PAQLQ total: median (IQR) for <12 years 6.1 (5.2–6.6) vs ≥12 years 5.6 (4.4–6.5), P = 0.018; symptoms domain: median (IQR) for <12 years 5.4 (4.5–6.6) vs ≥12 years 5.0 (3.7–6.1), P = 0.022]. PAQLQ activities and emotions domains were similar for both age groups (P = 0.092 and 0.242, respectively). FeNO was significantly higher for the older age group [median (IQR) for <12 years 18 (14–27) vs ≥12 years 21 (16–32), P = 0.023]. There were no differences for medians of lung function parameters between age groups (P > 0.05).

In line with the GINA criteria, 84.2% of the study population (n = 192) had “well-controlled” asthma, with 9.6% and 6.1% of patients identified as having “partially controlled” and “uncontrolled” asthma, respectively. Following combination of the 2 latter groups, it was determined that 15.7% of patients suffered from “not well-controlled” asthma. Older patients (≥12 years) more frequently displayed “not well-controlled” asthma than younger counterparts (<12 years) [22.9% (n = 19) vs. 11.7% (n = 17), respectively, P = 0.026]. Regarding gender, older female patients more frequently exhibited “not well-controlled” asthma defined by GINA criteria, while male patients in the younger age group displayed higher rates of “well-controlled” asthma (P = 0.012 and 0.005, respectively) (Table 2).

When the study population was analyzed for asthma phenotypes in childhood,²⁵ 33 (14.5%) of the patients had early-onset asthma (before 3 years old). There were no difference between the early and late-onset asthmatics for asthma control status (P = 0.880). FeNO and pulmoner function test parameters were also similar between the early and late onset asthmatics (P > 0.05). Atopic patients had more frequently controlled asthma than nonatopic asthmatic children (P = 0.021) (Table 2). Out of 110 atopic asthmatic children, 34 of them had multiple sensitizations. Nevertheless, there were no significant differences for asthma control status, asthma severity, or frequency of early-onset asthma between monosensitized and multisensitized patients (P > 0.05).

Comparison of the self-administered control tools revealed the only significant correlation, as determined by the Spearman correlation efficient, to be between PAQLQ and PACT or between PAQLQ and ACT (r = 0.658 and r = 0.789, P < 0.001, respectively). FeNO and lung function parameters showed no intercorrelation or correlation with other tests (P > 0.05) (data not shown).

Comparing PACT to GINA criteria for “well-controlled asthma” resulted in an AUC of the ROC curve of 0.79 (confidence interval [95% CI] 0.70–0.88, P < 0.001). The cutoff level of 21.5 had the highest sensitivity and specificity (0.609 and 0.882, respectively). As a result, the cutoff level of 22 (PACT ≥22 for “well-controlled asthma”) provided the highest comparability of PACT with GINA [correctly classified 93 (64.1%), κ = 0.221, P < 0.001]. Likewise, the ROC curve analyses of ACT resulted in an AUC of 0.86 (95% CI 0.76–0.96, P < 0.001). The cutoff level of 20.5 had the highest sensitivity and specificity (0.719 and 0.895, respectively). Ultimately, the cutoff level of 21 (ACT ≥21 for “well-controlled asthma”) provided the highest comparability of ACT with GINA [correctly classified 63 (75.9%), κ = 0.473, P < 0.001]. In the same way, the ROC curve analyses of PAQLQ revealed an AUC of 0.76 (95% CI 0.67–0.84, P < 0.001). The cutoff level of 5.85 had the highest sensitivity and specificity (0.591 and 0.828, respectively). Finally, the cutoff level of 5.9 (PAQLQ ≥5.9 “well-controlled asthma”) provided the highest comparability of PAQLQ with GINA [correctly classified 139 (62.9%), κ = 0.150, P < 0.001] (Table 3). Performance levels of different cutoff values of PACT and ACT are shown in Tables 4 and 5. We also assessed the predictive value of FeNO to foresee “not well-controlled” asthma, which was poor (AUC = 0.45, P = 0.75) (data not shown). PAQLQ had higher agreement with GINA for older patients (≥12 years old) than the younger age group (<12 years old) (κ = 0.326, P < 0.001 and κ = 0.151, P = 0.014, respectively). PAQLQ exhibited stronger correlation with ACT (r = 0.789, P < 0.001) than PACT (r = 0.658, P < 0.001).

Table 3.

Comparability of the Tests with Global Initiative for Asthma Criteria

	Cutofflevel	Comparability with GINA		κa	P
		Not well-controlled asthma, n (%)*	Well-controlled asthma, n (%)*
FeNO	>35 ppb	2 (6.1)	162 (83.1)	−0.106	0.109
FeNO	>20 ppb	15 (14.4)	96 (84.2)	−0.014	0.779
PACT	<22	15 (23.1)	78 (97.5)	0.221	<0.001
ACT	<21	17 (48.6)	45 (95.8)	0.473	<0.001
PAQLQ	<5.9	29 (27.9)	110 (94.8)	0.150	<0.001
FEV1	<80%	6 (27.3)	160 (85.6)	0.105	0.118
FEV1/FVC	<0.80	0 (0.0)	173 (84.0)	−0.027	0.450
FEF25–75	<80%	9 (15.5)	127 (84.1)	−0.004	0.947

Statistically significant results were written in bold.

^* Percents of row.

^a kappa test.

ACT/pACT, asthma control test/pediatric asthma control test; FEF_25–75, forced expiratory flow at 25% to 75% of forced vital capacity; FeNO, fractional exhaled nitric oxide; FEV₁, forced expiratory volume in 1 s; FEV₁/FVC, the ratio of “forced expiratory volume in 1 s” to “forced vital capacity”; GINA, Global Initiative for Asthma; PAQLQ, Pediatric Asthma Quality of Life Questionnaire.

Table 4.

Pediatric Asthma Control Test with Various Cutoff Points Compared with Global Initiative for Asthma Criteria for Assessing “Not Well-Controlled” Asthma in 145 Patients

Cut-off points	Sensitivity %	Specificity %	PPV %	NPV %	Correctly classified %	Kappa	P
≤17	84.4	47.1	28.6	92.3	80.0	0.245	0.002
≤18	77.3	58.8	25.6	93.4	75.2	0.232	0.002
≤19	73.4	64.7	24.4	94.0	72.4	0.223	0.001
≤20	69.5	64.7	22	93.7	69.0	0.186	0.005
≤21	60.9	88.2	23.1	97.5	64.1	0.221	<0.001
≤22	52.3	94.2	20.8	98.5	57.2	0.184	<0.001
≤23	44.5	94.1	18.4	98.3	50.3	0.139	0.002
≤24	35.2	100	17.0	100	42.8	0.113	0.003

The cutoff point with significantly highest statistical values was written in bold.

NPV, negative predictive value; PPV, positive predictive value.

Table 5.

Asthma Control Test with Various Cutoff Points Compared with Global Initiative for Asthma Criteria for Assessing “Not Well-Controlled” Asthma in 83 Patients

Cutoff points	Sensitivity %	Specificity %	PPV %	NPV %	Correctly classified %	Kappa	P
≤17	89.1	68.4	65.0	90.5	84.3	0.564	<0.001
≤18	81.2	73.7	53.8	91.2	79.5	0.486	<0.001
≤19	75.0	78.9	48.4	92.3	75.9	0.441	<0.001
≤20	71.9	89.5	48.6	95.8	75.9	0.473	<0.001
≤21	65.6	89.4	43.6	95.5	71.1	0.402	<0.001
≤22	50.0	94.7	36.0	97.0	60.2	0.284	<0.001
≤23	39.1	94.7	31.6	96.2	51.8	0.198	0.005
≤24	25	94.7	27.3	94.1	41.0	0.106	0.061

The cutoff point with significantly highest statistical values was written in bold.

NPV, negative predictive value; PPV, positive predictive value.

Discussion

In this study, comparability of the tools PACT, ACT, FeNO, PALQLQ, and lung function parameters in accordance with GINA criteria were evaluated in asthmatic children. According to GINA criteria, 84.2% of the patients assessed suffered from “well-controlled” asthma.

FeNO and lung function parameters were unsuccessful at revealing control status according to GINA criteria (Tables 2 and 3). PACT/ACT and PAQLQ were demonstrated to be able to determine “well-controlled asthma” to a certain extent consistent with GINA criteria (Table 3). Highest sum of sensitivity and specificity were obtained with the cutoff levels of 22 for PACT, 21 for ACT, and 5.9 for PAQLQ (Table 3). Nevertheless, only fair agreement occurred with kappa coefficient values of 0.221 and 0.150 for PACT and PAQLQ, respectively (P < 0.001). Highest agreement was obtained between ACT and GINA (κ = 0.473, P < 0.001). Percentages of patients correctly classified with PACT, ACT, and PALQLQ according to GINA criteria were 93 (64.1%), 63 (75.9%), and 139 (62.9%), respectively. Ultimately, ACT produced the best result for assessing asthma control in children aged ≥12 years old. However, PAQLQ generated better comparability for asthma control in children ≥12 years old than in those <12 years old.

In our study, girls in the older age group were more frequently seen to be affected by “not well-controlled” asthma. This may be due to the adverse impact of sex hormones on the control status of asthma, which is a particular problem for female adult patients.²⁶ This may be in support of the hypothesis that gender disparity hazardous for females in asthma control starts at puberty approximately at the age of 10 years.²⁶

Although there is no approved or validated gold standard for determining asthma control, GINA criteria are accepted worldwide as a standard used in daily clinical practice.^5,27 PACT and ACT are validated tools which are mostly used in clinical practice to assess asthma control in children. It should be noted that studies comparing the consistence of tools with the GINA criteria for assessment of asthma control produced variable results with different agreement rates and cut-off levels.^6–10

Koolen et al., compared PACT and ACT with GINA criteria in assessing asthma control of children, in a study with similar design and patient population to the present study.¹⁰ They determined an AUC of ROC curve analyses for PACT as 0.89 and ACT as 0.92, which were higher than our study (0.79 and 0.86, respectively). Koolen et al. established the cutoff value with the highest sum of sensitivity and specificity for ACT as ≤20 for uncontrolled asthma, which corresponds to our study. The cutoff values for PACT differ slightly between the 2 studies: ≤22 in their study and ≤21 in our study. The sensitivity and specificity of both PACT and ACT (sensitivity 82% and 76%; specificity 85% and 96%, respectively) with the calculated cutoff values were higher than our study (sensitivity 60.9% and 71.9%; specificity 88.2% and 89.5%, respectively). This may be because Koolen et al. (n = 145) evaluated more patients with uncontrolled (13.1%) and partly controlled asthma (33.8%) than in our study (6.1% and 9.6%, respectively).¹⁰ Another reason may be that patients with uncontrolled and partly controlled asthma were evaluated collectively in the “not well-controlled” asthma group in our study.

In another study of similar design, Voorend-van Bergen et al.⁸ (n = 228) found the same best cutoff value for PACT (≥22) for “well-controlled” asthma with respect to GINA criteria. They proposed to determine “well-controlled” asthma which is the primary goal of clinical asthma management according to GINA criteria. The AUC for PACT (0.81) determined by these authors was also similar to our results (0.79). On the contrary, their best cutoff value for ACT (≥23, AUC = 0.91) was higher than ours (≥21, AUC = 0.86).

In a recent study, Deschildre et al.⁷ compared PACT performance of assessing asthma control with respect to GINA criteria. As for our study, they grouped asthma levels of patients as “well controlled” and “not controlled” (combining partially controlled and uncontrolled, as was done in this study). Using ROC curve analyses, they found the same cutoff value for PACT (≤21, for “not controlled” asthma) with a higher sensitivity (76%), but lower specificity (81.5%) than those found in our study (60.9% and 88.2%, respectively). However, the most important distinction is that the rate of “not controlled” asthma in their study (76.5%) was much higher than for our study (15.7%).⁷

Voorend-van Bergen et al,⁸ established no difference in FeNO between the control groups of asthma as was found in our study. Similar to earlier studies, we found no correlation between asthma control scores, GINA criteria and FeNO or FEV_1.^28–30 Symptoms, lung functions and airway inflammation represent different domains of asthma phenotype and show poor agreement.⁸ Many children with uncontrolled asthma have normal lung functions between exacerbations.¹ A low FEV₁ percent predicted, particularly if it is <60%, identifies patients at risk of future asthma exacerbations independent of symptom levels.¹ If symptoms are few despite low FEV₁ (% predicted), limitation of lifestyle or poor perception of airflow limitation should be considered, which would be a marker of untreated airway inflammation.^31,32 In our study, none of the participants displayed FEV₁ <60%. Besides, there were no differences for FEV₁ (% predicted), FEV₁/FVC (ratio), and FEF_25–75 (% predicted) between control groups and no comparability with GINA criteria (Table 2).

ACT seems to be better than PACT for comparability with GINA in assessing control status in children. Higher correlation of PAQLQ and ACT and better comparability of PAQLQ and GINA were apparent in older children. This is probably because PACT is not only patient reported but also based on the parental report. It is common for parents to underestimate asthma symptoms.^33,34 With respect to control levels determined by the GINA criteria together with symptoms and medication diaries, PACT defined control status might be more disparate than that identified by ACT.

Besides, the correlation of PAQLQ and ACT was stronger than that of PAQLQ and PACT in our study. Voorend-van Bergen et al.⁸ showed similar results for this correlation. In addition, PAQLQ yielded higher agreement with GINA criteria in children ≥12 years old than those in <12 years old (κ = 0.326, P < 0.001 vs κ = 0.151, P = 0.014, respectively). Several studies have demonstrated the correlation between PAQLQ and ACT/PACT.^35–38 However, to the best of our knowledge, this is the first study to achieve a cutoff value for PAQLQ that has fair comparability with GINA criteria.

The most important limitation of our study was the reduced number of patients with “uncontrolled” and “partly controlled” asthma. This may be a consequence of the study design, in which before initiation of research, treatment adherence and inhaler technique were evaluated and adjusted if needed. Four weeks later, patients were called for the main visit, at which the adherence and technique of the patients had improved compared to the primary visit (data not shown). As a consequence, asthma control was better at the main visit.

Our study provides real-life data from daily clinical practice. This explains why the ratio of patients with “uncontrolled” asthma was lower than the above mentioned studies. Recently, real-world studies have become cause of choice because randomized controlled trials include selected patients populations that rarely represent the real-world situation.³⁹

Conclusion

ACT and PAQLQ seem to be the best instruments for clinicians to use in children to determine asthma control based on the GINA criteria, until further clarification about the use of instruments for assessing asthma control with large scale studies has been made.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.