Asthma is the most common chronic childhood disease in developed countries.1 In Spain, between 7.1% and 18.8% of children present the most typical characteristics of the disease.2,3 Most treatment-compliant patients are well-controlled, thanks to diagnostic and therapeutic advances, but good control is not achieved in up to 70% of childhood sufferers.4

Expert consensus papers agree that treatment education must form part of disease management, and emphasize it as an essential component for improving a patient's condition. Knowledge of the different aspects of the disease is a determinant factor in patient progress. For children and adolescents, this education must extend to family members, teachers and any adult responsible for the care or supervision of a young asthma patient.5–8

Teachers’ knowledge of asthma must be assessed in order to determine the need for educational activities, and to design their content and evaluate their efficacy. Children and adolescents spend long hours at school, under the supervision of their teachers. Thus, it is of interest to evaluate the knowledge of asthma among these professionals, to determine if training and/or informative sessions are required and what the content should be.9

Epidemiological studies in asthma knowledge can only be performed using a tool that is reliable and feasible from both an organizational and an economic point of view. Self-administered questionnaires, that are easy and inexpensive to complete, are useful, but they must be validated in a typical study population.

Before this study, only 3 papers evaluating teachers’ knowledge of asthma had been published. All 3 used self-administered questionnaires, but no data about validation techniques were provided.10–12 To date, no data on the validation of any questionnaire in Spanish on the knowledge of teachers in this area have been published.

The Newcastle Asthma Knowledge Questionnaire (NAKQ) for the evaluation of asthma knowledge was created and validated in English in 1990.13 It was subsequently used in the original language14–18 and in Spanish19 in different populations (primarily parents,13,15 educators,14 students14,18,19 and healthcare personnel17). In 2008, Praena Crespo et al.20 made a transcultural adaptation of the questionnaire to Spanish, and analyzed its validity and reliability in a population of parents of asthma patients. Hence, the NAKQ questionnaire has been widely used, and 2 studies have been published confirming its reliability and validity: the original in English13 and the Spanish adaptation,20 both performed in parents of asthmatics. However, reliability and validity are not intrinsic properties of questionnaires; their value depends more on their use in specific populations,21 so they must be verified in each study population.

The aim of our study was to determine the reliability of scores from the Spanish version of the NAKQ in a population of teachers, and its validity for quantifying knowledge of asthma among teachers in Spanish schools.

This was an observational, cross-sectional, descriptive study using a self-administered questionnaire on asthma knowledge completed by teachers in the region of A Coruña (Spain) between June and April 2009. This region comprises 9 municipal areas (Abegondo, Arteixo, Bergondo, Cambre, Carral, A Coruña, Culleredo, Oleiros and Sada) and covers an area of 470.7km2 with 384616 inhabitants.

Two-stage cluster sampling was performed. Participating schools were selected by proportional, stratified, randomized sampling, according to whether they were public or private and their educational level (public primary schools, public secondary schools or private centers). In each school, all teachers actively imparting any type of classes in the 2008–2009 school year, at any level of preschool, primary or compulsory secondary education were included. Teachers working exclusively in post-compulsory secondary education were not included.

In the 2008–2009 school year, 4089 teachers were actively employed in the area in 96 schools (46 public primary schools, 25 public secondary schools and 25 private schools) (data from the Education Department of the Regional Administration of Galicia). It was estimated that 873 teachers would need to be included (precision, ±4%; confidence, 95%), and a response rate of 60% was expected. On the basis of the mean number of teachers, 22 schools were included, stratified by public/private status and educational level (10 public primary schools, 6 public secondary schools and 6 private schools). A reserve list of randomly selected schools was generated in case any of the initially selected schools refused to participate.

The study was performed with the approval of the Clinical Research Ethics Committee of Galicia (approval number 2009/016) and the Education Department of the Regional Administration of Galicia, which sent a letter to the selected schools informing them of the study objectives. The board of each school was then contacted, the project was explained, and they were invited to participate.

Each teacher received a letter informing them of the study objectives and requesting their consent to participate. Pursuant to the Data Protection Act 15/1999, data confidentiality was guaranteed at all times.

The NAKQ20 was used to determine teachers’ knowledge of asthma. This questionnaire consists of 31 items: 25 are true/false questions and 6 have open answers. Correct responses score 1 and incorrect 0. Open answers were interpreted in the same way as the original version,13 as adapted by Praena Crespo et al.20 Total score is obtained by adding the score of the 31 items: the higher the score, the greater the knowledge of the disease.

Each teacher recorded their age, sex, academic training (diplomas/degrees), time of teaching experience and levels taught. All participants were asked if they or anyone in their immediate setting had asthma (spouse/partner, children, parents, siblings, other family members or close friends). They were also asked if they felt their knowledge of asthma was sufficient.

Of the 22 selected schools, 3 refused to participate, so these were replaced with schools from the reserve list. In the end, 24 schools (10 public primary education, 6 public secondary education and 8 private) were included, and 864 teachers who met the inclusion criteria participated.

The percentage of participation was 62.1%, and 537 participants returned the completed questionnaire (56.8% from public primary education, 62.1% from public secondary education and 67.2% from private schools). Characteristics of the participants are shown in Table 1.

Mean score on the NAKQ questionnaire was 15.7±5.3 (median, 17.0) (Fig. 1). The floor and ceiling effect percentages were 2.0% (score=0) and 0% (score=31). On average, 50.6% of participants answered the items correctly. The percentage of correct results for each item ranged from 89.6% to 1.10% (Online Appendix: Table 1).

Fig. 1.

Distribution of Newcastle Asthma Knowledge Questionnaire (NAKQ) scores among teachers. SD: standard deviation.

(0.06MB).

Internal consistency analysis gave a high Cronbach's alpha coefficient of 0.824 (95% CI: 0.802–0.845).23 As each item was excluded, in a stepwise fashion, the index ranged between 0.811 and 0.828.

Significant differences were observed in the concurrent validity of the NAKQ score, depending on the personal contact of teachers with asthma and their self-assessed knowledge of the disease (Table 2). Scores were lower for teachers with no contact with asthmatics (15.1±5.6) than for asthmatic teachers or those with a spouse/partner/children/parents/siblings with asthma (17.7±3.3) and in teachers with other family members/friends who were asthmatic (16.1±5.4; P<.001). Lower scores were found in teachers who considered their knowledge of asthma to be insufficient (16.1±5.0) compared to those who considered it to be sufficient (18.7±2.2; P=.006). Other variables associated with poorer knowledge were older age (Spearman's rho=−0.192; P<.001), longer teaching experience (Spearman's rho=−0.160; P<.001) and female sex (15.4±5.3 vs 17.0±4.8; P<.001) (Table 2).

Test–retest reliability was evaluated in 48 teachers with similar characteristics in terms of age, years of teaching experience, academic training and personal/family history of asthma, although there were more men in this group (61.7% vs 25.9%; P<.001). An intraclass correlation coefficient of 0.863 (95% CI: 0.768–0.920) was obtained for the overall questionnaire score. Using Bland–Altman methodology, the average difference between the scores in the 2 questionnaire submissions was 0.3±2.3 (95% CI concordance: −4.2; 4.8) (P=.330) (Fig. 2). For each item, the percentage of concordance between the 2 items ranged from 68.7% to 100% (kappa index: 0.33–1) (online Appendix: Table 2).

Fig. 2.

Newcastle Asthma Knowledge Questionnaire (NAKQ) test–retest reliability. Bland–Altman graph. SD: standard deviation; CI: confidence interval.

(0.11MB).

In studies published on asthma knowledge among Spanish teachers, Callén and Garmendia10 found that 95% of teachers were unaware of inhalation techniques or the effects of bronchodilators; Rodríguez Fernández-Oliva et al.12 reported that 64% did not know the first steps to take in the event of a crisis, and according to Cobos and Picado,11 91% of teachers admitted that their knowledge of the disease was poor. These and other findings suggest that there are large gaps in knowledge that could affect the course of the disease in patients. However, if an educational intervention program is to be designed, new data obtained from validated tools will be necessary.

The sample size of our study, in contrast to previous studies, was large enough to be accurate (n=537): the original study was performed in 138 parents,13 and the transcultural adaptation to Spanish was performed in 157.20

Score outcomes were similar to those of other studies in subjects with average knowledge, and no significant floor or ceiling effects were found, suggesting that the questionnaire is capable of detecting especially low or especially high levels of knowledge.

Score reliability was measured in 2 ways, according to conventional test theory.21 Firstly, internal consistency was evaluated with Cronbach's alpha. The statistical interpretation of this test is an estimate of the proportion of real variance contained in the total variance observed in the scores obtained.21 The value achieved, 0.824 (95% CI: 0.802–0.845), which did not change as each questionnaire item was excluded, is high and indicates that 82% of the total observed variance is real variance and the rest is only error variance. Compared to other studies, this value is similar to the 0.72 (95% CI: 0.65–0.78) obtained by Praena Crespo et al.20 The original study by Fitzclarence and Henry13 did not report this value, and the reliability study was limited to the test–retest evaluation. No significant differences were found between the Cronbach's alpha obtained in our study and by Praena Crespo et al.20 (single-tailed significance, P=.20). This Cronbach's alpha suggests high score reliability. In general, values higher than 0.6 are acceptable and scores higher than 0.8 are considered excellent. Values close to 1 would indicate redundancy of items.21

Secondly, a test–retest design with an interclass correlation coefficient for the total score between the 2 completions, interpreted similarly to the Cronbach's alpha, obtained an excellent internal consistency value of 0.863 (95% CI: 0.768–0.920). Praena Crespo et al.20 reported a similar parameter, Kendall's tau-b, with a value of 0.856, while in the original study, an equally high Kendall's tau of 0.94 was reported.13

The test–retest analysis showed adequate reproducibility, confirming that scores were consistent over repeated applications of the questionnaire.21 The percentage of concordance was high for all questions (from 68.7% to 100%). A better method for determining the degree of concordance is the kappa index, which takes into account (by subtraction) concordance due to chance. Kappa values range between −1 (complete lack of concordance), 0 (concordance due to chance) and 1 (full concordance). In our case, the kappa indexes for each question showed excellent consistency between repeated applications (kappa between 0.33 and 1), slightly higher than that reported by Praena Crespo et al.,20 who found a kappa≤0.2 for at least 8 items. The original validation study did not report concordance values.13

With respect to the validity of the scores obtained for quantifying asthma knowledge, the main advantage of our study design lies in its attempt to verify whether teachers with more contact with asthmatic patients or those who felt they had sufficient knowledge of asthma returned higher scores (concurrent validity). This was confirmed by the results, although the difference between groups with high knowledge and those with low knowledge was less than in the 2 previous reliability and validity studies, possibly because the high knowledge population (parents of asthmatic children and a high educational level) was a highly selected group. Another important aspect of validity, sensitivity to change, was evaluated using scores obtained before and after an educational intervention in a random sample of teachers in schools in San Sebastian, Spain. A significant increase was observed in the level of asthma knowledge (16.1±3.4 vs 22.3±4.1; P<.001), suggesting that the sensitivity to change of the Spanish version of the NAKQ is sufficient for use in this context.23

Ours is the third study to analyze the reliability of NAKQ scores and their validity for measuring asthma knowledge: in our case, we assessed the Spanish version of the questionnaire in teachers. In all 3 studies, the reliability and validity results are satisfactory, and this certainly provides evidence that the NAKQ in these populations is reliable and measures what it aims to measure: asthma knowledge in these populations. However, the potential publication bias and the possibility that studies with poor results were not published must be taken into account. Notwithstanding, in our population, our scores were both reliable (with little random error) and valid (really measure what we want to measure).

The limitations of this study are associated with sample selection and the characteristics of the study population. To control for selection bias, a randomized sample of public and private schools in the study area was determined, thus ensuring results that could be extrapolated to the original population, unlike other studies that used non-probabilistic sampling.20 The participation rate was similar to that of similar studies,14 and comparable between the various private and public schools and educational levels. No data are available on the sociodemographic characteristics of the teachers who did not complete the questionnaire. This prevented us determining whether there is any bias between the population that participated and the population that did not. Teachers who participated may have been more aware of asthma issues, and thus may have had greater knowledge. However, this possibility can be ruled out, as 63.3% of the participants stated that they had no contact with asthmatics, and only 6.3% thought their knowledge of the disease was adequate.

With regard to the population characteristics, the geographical limitations and linguistic peculiarities may mean that the results are not generally applicable. However, given the practically universal knowledge of Spanish among the subjects, it is unlikely that our results cannot be applied to teachers from other regions in Spain, no matter how different.

Another possible limitation derives from the collection of self-reported data that may affect the internal validity of the study. Both the diagnosis of asthma in teachers or in close family members and the assessment that they made of their knowledge is self-declared, thus exposing the study to bias in the evaluation of the concurrent validity of the questionnaire. Indeed, the reason for conducting this study was the lack of validated instruments for assessing asthma knowledge among teachers, making it difficult to evaluate the criterion validity in this context.

Our study confirms that this questionnaire is a valid and reliable tool for the scientific community to use in the evaluation of asthma knowledge among teachers in Spanish schools, and one that may be useful for planning and evaluating educational interventions aimed at improving this knowledge and for helping to optimize asthma control in Spanish schoolchildren.

To conclude, our results provide quality data to show stakeholders that scores obtained with the Spanish version of the NAKQ are reliable and valid for measuring the level of asthma knowledge among teachers in Spanish schools.

All authors are responsible for the study and participated in its concept and design, analysis and interpretation of results, writing and correction of the manuscript and are in agreement with its content.

The authors declare that they have no conflict of interests.