 |
voltar ao início da página |
ADAPTATION AND VALIDITY OF THE ATS-DLD-78-C QUESTIONNAIRE FOR ASTHMA DIAGNOSIS IN CHILDREN UNDER 13 YEARS OF AGE
AUTHORS
ALZIRA ROSA ESTEVES
MD, Division of Ambulatorial and Social Pediatrics, Department of Pediatrics,
Universidade Federal de São Paulo, Escola Paulista de Medicina, São Paulo, Brasil.
DIRCEU SOLÉ
MD, Associate Professor - Division of Allergy, Clinical Immunology and Rheumathology -
Dept. of Pediatrics - Universidade Federal de São Paulo, Escola Paulista de Medicina -
São Paulo, Brasil.
MARCOS BOSI FERRAZ
MD,PHD,Professor, Division of Rheumatology, Dept. of Medicine, Universidade Federal de
São Paulo, Escola Paulista de Medicina - São Paulo, Brasil.
Author responsible for correspondence:
Alzira Rosa Esteves
Rua Luis Goes, 206/121
São Paulo, Capital
CEP 04043000
Brasil.
ABSTRACT; INTRODUCTION; MATERIALS AND METHODS; RESULTS; DISCUSSION; REFERENCES
The objective of this work was to translate into Portuguese, adapt and validate the ATS-DLD-78-C questionnaire to be used as a diagnostic tool in children with asthma.After being translated into Portuguese, the questionnaire was analysed by ten pediatricians and then by a panel consisted of four other pediatricians who selected nine questions, atributing a score to each one and nominating it as M-ATS-DLD-78-C ( M- ATS).
The questionnaire was validated through its application to the parents or guardians of children aged from four months to 13 years old ( N=79).These children were divided in two groups: one from the Pediatrics Ambulatory of Hospital São Paulo (N=34) and the other from the Health Basic Unit of Embú City (N=45). Half of each group was asthmatic and the other half non asthmatic children.
The same questionnaire was applied 15 days later to the Hospital São Paulo`s group.The reliability of the questionnaire was obtained through Pearson's correlation coefficient (R=0.891) and through Kappa's test (K=0.862), by a comparison of the first and second questionnaire application in the Hospital São Paulo`s group.
With the sensibility and specificity of the overall score of both groups, a ROC curve was constructed for each group thus determining a global cut-off score of seven points, making it possible to differentiate asthmatic ( global score ³7 points) from non- asthmatic children ( global score < 7points).
M-ATS questionnaire seems to be not only an important instrument to differentiate asthmatic from non- asthmatic children, but also presents a high reliability to identify asthma in children younger than 13 years of age.
Keywords: Asthma, questionnary, children.
Epidemiological studies have provided important data about asthma occurrence, etiology and natural history. Epidemiological inquiries have made it possible to study indicators such as prevalence, incidence, morbidity, mortality and risk factors. The best method to detect changes in the occurrence of asthma in a population is to estimate its prevalence in cross sectional studies, at different times and using the same method (1).
Data revision on asthma prevalence in western countries have shown increased advance since 1950 (1,2,3).In the United States, this upward trend in asthma prevalence has been noticed since the 1950(4) . In England, BURNEY, CHINN, RONNA (1990))(5) reported this trend from 1973 to 1986 (6.9% increase among boys and 12.8% among girls); and BURR et al (1989)(6) , after studying 12 year-old children in Wales in 1973 and in 1988, observed an increase from 6% to 12%.
Some asthma incidence data was observed with asthma incidence (7-9) , and its severity measured by hospital admissions (10-14) and mortality rates (15,16)
These data show the importance of asthma diagnosis mainly in children, and indicate that further research to formulate mechanisms and strategies to identify the disease and the risk factors associated with the illness are necessary in order to provide a more adequate control.
Epidemiological studies of asthma are dificult . In many cases, asthma diagnosis is not easy to establish.Symptoms as wheezing, shortning of breath, cough and expectoration themselves do not define the diagnosis, so a report of recurrent crisis is important. Another relevant factor is that physical examination except during an acute exacerbation may not show any alteration. One more factor to be concerned is that the extension of the airway obstruction and its variability is relevant for the asthma diagnosis (17)
The evaluation of non-specific bronchial hyperresponsiveness (BH) to histamine and methacholine seems to be an objective measure for the asthma diagnosis. However, it presents some restrictions. As well as asthma may occur with absence of BH; BH may be present in normal non-asthmatic individuals and these tests are dificult to administer to children aged under 6 years of age (18).
The above mentioned factors make the asthma related studies difficult to be elaborated, worsening the possibility of having a standard comparison among them. Several instruments have been developed and the written questionnaires are the most used ones in the epidemiological studies. The Internatinal Study of Asthma and Alergies in Childhood (ISAAC) is an example (19).
Before effectively administered to a population, any WQ with a differenciate function must be validated, that is, analyzed with reference to its capacity to separate the presence or absence of the investigated disease . The WQ reproduction capacity should also be evaluated, i.e., its capacity to identify cases and controls, in similar ways in successive interviews.
One of the most used WQ on asthma studies is the ATS-DLD-78, elaborated by the American Thoracic Society (ATS). It is composed of 46 questions about respiratory diseases, and it has already been released for administration in children over 13 years of age (20).
Afterwards, ATS has adapted this WQ to parents and guardians of children under 13 years of age (ATS-DLD-78-C), but this has not been validated so far.
Due to the importance of the knowledgment about asthma prevalence in children in our area, mainly to apply further preventive and therapeutic strategies we have adapted, reduced and validated the ATS-DLS-78-C questionnaire after its translation to Portuguese. A discriminative instrument for asthma epidemiological survey in children under 13 years of age is necessary.
The ATS-DLD-78-C WQ, originally written in English, was translated into Portuguese and some words were adapted to Portuguese language. To turn this QW more objective for asthma diagnosis and easier to be applied, it was reduced. A group of ten pedriatricians were asked to choose, at first, 20 questions considered the most important for asthma diagnosis in children. Afterwards, they were requested to indicate 10 questions that they considered the most relevant for asthma diagnosis.
The questions chosen by four or more pediatricians (13 altogether) were then submitted to a "group of analysis" composed of three general pediatricians who were asked to chose and score questions according to their importance for the asthma diagnosis.
Thereby nine questions were selected, and the final WQ named ATS-DLD-78-C-M (M-ATS), with a 17 total score .
Before the administration of the WQ, a back translation to English was done and no significant differences were observed between the original and the final one.
To validate the WQ, they were admnistered to child's parents by two trained layperson interviewers. Two groups of children aged from zero to 13 years old: cohort BUS (Basic Unit Health) and cohort SP (Hospital São Paulo) participated in this study.
The selection of the children was done by the author, on the days which they attended at the health centers. Each record was analyzed based on the presence or absence of asthma diagnosis and therapies. Notes were also taken at the last clinical appointments.
The BUS cohort consisted of 34 children, 15 of them were asthmatics and attended in two different health centers in Embú (28 Km from São Paulo city). The families of these children have income per capita lower than US$50.00 a month (half of the local minimum wage), and the city has a child mortality rate of 35 / 1,000 live newborns, being respiratory diseases the second mortality cause.
The asthmatic children from BUS cohort had attended the program "Attention to Children with Wheezing" developed in those health centers. The non-asthmatic children were selected at random, from the routine appointments or emergency (Table 1).The mean age of the non-asthmatic children in this group was 28 months, 42% girls and 58% boys. The mean age among the asthmatics was 71 months, 33.3% girls and 66.7% boys.
The HSP cohort was composed of 45 out-patient children from the Hospital São Paulo pediatric clinic, 22 without and 23 with asthma diagnosis. Hospital São Paulo is a school hospital linked to the Universidade Federal de São Paulo/Escola Paulista de Medicina, which provides tertiary health care to families with low and medium income, from different regions of São Paulo and other municipalities .The asthmatic children from the HSP cohort, attended the Allergy and Immunology Clinic of the Pediatric Department and the non-asthmatic attended clinics of other pediatric specialties (surgery, cardiology, gastroenterology, endocrinology, adolescence and hematology) (Table 2) .
The mean age of the non-asthmatic children in this group was 72 months, 31.8% girls and 68.2% boys. Among the children with asthma, the mean age was 95 months, 39% girls and 61% boys.
The interviews were carried out at the out patient facilities. After each interview, the author attributed the corresponding score to each questioner and calculated the global score for each child (Tables 1 and 2).
Eleven asthmatic children and 5 non-asthmatic children from the cohort HSP were submitted again to the WQ by the same interviewer in an interval of 15 days. The validation of the M-ATS WQ was done by the calculation of sensitivity (proportion of diseased with positive test) and specificity (proportion of non-diseased with negative test) for each total score in both groups of children.
To establish the cut-off points of the global scores in order to separate asthmatics from non-asthmatics, a ROC (Receiving Operating Characteristic) curve was drawn for both groups of study. This curve shows the balance between sensitivity and specificity, built by the representation of sensitivity and by (1-specificity)(21). The Likelihood Ratio was also assessed.
The questionnaire intra-observer reliability in the diagnosis or not of asthma was calculated according to Kappa Test. The Kappa test compared the number of patients that obtained global score higher or equal (asthmatic) and lower (non-asthmatic) than the cut-off point in both consecutive interviews (Table 4).
The Pearson's correlation coefficient was calculated by comparing the global scores obtained from the HSP first interview with the ones obtained at the second interview (to HSP cohort). In both tests, the rejection level of 5% has been fixed for the null hypothesis (Table 5).
Table 3 shows sensitivity and specificity values for each global score obtained after administration of the M-ATS QW in BUS and HSP cohorts.
According to the ROC curves for each cohort (Figures 1 and 2) a cut-off point for each one was coincidentally a global score of 7 points . The sensitivity and specificity values for the cut-off points in both cohorts were high, 0.87 and 0.95 respectively in HSP cohort and 0.93 and 1.00 in BUS cohort.
The Likelyhood ratio (LR) for a positive test (score > 7 points) and for a negative test (score < 7 points), in both groups, is shown in Table 6. It indicates that in the HSP cohort the occurrence of global score > 7 points (positive test) is 21.15 times more frequent among asthmatic children than among non-asthmatic children, and that the chance of occurrence of global score < 7 points (negative test) is 0.14 times higher in asthmatic patients than in non-asthmatic patients . In the BUS cohort, LR was equal to infinite for global score > 7 points and 0.07 for a global score < 7 points .
To evaluate the reliability in classifying on asthma patients, the Kappa test was carried out indicating excellent agreement between the first and second interviews (Table 4). Pearson's correlation coefficient (PCC) calculated between the scores obtained in the first and second interviews was 0.891 (p<0.05) (Table 5).
The increase of asthma incidence, prevalence, hospitalization and mortality in childhood in the last decades, as reported by epidemiological studies, have concerned public health throughout the world.
Prevalence data are particularly important, as they allow the elaboration of measures and health programs to extend the best treatment to asthmatic patients, reducing hospitalization rates and sequela, and establishing prophylactic actions for children under higher risk of developing the disease.
To carry out epidemiological studies, tools are used with discriminative goals. The reliability of these tools must be evaluated before use and this is possible by analyzing its reapplicability and discriminative validation.The most commonly used tools have been questionnaires, which can be administered to the population by an interviewer or completed by the interviewed.
The objective of our study was to reduce and validate the ATS-DLD-78-C questionnaire. The reduction from 46 to 9 questions aimed to obtain a more specific questionnaire for asthma diagnosis and speed up its administration .
Evaluation of the initial translation showed no important changes in relation to the original WQ. Besides the reduction of the number of questions, an analysis group attributed scores to each question, making it possible to compare the global score with the respective clinical diagnosis obtained from each child's record. Sensitivity and specificity for each value was thereby calculated for each global score.
With the sensitivity and specificity data a ROC curve was drawn for each cohort indicating the global score which discriminates asthmatics and non-asthmatics. In both cohorts, BUS and HSP, the cut-off points were 7 points coincidently. The sensibility and specificity values for the cut-off point in both cohorts were high.
The Likelihood ratio of rate of probability with positive test in cohort HSP was 21.15. This happened probably because the HSP control cohort was formed by children with chronic and more severe illness in relation to children from the BUS control cohort, which is a primary health care service. This suggests that the parents or guardians of children with chronic or more severe illness give more accurate answers during the interview.
Regarding reapplicability of the M-ATS, the Kappa test analyzed the agreement in the classification of asthmatic or non-asthmatic children in two successive interviews, without any chance influence . We have obtained excellent agreement between the global scores from the first and second interviews, with Pearson's correlation coefficient of 0.89. Such facts indicate that the questionnaire shows good reapplicability.
Our data reveal that the M-ATS questionnaire is a good tool for identification and epidemiological research of asthma in childhood.
In conclusion, the reduction of the ATS-DLD-78-C questionnaire from 46 to 9 questions resulted in a faster and more objective administration tool as far as asthma diagnosis in children younger than 13 years of age is concerned. Its translation to Portuguese has not altered the original meaning of the English version.
TABLE 1: Children from the UBS cohort, respective global score and clinic diagnosis
TABLE 2: Children from the HSP cohort, with respective global score and clinic diagnosis
TABLE 5: Global score in both HSP interviews.
TABLE 6: Likelyhood Ratio obtained with positive and negative tests in BUS and HSP cohorts.
(1) Gregg I. Epidemiological research in asthma: the need for a broad perspective. Clinical Allergy 1986;16:17-23.
(2) Mitchell EA. Increasing Prevalence of Asthma in Children. The NZ Need J 1983; 22:463-4.
(3) Chavarria, AG. Consideraciones sobre epidemiologia del asma in Mexico. Alergia, 1980;633:97-101.
(4) Taylor WR. & Newacheck PM. - Impact of Childhood Asthma on Health. Pediatrics 1992; 90:657-662.
(5) Burney PGJ.; Chinn S.; Rona RJ. - Has the prevalence of asthma increased in children? Evidence from the national study of health and growth, 1973-86. Br Med J 1990; 300:1306-10.
(6) Burr ML.; Butland BK.; King S; Vaugham W.- Changes in Asthma Prevalence: two surveys 15 years apart. Archives of Disease in Childhood 1989; 64:1452-1456.
(7) Leeder SR., Corkhill RT., Irwing LM, Holland WW. - Influence of family factors on asthma and wheezing during the first five years of life. Br. J Prev Soc Med 1976; 30:213-8.
(8) Peat JK., Woolcock A J, Leeder SR., Blackburn CR. - Asthma and bronchitis in Sydney Schoolchildren. Prevalence during a six-year study. Am J Epidemiol 1980; 111:721-7.
(9) Dodge R. & Burrows B. - The Prevalence and Incidence of Asthma and Asthma-like symptoms in a general population sample. American Review of Respiratory Disease 1980; 122:567-575.
(10) Mitchell, E A & Dawson KP. - Why are hospital admissions at children with acute asthma increasing? Enr Resp J 1989; 2:470-472..
(11) Mitchell E A & Borman B. - Demographic Characteristics of asthma admissions to hospitals. N Z Med J 1986; 99:576-79..
(12) Evans R.- Recent observencion refleting increases in mortality from asthma . J Allergy Clin Immunol 1987; 80:377-379.
(13) Mullally DI., Howard W A, Hubbard TJ, Grauman JS. - Increased hospitalizations for Asthma Among Children in the Washington, D.C. Area during 1961-1981. Am Allergy 1984; 53:15-9..
(14) Gerstman BB., Bosco LA, Tomita DK. - Trends in the prevalence of asthma hospitalization in the 5 to 14 years-old. Michigan Medicaire population, 1980 to 1986. J Allergy Clin Immunol 1993; 91:838-43..
(15) Stolley PD. - Asthma Mortality. American Review of Respiratory Disease 1972; 105:883-85.
(16) Jackson RT., Beaglehole R; Rea HH., Sutherland DC. - Mortality from asthma: a new epidemic in New Zealand. British Medical Journal 1982; 285:771-774..
(17) U.S. Departament of Health and Human Services. Public Health Service National Institute of Health. International Consensus Report on Diagnosis and Managment of Asthma. Bethesda, Maryland 1992. Publication nº 92 - 3091.
(18)Coutlas DB & Samet JH. - Epidemiology and Natural History of Childhood Asthma. Tinkelman DG., Falliers CJ, Naspitz CK, ed - Childhood Asthma: pathophysiology and tratment. New York, Marcel Dekker Luc 1987. p. 131-57.
(19) Asher M.I., Keil U., Anderson H.R., Beasley R., Crane J., Martinez F., Mitchell E.A., Pearce N., Sibbald B., Stewart W., Strachan D., Weiland S.K., Williams H.C. - Internacional study of asthma and allergies in childhod (ISAAC): rationale and methods. Eur Respir J, 1995, 8, 483-491.
(20) Ferris BG. - Recomended Respiratory Disease Questionnaires for use with Adults and Children in Epidemiological Research. Am Ver Resp Dis 1978; 188:1-79.
(21) Mari JJ & Williams P.- A comparison of the validity of two psychiatric screening questionnaires (GHQ-12 and SRQ-20) in Brazil, using Relative Operating Characteristic (ROC) analysis. Psycological Medicine 1985; 15:651-659.
APPENDIX
QUESTIONNAIRE ATS-DLD-78-C-M