• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Clin Epidemiol. Author manuscript; available in PMC Aug 1, 2012.
Published in final edited form as:
PMCID: PMC3124572
NIHMSID: NIHMS271667

High validity of mother-reported use of anti-asthmatics among children: a comparison with a population-based prescription database

Abstract

Objectives

To examine the validity of: 1) maternal questionnaire report of children's use of anti-asthmatics using a prescription database as the reference standard, 2) dispensed anti-asthmatics as a measure of asthma using maternal report of children's asthma as the reference standard.

Study Design and Setting

3394 children in the Norwegian Mother and Child Cohort Study (MoBa) aged seven were linked to the Norwegian Prescription Database (NorPD). Maternal report of both children's use of anti-asthmatics during the preceding year and of the presence of asthma was compared with data on dispensed anti-asthmatics.

Results

2056 mothers responded and reported use of anti-asthmatics the previous year in 125 of 147 children who had been dispensed anti-asthmatics (sensitivity 85.0%). Of 1909 children with no dispensed anti-asthmatics, 1848 had no maternal report of anti-asthmatic use (specificity 96.8%). Mothers reported current asthma in 133 (6.5% of 2056) children, including 122 (5.9%) reported as verified by a doctor. Of these 122, 98 had been dispensed anti-asthmatics during the preceding year (sensitivity 80.3%). Only 1.2% of the children without reported asthma were dispensed anti-asthmatics.

Conclusion

Mother-reported use of anti-asthmatics during the previous year among 7 year old children is highly valid. Dispensed anti-asthmatics would be a useful proxy for the presence of current asthma when disease data are not available.

Keywords: asthma, children, mother-reported, pharmacoepidemiology, prescription database, validity

Introduction

Asthma is the most common chronic disease among children in developed countries. In epidemiological studies, questionnaire reports of asthma diagnosis and medication use have been a central source of information for assessing the prevalence of asthma [1, 2]. However, self or parental report of asthma may not be a sufficient way of defining asthma in studies of risk factors [3, 4]. The validity of parental reporting of pediatric asthma and asthma drug exposure has also been questioned because of possible recall bias which may influence the prevalence estimates and surveillance of trends in asthma [5-7]. Therefore, it is useful to have more objective data to compare with parental reports in different populations. In large birth cohorts of children, rates of response to questionnaires may decrease over time as children get older. Thus, the availability of population-based prescription registry data can serve as surrogate disease endpoints among non-responders as well as enabling the assessment of non-response bias. Prescription data are a source of objective data that have become available in some countries, recently also in Norway [8].

Most studies on the validity of self-reported drug use have been performed in adult populations [7, 9]. Worldwide, there are very few validation studies on maternal reporting of drug use among children, using registry data as the reference source. Pharmacy data in prescription databases may be a feasible reference standard in validity studies of parental reporting on children's use of anti-asthmatics [10]. Norway has had a complete, nationwide prescription database since January 2004 [11].

The objectives of this study were to examine:

  1. the validity of maternal report of children's use of anti-asthmatics in a pregnancy cohort, the Norwegian Mother and Child Cohort Study (MoBa), using the Norwegian Prescription Database (NorPD) as the reference standard, and
  2. the validity of dispensed anti-asthmatics recorded in NorPD as a measure of asthma using maternal report of children's asthma in MoBa as the reference standard.

1. Materials and methods

The present study is based on the MoBa conducted by the Norwegian Institute of Public Health which is a prospective nationwide pregnancy cohort enrolling pregnant women and their partners throughout Norway [12]. Data from MoBa on the children's use of anti-asthmatics and asthma history as reported by mothers in the year after the child turned seven years were compared with data on dispensed anti-asthmatics recorded in the nationwide NorPD. Data from MoBa and NorPD were linked using the unique 11-digit personal identity number, assigned to all individuals living in Norway.

a. Norwegian Mother and Child Cohort Study (MoBa)

During 1999-2008, more than 100 000 pregnancies were included. The mother received three questionnaires during pregnancy while the father received one. After birth the parents receive several questionnaires at different time points. Our analysis is based on the first follow-up of children in MoBa who reached the age of 7 years. In the period March 2008 - June 2008, 3394 mothers in MoBa were mailed a questionnaire which included several items on asthma and allergy. Of the 3394 mothers who received the mailing, 2 056 responded giving a 60.6 % response rate. The mothers reported the date of questionnaire completion, and for those with missing date we used the return date minus 5 days. One was excluded from the study because of an invalid date. Non-responders were also linked to the prescription database to examine if the prevalence of dispensed anti-asthmatics in NorPD differed between non-responders and responders.

b. The Norwegian Prescription Database (NorPD)

From January 2004 all Norwegian pharmacies have been required to send data electronically each month to the Norwegian Institute of Public Health on all prescribed drugs (irrespective of reimbursement or not) redeemed by individuals in ambulatory care [11]. Variables from NorPD used in our study were: unique encrypted personal identity number, date of dispensing and information on drugs dispensed. All medicines in Norway are classified according to the Anatomical Therapeutic Chemical Classification System (ATC) [13]. Asthma drugs (called anti-asthmatics throughout this manuscript) included in our study (table 1): inhaled β2-agonists, inhaled glucocorticoids, combination inhalers with β2-agonists and glucocorticoids, and leukotriene receptor antagonists. We did not include oral mixtures with β2-agonists because they are mainly used to treat children for incidental, unspecific acute upper respiratory tract symptoms and infections. For both aims, data from NorPD was extracted for each responder for a 365 day period preceding the date of completing the MoBa questionnaire. A similar 365 day period with prescription data was extracted for each non-responder using the 15th in the month the questionnaire was sent out as the reference date. As a measure of robustness, we analyzed the time period 366 to1095 days (1-3 years) preceding the completing date to assess the discordance between maternal reports in the questionnaire and the recording of dispensed drugs in NorPD.

Table 1
List of anti-asthmatic drug categories with Anatomical Therapeutic Chemical (ATC) codes and generic names of each substance included.

c. Analysis strategy

One-year prevalence of being dispensed at least one anti-asthmatic in 2008 is presented for responders and non-responders.

The following MoBa question was used in the validity analysis to answer aim 1:

Q1. “Has the child used spray, inhaler or other medications for asthma during the past year?” (no/yes).

  1. “If yes, name of medication used on a regular basis” (the listed names were coded according to their ATC codes [13])
  2. “If yes, name of medication used during attacks” (the listed names were coded according to their ATC codes [13])

Data on dispensed drugs as recorded in NorPD is the reference standard in this validity analysis. Two measures of validity (sensitivity and specificity) were calculated with 95% confidence intervals (CI) by the continuity corrected score interval method [14]. The positive predictive value (PPV) and negative predictive value (NPV) of mother-reported use of anti-asthmatics were also estimated. Validity was assessed in two steps. First, we used question 1, i.e. the sensitivity measures to what degree mother-reported use of anti-asthmatics (no/yes) in the MoBa questionnaire correctly identifies individuals who had been dispensed one of the four anti-asthmatic groups (table 1) at least once in the 365 days before completion of the MoBa questionnaire [9]. Specificity is the proportion reporting “no use” in MoBa among children with no records of dispensed anti-asthmatics in NorPD during the 365 day period.

Second, we used the answers to question 1a together with 1b to perform the same kind of validity analysis on the substance level for salbutamol and fluticasone which were the two most frequently reported anti-asthmatics used by MoBa children.

To answer the second aim of our study we extracted information on childhood asthma from the MoBa questionnaire for 7 year old children. The questionnaire queried about 30 illnesses or health problems including asthma (“Has the child ever experienced any of the following long-lasting illnesses or health problems?”). For each condition the mothers were asked to tick whether the child had ever experienced the condition, if the child still had the condition, and if it had been verified by a physician. On the basis of these three answers regarding asthma, the 2056 responders were divided into four mutually exclusive groups; current asthma with physician diagnosis, current asthma but without physician diagnosis, asthma ever but not current, no report of asthma. The validity was estimated in the same way as for aim 1 but using MoBa as the reference standard. Both children with and without dispensed anti-asthmatics in NorPD during the previous 365 days were included in the validity calculations.

Furthermore, to examine how many children in each of the four groups may have used anti-asthmatics which had been dispensed more than 365 days before completion of the questionnaire, the period was extended to 1095 days (3 years).

Statistical analyses were conducted using SPSS 17.0.

Ethics

The MoBa study was approved by the Regional Committees for Medical Research Ethics (S-95113) and The Norwegian Social Science Data Services (01/4325-6). Informed consent was obtained from mothers before inclusion and the record-linkage was approved by the Norwegian Data Inspectorate.

2. Results

a. Prevalence of dispensed anti-asthmatics as recorded in NorPD

The one-year prevalence of dispensed anti-asthmatics in 2008 recorded in NorPD was 7.6 % (95 % CI 5.9-9.4) among responders and 7.9 % (5.8-9.9) among non-responders to the MoBa seven year questionnaire.

b. Mother-reported overall use of anti-asthmatics versus dispensed anti-asthmatics

Mothers reported use of anti-asthmatics among their children during the preceding year in 125 of 147 children (table 2) who redeemed prescriptions in NorPD giving a sensitivity of 85.0%. Of 1909 children with no record of dispensed anti-asthmatics in NorPD 1848 had no maternal report of anti-asthmatic use in MoBa, giving a specificity of 96.8%. Among the 61 children who had no anti-asthmatics dispensed in the last 365 days despite mother-reported use in MoBa, slightly more than hal of the children (32) had been dispensed anti-asthmatics during the period 366-1095 days before completion of the questionnaire.

Table 2
Mother-reported use of anti-asthmatics in the Mother and Child Cohort Study (MoBa) questionnaire (n=2056) versus dispensed anti-asthmatics in the Norwegian Prescription Database (NorPD)

c. Mother-reported use of specific anti-asthmatic substances

The sensitivity of mother-reported use of salbutamol during the preceding year was 71.8% and specificity 98.2% (table 2). Among the 36 individuals who had no salbutamol dispensed according to NorPD within 365 days before the filling in date despite mother-reported use in MoBa, 24 had received salbutamol 366-1095 days before completion of the questionnaire.

The sensitivity of mother-reported use of fluticasone during the preceding year was 74.1% and specificity 99.2% (table 2). Among the 15 individuals who had no fluticasone dispensed according to NorPD within 365 days before the filling in date despite mother-reported use in MoBa, all but one had received fluticasone 366-1095 days before completion of the questionnaire.

d. Dispensed anti-asthmatics versus mother-reported asthma diagnosis

Mothers reported current asthma in 133 (6.5% of 2056) children, including 122 (5.9 %) who were reported to be diagnosed by a doctor. Ninety-eight of these 122 children had been dispensed anti-asthmatics in the 365 days preceding the completion date of the MoBa questionnaire giving a sensitivity of 80.3%. Among those with current asthma but not verified by a physician the sensitivity was 54.6% (table 3).

Table 3
Dispensed anti-asthmatics in the Norwegian Prescription Database (NorPD) versus maternal report of children's asthma diagnosis in the Mother and Child Cohort Study (MoBa) questionnaire (n= 2056)

When increasing the period to 1065 days before completing date 115 of the 122 children with current doctor-diagnosed asthma had been dispensed anti-asthmatics, increasing the sensitivity to 94.3 % and decreasing specificity to 92.8 % for the children with current and physician diagnosed asthma. For those with current asthma but not verified by a doctor the sensitivity increased to 72.7 % with specificity 87.9%. 1805 children were classified as no asthma because they had not answered yes to any of the asthma questions. Only 1.2 % of these children received anti-asthmatics during the year before the questionnaire.

Discussion

In this large population-based sample, we showed that mother-reported use of anti-asthmatics in general and of specific brand names of anti-asthmatics in children aged 7 years had high sensitivity and specificity when using NorPD as the reference standard. In addition, we showed that the sensitivity for dispensed anti-asthmatics during the past year recorded in NorPD was 80% for children with current physician diagnosed asthma, increasing to 94% when taking into account anti-asthmatics dispensed during the previous three years, using maternal report of their child's asthma as the reference standard. Only 1 % of children without reported asthma were dispensed anti-asthmatics during the previous year. This result suggests that the prescription database can provide useful surrogate information on the presence or absence of asthma when questionnaire data are not available. In prospective birth cohorts requiring active follow-up, non-response to questionnaires as children get older can be an important source of bias. In this study, the prevalence of dispensed anti-asthmatic medications was similar in responders and non-responders, suggesting a lack of response bias with respect to asthma risk factors. The availability of prescription data for the non-responders can enhance the validity of studies of early life factors in relation to the development of asthma.

Strengths and limitations

The MoBa questionnaire

The design and phrasing of the questions may influence the overall validity. In our questionnaire we used a question about overall use of anti-asthmatics (no/yes) and also asked mothers to write the trade names of anti-asthmatics used regularly and/or used for attacks. The open-ended questions on trade names of anti-asthmatics lead to a lower sensitivity than for the overall question on use of anti-asthmatics, but still the sensitivity was over 70% for both salbutamol and fluticasone. This is in accordance with Klungel et al. who have shown that open-ended questions had lower sensitivity than questions that specified the exact name of the disease and type of drug [15]. As shown by Nilsen et al the proportion of nulliparous women 25-34 years of age is higher and the proportion of smokers lower in MoBa compared to the total population [16]. In contrast, the prevalence of pregnancy complications and diseases are similar. For example, the prevalence of asthma is about 4% both in MoBa and in the total population. The results of this study suggested that prevalence estimates of some exposures and outcomes could be biased because of self-selection, but this is not necessarily the case for estimates of exposure-outcome associations[16].

The Norwegian Prescription Database

A strength of our study is the use of the unique 11-digit identification number which ensured valid data-linkage between the questionnaire and the prescription database [11]. In addition, prescription data is considered to be more complete than medical records and information elicited from interviews and questionnaires [17-19]. The completeness and accuracy of NorPD is high, because of legislation and other incentives motivating pharmacies to collect and send the data electronically to the national database on all dispensed drugs to individuals in ambulatory care. Coverage of NorPD for total pharmacy records in Norway to individual outpatients is 100% [8]. However, it is important to realize that none of the prescription databases can address whether the medication was actually used, which is what the questionnaire asks. Since only information about drugs dispensed and purchased by patients is entered into NorPD, primary non-compliance, failure to fill the prescription, is not an issue in our study [20].

If the child was dispensed an anti-asthmatic for indications other than asthma, e.g. acute upper respiratory tract infection or bronchitis, the mother may not report this drug use when asked specifically about asthma. However, this was very rare in our study, only 1% of children without mother-reported asthma had been dispensed anti-asthmatics. Inconsistency between the database and the questionnaire also occurs when mothers report use of anti-asthmatics even though no anti-asthmatics were recorded in the database for the recall period of 12 months. In our study 61 children with mother-reported anti-asthmatics, 36 children with salbutamol and 15 with fluticasone were not recorded as being dispensed these drugs in the 365 days before date of reporting. However, when we increased the period to 1095 days before the completing date, 32 of 61 children with maternal report of anti-asthmatics, 24 of 36 with maternal report of salbutamol, and 14 of 15 children with maternal report of fluticasone had received anti-asthmatics according to the prescription database. These children may have asymptomatic phases of their asthma without the need for anti-asthmatics and therefore their asthma drugs may be used intermittently for longer periods. However, discordance seems to be a small problem in our study with high estimated sensitivity and specificity even for the open-ended questions where we asked the mothers to report the brand names of the anti-asthmatics. Of note, we would not expect 100% concordance because mothers were asked about use of drugs during the previous year as opposed to dispensing of drugs which is what is recorded in the prescription database. For a variety of reasons, a small number of children will not necessarily use the dispensed drugs.

The response rate to our questionnaire was 60%. Therefore selection bias may have occurred and could theoretically bias the concordance between questionnaire and registry data in either direction. However, the non-responders to the questionnaire had the same prevalence of dispensed anti-asthmatics as the responders and therefore it is unlikely that the 60% response rate biased our results.

Other studies

There are few validation studies on parental-reported drug use in children using complete prescription data as the reference source. Sensitivity and specificity in this study correspond well with a Danish study on parental-reported use of anti-asthmatics in children aged 6-8 years which found a sensitivity of 91.6% and specificity of 95.6% [10]. A diagnosis of asthma is difficult to make before the child is 4-5 years old and asthma symptoms may therefore not be communicated as asthma in a survey [21]. The older age of the children in our study, 7 years, may partly explain why we found a higher validity than studies of younger children. According to a study by Skurtveit et al, self-reported use of anti-asthmatics among adolescents 15-16 years old had a lower sensitivity (75.0%) than we found among children aged 7 years [18]. This may be explained by a lower adherence in adolescents.

Data presented by Moth et al [22] suggest that register-based data on redeemed prescriptions can be utilized to identify asthmatic school children. Our data support the value of prescription data on dispensed asthma drugs to identify asthmatic children when questionnaire data are not available. Where children were reported to have current asthma verified by a doctor, 80% had redeemed a prescription for an anti-asthmatic in the past year and 94% during the past three years. The very low proportion of dispensed asthma drugs among children without a report of current diagnosed asthma also supports the value of prescription data to characterize children as currently asthmatic versus non-asthmatic in epidemiologic analyses when questionnaire responses are not available.

Conclusion

Our study confirms that mother-reported use of anti-asthmatics during the previous year among 7 year old children is highly valid. In addition, our data suggests that prescription data on dispensed anti-asthmatics would be a useful proxy for the presence or absence of current asthma when disease data are not available.

Acknowledgments

The Norwegian Mother and Child Cohort Study is supported by the Norwegian Ministry of Health, NIH/NIEHS (contract no N01-ES-85433), NIH/NINDS (grant no. UO1 NS 047537-01), and the Norwegian Research Council/FUGE (grant no.151918/S10). Dr. London is supported by the Intramural Research Program, NIEHS, NIH, DHHS.

References

1. Worldwide variation in prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and atopic eczema: ISAAC. The International Study of Asthma and Allergies in Childhood (ISAAC) Steering Committee. Lancet. 1998;351:1225–32. [PubMed]
2. Burney PG, Laitinen LA, Perdrizet S, Huckauf H, Tattersfield AE, Chinn S, Poisson N, Heeren A, Britton JR, Jones T. Validity and repeatability of the IUATLD (1984) Bronchial Symptoms Questionnaire: an international comparison. Eur Respir J. 1989;2:940–5. [PubMed]
3. Pekkanen J, Sunyer J, Anto JM, Burney P. Operational definitions of asthma in studies on its aetiology. Eur Respir J. 2005;26:28–35. [PubMed]
4. van Wonderen KE, van der Mark LB, Mohrs J, Bindels PJ, van Aalderen WM, Ter RG. Different definitions in childhood asthma: how dependable is the dependent variable. Eur Respir J. 2009 Dec 23; [Epub ahead of print] 2009. [PubMed]
5. Roberts EM. Does your child have asthma? Parent reports and medication use for pediatric asthma. Arch Pediatr Adolesc Med. 2003;157:449–55. [PubMed]
6. Crane J, Mallol J, Beasley R, Stewart A, Asher MI. Agreement between written and video questions for comparing asthma symptoms in ISAAC. Eur Respir J. 2003;21:455–61. [PubMed]
7. Smeeton NC, Rona RJ, Oyarzun M, Diaz PV. Agreement between Responses to a Standardized Asthma Questionnaire and a Questionnaire following a Demonstration of Asthma Symptoms in Adults. Am J Epidemiol. 2006;163:384–91. [PubMed]
8. Furu K, Wettermark B, Andersen M, Martikainen JE, Almarsdottir AB, Sorensen HT. The Nordic countries as a cohort for pharmacoepidemiological research. Basic Clin Pharmacol Toxicol. 2010;106:86–94. [PubMed]
9. West SL, Strom BL, Poole C. Validity of pharmacoepidemiology drug and diagnosis data. In: Strom BL, editor. Pharmacoepidemiology. Chichester: John Wiley & Sons Ltd; 2005. pp. 709–65.
10. Wogelius P, Poulsen S, Sorensen HT. Validity of parental-reported questionnaire data on Danish children's use of asthma-drugs: a comparison with a population-based prescription database. Eur J Epidemiol. 2005;20:17–22. [PubMed]
11. Furu K. Establishment of the nationwide Norwegian Prescription Database (NorPD) - new opportunities for research in pharmacoepidemiology in Norway. Nor J Epidemiol. 2008;18:129–36.
12. Magnus P, Irgens LM, Haug K, Nystad W, Skjaerven R, Stoltenberg C. Cohort profile: the Norwegian Mother and Child Cohort Study (MoBa) Int J Epidemiol. 2006;35:1146–50. [PubMed]
13. WHO Collaborating Centre for Drug Statistics Methodology. Norwegian Institute of Public Health Guidelines for ATC classification and DDD assignment. Oslo, Norway: 2005.
14. Vollset SE. Confidence intervals for a binomial proportion. Stat Med. 1993;12:809–24. [PubMed]
15. Klungel OH, de Boer A, Paes AH, Herings RM, Seidell JC, Bakker A. Influence of question structure on the recall of self-reported drug use. J Clin Epidemiol. 2000;53:273–7. [PubMed]
16. Nilsen RM, Vollset SE, Gjessing HK, Skjaerven R, Melve KK, Schreuder P, Alsaker ER, Haug K, Daltveit AK, Magnus P. Self-selection and bias in a large prospective pregnancy cohort in Norway. Paediatr Perinat Epidemiol. 2009;23:597–608. [PubMed]
17. West SL, Savitz DA, Koch G, Strom BL, Guess HA, Hartzema A. Recall accuracy for prescription medications: self-report compared with database information. Am J Epidemiol. 1995;142:1103–12. [PubMed]
18. Skurtveit S, Selmer R, Tverdal A, Furu K. The validity of self-reported prescription medication use among adolescents varied by therapeutic class. J Clin Epidemiol. 2008;61:714–7. [PubMed]
19. Haukka J, Suvisaari J, Tuulio-Henriksson A, Lonnqvist J. High concordance between self-reported medication and official prescription database information. Eur J Clin Pharmacol. 2007;63:1069–74. [PubMed]
20. Beardon PH, McGilchrist MM, McKendrick AD, McDevitt DG, MacDonald TM. Primary non-compliance with prescribed medication in primary care. BMJ. 1993;307:846–8. [PMC free article] [PubMed]
21. Global Initiative for Asthma (GINA) NIH Publication No 02-3659 Issued January, 1995(updated 2002) Management Segment (Chapter 7) 2006. Global Strategy for Asthma Management and Prevention. Updated 2005 from the 2004 document.
22. Moth G, Vedsted P, Schiotz P. Identification of asthmatic children using prescription data and diagnosis. Eur J Clin Pharmacol. 2007;63:605–11. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...