pharmacoepidemiology and drug safety 2010; 19: 400407

Published online 24 November 2009 in Wiley InterScience (www.interscience.wiley.com) DOI : 10.1002/pds.1887

ORIGINAL REPORT

Factors associated with medicine use among children aged under 12 yearsa population survey in Finlandy
meen-Anttila PhD1*, Leena Lindell-Osuagwu MSc (Pharm)1, Kati Sepponen MSc (Pharm)1, Katri Ha Kirsti Vainio PhD1 , Pirjo Halonen MSc (Stat)2 and Riitta Ahonen PhD1
1 2

Department of Social Pharmacy, University of Kuopio, Kuopio, Finland Centre for Information and Learning Resource Services, University of Kuopio, Kuopio, Finland

SUMMARY
Purpose The goal was to determine the prevalence of medicine use and to provide population-based information on factors associated with medicine use, including prescribed and over-the-counter (OTC) medicines, in children aged under 12 years. Methods A cross-sectional population survey of a random sample of children aged under 12 years (n 6000) was carried out in Finland in spring 2007, with a response rate of 67%. A questionnaire was sent to their parents. Current use of medicines prescribed by a physician and use of OTC medicines during the preceding two days were the main outcome measures. Results The prevalence of current prescribed medicine use was 17%, and the 2-day prevalence of OTC medicine use 17% (vitamins excluded). The use of prescribed medicines was higher among children with illnesses diagnosed by a physician. Additionally, predictors for the use of prescribed medicines were young age among boys, health status worse than good, and the use of any prescribed medicines by a parent. The predictors for the use of OTC medicines by the child were young age and fairly good to poor health status, and the use of OTC medicines by a parent. However, children with any illnesses diagnosed by a physician were less likely to use OTC medicines. Conclusions A considerable proportion of children use prescription and OTC medicines in Finland. Young age and health-related factors, morbidity and health status, and parental medicine use predicted childrens medicine use. Further studies are needed to examine the association between parents and childrens medicine use. Copyright # 2009 John Wiley & Sons, Ltd. key words medicine use; child; prescribed medicines; over-the-counter medicines; survey Received 11 March 2009; Revised 30 September 2009; Accepted 6 October 2009

INTRODUCTION Childrens medication has increasingly gained attention in the European Union and in the United States, and new regulations have been enacted accordingly.1,2 Besides accentuating the need for clinical trials with children, these regulations emphasize the need to gather all the available information on medicine use among children. Such information is also needed by
meen-Anttila, Department of Social Pharmacy, * Correspondence to: K. Ha University of Kuopio, P.O. Box 1627, FIN-70211 Kuopio, Finland. E-mail: katri.hameen-anttila@uku. y No external funding was obtained to conduct this population survey. All the work was done at the Department of Social Pharmacy, University of Kuopio. The costs have been covered by the Department. The authors do not have any conicts of interests.

health care professionals. However, few studies have investigated the prevalence of medicine use among children, especially before adolescence.35 The factors which have been found to be associated with higher prevalence of medicine use include poor health status,3,6 maternal medicine use,7 and young age.8,9 Previous studies have found that the association between medicine use and the gender of the child is twofold: in younger age groups, the use of medicines is more common among boys than girls, while in adolescence it is more common among girls.6,8,10,11 Based on studies conducted in the United States, the level of the mothers education and household income seem to correlate with the childs medicine use.3,7 Mothers with more education and higher household incomes are more likely to give OTC medicines to their

Copyright # 2009 John Wiley & Sons, Ltd.

factors associated with medicine use among nnish children

401

children than those with less education and lower incomes. Previous studies on childrens medicine use have several limitations. Most of the studies have reported the use of prescribed medicines on the basis of physician records10,12,13 or national prescription registers.8,9 Other studies have focused on the use of single but widely used and somehow problematic groups of medicines, e.g., antibiotics.1416 Medicine use among children has also been studied as a part of larger health surveys, producing quite supercial knowledge about medicines.5,11,17 Most of the studies have focused on adolescents, and children have often been recruited to the studies through their schools, which raises questions about the generalizability of the results.6,1820 Moreover, only a few studies describe OTC medicine use among children.3,21 The aim of this study was to assess the prevalence of medicine use, including prescribed and OTC medicines, and to identify the factors associated with medicine use among children aged under 12 years in a Finnish general population. METHODS Context: Finnish health care In Finland, health care is nanced mainly by taxes, and primary health care is available to all citizens for ambulatory medical services in health centers, free of charge for children and for a small fee for adults. Outpatient visits can also be made to private health care clinics. All children are covered for health care in the public sector from birth, and also during the mothers pregnancy. The maternity and child health care system includes prevention, health checkups, and health education until the child is of school age, after which the school health care system takes over these tasks. Medicines, both OTC and prescription, can only be bought from pharmacies. The Social Insurance Institution reimburses the patient for a part of the cost of prescribed medicines.

person who usually takes care of the childs medication was instructed to answer the questionnaire. The childs name was printed on the questionnaire in order to specify the child in families with two or more children. Two reminders were sent. The questionnaire was available in Finnish and Swedish, both of which are ofcial languages in Finland. The six-page questionnaire contained the following ve sections: background information of the child, health status of the child, the childs medicine use, sources of information about medicines, and background information of the responding parent, including the parents own medicine use. The questionnaire was pilot-tested with a convenience sample of mothers (n 61) selected by the research group, and modied on the basis of the pilot. The sample size was calculated on the basis of the information on medicine use among children from a previous population study in Finland.22 We used a desired signicance level of a < 0.05 and the expected power of 80%. In order to gain statistically signicant differences in medicine use between the age groups of 02, 36, and 711 years, the sample needed for each group was 1315, so altogether 3945 children were needed. A sample size of 6000 with a response rate of 70% was expected to be sufcient. Of the 6000 questionnaires sent, 4121 were returned, yielding a response rate of 69% (Figure 1). Although the name of the child was printed on the questionnaire, some parents (n 89) lled in the questionnaire for one

Design and study population This cross-sectional population survey was conducted in Finland in spring 2007. A random sample (n 6000) of children under 12 years of age was taken. The sampling was conducted by the Finnish Population Register Center from its database, which contains constantly updated information on everyone living permanently in Finland. A questionnaire was sent to one parent of each child, primarily to the mother. The
Copyright # 2009 John Wiley & Sons, Ltd.

Figure 1. Formation of the nal study population

Pharmacoepidemiology and Drug Safety, 2010; 19: 400407 DOI: 10.1002/pds

402

meen-anttila k. ha

ET AL.

of their other children. These 89 were not included in the study population, and the analysis was done with 4032 responses giving a nal response rate of 67%. The nal study population was representative in age and gender of children aged under 12 years and who were living in Finland. The analysis of the non-respondents showed no differences from the target population in age, gender, or regional distribution, and the response rate was considered sufcient to produce non-biased results.23 Questionnaires were completed by the mother (95%), father (4%), or both (1%). Of the respondents, 67% were working or studying, and 28% were at home with children (Table 1). The mean age was 36 years (range 1861). Children from all six regions of Finland were included: Southern, 40.7% (the actual proportion of the children in that area is 41.1%); Western, 32.6% (34.9%); Eastern, 12.8% (9.9%); Oulu region, 10.1% land 0.5% (10.4%); Lapland, 3.4% (3.3%); and A (0.5%). National and local ethical guidelines for researchers were followed in carrying out the study. The anonymity of the respondents was assured by deleting the name of the child from the questionnaire as soon as it was returned. The data management and disposal of all personal data were conducted in accordance with national privacy protection laws. Denitions of medicine use The indicators for medicine use were taken from previous studies carried out in Finland to allow comparisons of the results.5,22 The use of prescribed medicines was determined by asking the parents to report the current use of any medicines prescribed by a physician, and write a list of all medicines. They were also asked to report the use of OTC medicines during the preceding 2 days. All listed medicines were coded according to the ATC classication system.24 The reported use of vitamins was excluded in this study, since the main focus was on medicine use meant to relieve symptoms and treat illnesses. Statistical methods Data were analyzed with SPSS for Windows statistical software, Release 14.0 (SPSS Inc., Chicago, IL, USA). The categorical variables were cross-tabulated and their potential dependencies were estimated with x2 tests. P values less than 0.05 were considered signicant. In order to determine the adjusted odds ratios (ORs) of various factors for medicine use by the child, logistic
Copyright # 2009 John Wiley & Sons, Ltd.

regression analyses were conducted for prescribed and OTC medicine use, and the total exposure to medicines (prescribed and/or OTC medicine use). Prior to the regression analyses, interdependencies (Spearman correlation coefcient or contingency coefcient) of the variables shown in Table 1 were calculated. Relatively low values (range r 0.0020.423 Spearman; C 0.0080.397 contingency coefcient) suggested that all variables were sufciently independent to be included in the initial models. Next, variables were deleted from the models one by one on the basis of their statistical non-signicance. Changes of the b values were assessed at each step, a 20% change being considered a threshold value. The dependence between health status and the number of symptoms experienced (Spearman r 0.370) was assessed by deleting the number of symptoms experienced from the models. This increased the b values of health status considerably, and suggested that the number of symptoms experienced might be a confounder. Therefore, the nal models are adjusted for this variable. Furthermore, age and sex seemed to interact and this was explored by using the interaction variable in the model. The ORs in the presence of interaction were estimated and even though the likelihood ratio test showed no statistical signicance, some of the ORs were found statistically signicant in a detailed inspection within ages and genders.25 Thus, the interaction variable was included in the nal models to discover this phenomenon in a more detailed way. The results are presented as ORs together with their 95% condence intervals (CIs).

RESULTS About one in six of the children (17%) were currently using prescribed medicines, and an equal proportion (17%) had used one or more OTC medicines (vitamins excluded) on the preceding 2 days. A total of 151 children (4%) used both prescribed and OTC medicines. Children had on average 1.3 symptoms reported by the parent from a list of symptoms (range 013), most commonly common cold (33% of the children), rash (24%), atus (11%), and growing pains (9%). Altogether, 18% of the children had at least one illness diagnosed by a physician. The most common illnesses reported were allergy (4% of the children), asthma (4%), atopic eczema (3%), and ear infection (2%). A total of 169 children (4%) had two or three illnesses diagnosed by a physician. The majority of the parents (96%) rated their childs health status as good or fairly good (Table 1).
Pharmacoepidemiology and Drug Safety, 2010; 19: 400407 DOI: 10.1002/pds

Table 1. Characteristics of the study population and childs medicine use by selected characteristics of the child and the respondent (N 4032) Whole study population No. (% of the responses) Prescribed medicine use No. (%) Pearson x2 p Pearson x2 p 0.556 470 (24.4) 601 (31.2) 855 (44.4) 0 0.001 534 (25.4) 686 (32.6) 886 (42.1) 0 120 (22.8) 113 (16.7) 134 (15.2) 24 129 (24.6) 99 (14.7) 107 (12.4) 43 <0.0005 215 (40.9) 189 (28.0) 219 (25.1) 34 77 (16.7) 104 (17.4) 128 (15.3) 34 103 (22.4) 103 (17.4) 114 (13.8) 47 160 (34.7) 175 (29.4) 222 (26.7) 39 <0.0005 0.011 OTC medicine use (vitamins excluded) No. (%) Total exposure to medicines (prescribed and/or OTC medicine use) No. (%) Pearson x2 p

Characteristic

Copyright # 2009 John Wiley & Sons, Ltd.

<0.0005 0.184 1662 (41.4) 1325 (33.0) 1027 (25.6) 18 <0.0005 411 (12.9) 186 (30.4) 53 (40.5) 106 65 (4.9) 192 (15.1) 396 (30.6) 135 <0.0005 3281 (81.8) 559 (13.9) 169 (4.2) 23 214 (6.6) 323 (58.6) 133 (78.7) 80 507 (15.7) 108 (20.3) 32 (19.8) 111 0.017 668 (20.8) 360 (65.5) 141 (83.9) 95 3249 (80.9) 631 (15.7) 135 (3.4) 17 <0.0005 1353 (34.0) 1298 (32.6) 1334 (33.5) 47 74 (5.6) 189 (14.7) 409 (31.2) 102 353 (11.0) 232 (37.3) 82 (61.7) 75 282 (17.3) 237 (18.1) 155 (15.3) 75 255 (15.7) 243 (18.8) 151 (15.1) 108 0.029 478 (29.3) 425 (32.6) 269 (26.6) 90 <0.0005 703 (22.0) 365 (58.8) 102 (76.7) 89 <0.0005 135 (10.2) 357 (27.9) 682 (51.9) 117 <0.0005 0.007 <0.0005

factors associated with medicine use among nnish children

<0.0005

Child Female: age, years 02 36 711 Missing cases Male: age, years 02 36 711 Missing cases Mean age 5.57 Median age 6.00 Range 011 Birth order First Second Third Missing cases Health status (rated by the parent) Good Fairly good Moderate or poor Missing cases Symptoms experienced, no. (reported by the parent from a list of symptoms) 0 1 2 Missing cases Mean 1.31 Range 013 Illnesses diagnosed by a physician, no. 0 1 2 Missing cases Mean 0.23 Range 03 0.019 862 (21.6) 2866 (71.9) 256 (6.4) 48 165 (19.5) 470 (16.6) 31 (12.3) 105 156 (18.6) 466 (16.6) 27 (10.9) 136 0.017 275 (32.5) 837 (29.7) 55 (22.2) 119

Pharmacoepidemiology and Drug Safety, 2010; 19: 400407 DOI: 10.1002/pds

Parent (95% mothers) Age, years 30 3145 46 Missing cases Mean 35.9 Median 36 Range 1861

0.008

403
(Continues)

Table 1. (Continued)

404

Characteristic OTC medicine use (vitamins excluded) No. (%) 0.013 364 (15.4) 286 (18.4) 116 0.001 2853 (71.8) 1118 (28.2) 61 0.509 252 (6.3) 2456 (61.4) 1291 (32.3) 33 0.942 2698 (67.3) 1119 (27.9) 194 (4.8) 21 0.915 797 (20.7) 1292 (33.5) 1419 (36.8) 347 (9.0) 177 127 (16.4) 222 (17.4) 242 (17.2) 56 (16.3) 231 135 (17.5) 227 (17.8) 216 (15.6) 49 (14.3) 260 78 109 0.235 230 (29.8) 396 (31.1) 413 (29.6) 92 (26.8) 245 449 (16.9) 189 (17.2) 34 (17.7) 410 (15.5) 209 (19.1) 33 (17.4) 0.028 767 (28.9) 352 (32.1) 56 (29.3) 92 0.486 35 (14.5) 410 (16.9) 224 (17.5) 90 37 (15.4) 402 (16.7) 213 (16.8) 121 0.865 63 (26.0) 720 (29.9) 390 (30.6) 104 0.161 442 (15.7) 220 (20.0) 111 419 (15.0) 228 (20.9) 141 781 (27.8) 379 (34.6) 125 0.361 <0.0005 614 (26.0) 558 (35.5) 97 <0.0005 Pearson x2 p Total exposure to medicines (prescribed and/or OTC medicine use) No. (%)

Whole study population No. (% of the responses) <0.0005 2409 (60,2) 1591 (39.8) 32 330 (13.9) 341 (21.7) 83

Prescribed medicine use No. (%)

Pearson x2 p

Pearson x2 p

Copyright # 2009 John Wiley & Sons, Ltd.

<0.0005

Prescribed medicine use by the parent No prescribed medicine use One or more prescribed medicines Missing cases OTC medicine use by the parent No OTC medicine use One or more OTC medicines (vitamins excluded) Missing cases Education Junior high school or less (9 years) Senior high school / vocational school (1113 years) Polytechnic, college or university degree ( 15 years) Missing cases Working status Working or studying Home with children Not working (including persons on sick leave, retired, and unemployed) Missing cases Monthly net income of the household s 1999 20002999 30003999 400010000 Missing cases

meen-anttila k. ha
ET AL.

Pharmacoepidemiology and Drug Safety, 2010; 19: 400407 DOI: 10.1002/pds

factors associated with medicine use among nnish children

405

The most commonly used prescription medicines were medicines for obstructive airway diseases (6% of the children). Furthermore, corticosteroids for dermatological use (3%) and antibiotics for systemic use (3%) were commonly used. The analgesics including NSAIDs were the most commonly used OTC medicines (7% of the children), followed by cough and cold preparations (2%), and preparations for the treatment of wounds and ulcers (2%). Factors associated with the use of prescribed medicines The use of prescribed medicines was commonest among younger boys and children who had symptoms, illnesses diagnosed by a physician, and health status rated by the parent as worse than good (Table 1). Furthermore, two characteristics of the respondents were associated with the childs use of prescribed medicine: children with young mothers were more likely to use prescribed medicines than other children, and the mothers own use of prescribed and OTC medicines was associated with the childrens use of prescribed medicine. The results of the logistic regression analysis are presented in Table 2. The following variables were signicantly associated with the childs use of prescribed medicines, when adjusted with other variables: young age among boys, health status assessed worse than good by the parent, number of illnesses diagnosed by a physician, and use of prescribed medicine by the parent. Factors associated with the use of OTC medicines The childs characteristics that associated univariably with OTC medicine use were age, birth order, health status, number of symptoms experienced, and illnesses diagnosed by a physician (Table 1). Children with worse than good health status, symptoms and diagnosed illnesses were more likely to use OTC medicines than others. Furthermore, young children and second-borns were more likely to use OTC medicines than others. Several characteristics of the responding parent were associated with the childs use of OTC medicine: parents who gave OTC medicines to their children tended to be young, used prescribed and OTC medicines themselves, and were at home with their children (Table 1). According to the logistic regression analysis, the young age of the child, health status assessed worse than good, and the use of OTC medicine by the parent were signicantly associated with the childs use of
Copyright # 2009 John Wiley & Sons, Ltd.

OTC medicine, when adjusted with other variables (Table 2). However, children were less likely to use OTC medicines if they had some illness diagnosed by a physician. Factors associated with the total exposure to medicines (prescribed and/or OTC medicines) The characteristics that associated univariably with total exposure to medicines (prescribed and/or OTC medicines use) followed the same pattern as with the prescribed and OTC medicine use (Table 1). The use of any medicines was commonest among younger children and children with symptoms, children with illnesses diagnosed by a physician, with those health status rated by the parent as worse than good, young parent, and parent who also used prescribed or OTC medicines herself. As with OTC medicine use, the birth order of the child was associated with the total exposure to medicines by the child: second-borns were more likely to use medicines than other children. The following variables were signicantly associated with the childs use of any medicines (prescribed and/or OTC medicines use), when adjusted with other variables: young age, health status assessed worse than good by the parent, number of illnesses diagnosed by a physician, and use of prescribed or OTC medicine by the parent (Table 2). DISCUSSION Our study supports earlier ndings which suggest that poor health status3,6 and young age of the child8,9 as well as maternal medicine use7 is associated with a childs medicine use. According to our study, the association between medicine use and gender is not as straightforward as some earlier studies suggest when they claim that at young ages medicine use is more common among boys than girls.6,8,10,11 The estimation of ORs in the presence of interaction between age and gender revealed that age is a stronger predictor than gender in explaining the difference in medicine use. Among boys this association is clearer. The socioeconomic characteristics of the responding parent, such as education, working status, or the monthly net income of the household, did not have an inuence on childrens medicine use in this study, in contrast to previous ndings that mothers with more education and higher household incomes tend to give more OTC medicines to their children.3,7 The differences between the results of our present study and those of previous studies conducted in the United States may be due to cultural differences, and to the health care system in
Pharmacoepidemiology and Drug Safety, 2010; 19: 400407 DOI: 10.1002/pds

406

meen-anttila k. ha

ET AL.

Table 2. Odds ratios for factors associated with reported over-the-counter (OTC), prescribed medicine use, and total exposure to medicines among children aged under 12 years in Finland, using logistic regression multivariate analysis Variable Model I: Prescribed medicine use (N 3870, R2 0.471) OR (95%CI) Child Interaction: Sex & Age Female: age, years 02 36 711 Male: age, years 02 36 711 Health status (rated by the parent) Good Fairly good Moderate or poor Symptoms experienced, number 0 1 2 Illnesses diagnosed by a physician, no. 0 1 2 Parent Prescribed medicine use by the parent No prescribed medicine use One or more prescribed medicines OTC medicine use by the parent No OTC medicine use One or more OTC medicines (vitamins excluded)
y

p-value

Model II: OTC medicine use (N 3812, R20.170) OR (95%CI)

p-value

Model III: Total exposure to medicines (prescribed and/or OTC medicine use) (N 3611, R2 0.362) OR (95%CI)

p-value

0.290 1.00 1.18 (0.771.80) 0.70 (0.461.05) 1.00 0.75 (0.511.11) 0.51 (0.350.75) 1.00 1.73 (1.332.26) 2.48 (1.504.09)
y

0.282 1.00 0.81 (0.581.13) 0.55 (0.400.76) 1.00 0.58 (0.420.80) 0.53 (0.390.72) 1.00 1.84 (1.472.31) 2.35 (1.553.56)
y

0.179 1.00 0.82 (0.601.12) 0.57 (0.420.76) 1.00 0.55 (0.410.74) 0.44 (0.330.58) 1.00 2.12 (1.722.63) 2.72 (1.664.44)
y

<0.0001

<0.0001

<0.0001

<0.0001

<0.0001

<0.0001

<0.0001 1.00 16.72 (13.1421.27) 36.88 (24.0756.51) <0.0001 1.00 1.66 (1.332.06)

0.015 1.00 0.75 (0.580.98) 0.59 (0.380.92)

<0.0001 1.00 5.50 (4.406.88) 13.71 (8.5222.07) <0.0001 1.00 1.43 (1.211.70)

0.004 1.00 1.33 (1.091.61) 1.00 1.20 (1.011.44)

0.048

Variable not included in the model. Adjusted for symptoms experienced.

Finland, where free health care is available for all children regardless of the socio-economic status of the family. This study shows that a considerable proportion of children aged under 12 years were using prescribed medicines and OTC medicines, the prevalence being 17% for both. Comparison of our result with the prevalence rates in other studies is difcult since the outcome measure of medicine use differs considerably in different studies, e.g., 2 days,5,26 1 or 2 weeks,18,20 1 month,11,17 or even 3 months19 to 1 year.8,9,10,27 Thus, there is a need to agree on the indicators for medicine use in order to be able to compare the results from different countries. We strongly suggest that the present use of medicine should be the outcome measure when prevalence rates of drug use are studied, because of minimal recall bias.
Copyright # 2009 John Wiley & Sons, Ltd.

This study was based on a nationally representative sample of children aged under 12 years with a response rate of 67%. The responses represent Finnish children in age and gender, but their regional distribution differed from that of the target population. However, we did not nd any signicant differences in medicine use in different regions, so, the results may be generalized to represent medicine use among children aged under 12 years in Finland. We report the use of both prescribed and OTC medicine by children, which is a strength of this study since, typically, studies on medicine use among children report only prescription medicine use4,68 omitting OTC medication. Parents were asked to report the current use of prescribed medicine and the use of OTC medicine during the previous 2 days, which decreases the possibility of recall bias.
Pharmacoepidemiology and Drug Safety, 2010; 19: 400407 DOI: 10.1002/pds

factors associated with medicine use among nnish children KEY POINTS  A considerable proportion of children aged under

407

12 years use prescription and OTC medicines in Finland.  Young age and health-related factors, morbidity and health status, and parental medicine use predicted childrens medicine use.  The indicators for medicine use should be agreed upon so that results from different countries can be compared. We strongly suggest that present use of medicine should be the outcome measure when prevalence rates of drug use are studied, because of minimal recall bias and improved possibility to compare results between studies.

CONCLUSIONS A considerable proportion of children aged under 12 years use prescription and OTC medicines in Finland. Young age and health-related factors, morbidity and health status, and parental medicine use predicted childrens medicine use. However, our results do not support some previous ndings about differences in medicine use according to the socioeconomic characteristics of the parent. The nding of the association between parents and childrens medicine use is interesting and needs further investigation.

REFERENCES
1. Regulation (EC) No 1901/2006 of the European Parliament and of the Council on medicinal products for paediatric use and amending Regulation (EEC) No1768/92, Directive 2001/20/EC, Directive 2001/83/EC and Regulation (EC) No 726/2004 (12 December 2006). 2006; http:// eur-lex.europa.eu/LexUriServ/site/en/oj/2006/l_378/ l_37820061227en00010019.pdf [accessed 8 January 2009]. 2. Food and Drug Administration, Department of Health and Human Services. Regulations requiring manufacturers to assess the safety and effectiveness of new drugs and biological products in pediatric patients; nal rule. Federal Register, Rules and Regulations 1998; Vol. 63: 6663166672. 1998; http://www.fda.gov/CDER/guidance/pedrule.htm [accessed 8 January 2009]. 3. Kogan MD, Pappas G, Yu SM, Kotelchuck M. Over-the-counter medication use among US preschool-age children. JAMA 1994; 272(13): 10251030. 4. Eggen AE. Patterns of medicine use in general population (080 years). The inuence of age, gender, diseases and place of residence on drug use in Norway. Pharmacoepidemiol Drug Saf 1997; 6(3): 179187. kkinen U, Klaukka T, Klavus J, Lehtonen R, Aro S. Health 5. Arinen S, Ha and the use of health services in Finland. Main ndings of the Finnish Health Care Survey 1995/96 and changes from1987. Health Care 1998; 5: 5358.

6. Tobi H, Mejer WM, Tuinstra J, de Jong-van den Berg LTW. Socioeconomic differences in prescription and OTC drug use in Dutch adolescents. Pharm World Sci 2003; 25(5): 203206. 7. Maiman LA, Marshall H, Becker H, Katlic AW. Correlates of mothers use of medications for their children. Soc Sci Med 1986; 22(1): 41 51. 8. Thrane N, Srensen HT. A one-year population-based study of drug prescriptions for Danish children. Acta Paediatr 1999; 88(10): 1131 1136. 9. Madsen H, Andersen M, Hallas J. Drug prescribing among Danish children: a population-based study. Eur J Clin Pharmacol 2001; 57(2): 159165. 10. Straand J, Rokstad K, Heggedal U. Drug prescribing for children in general practice. A report from the Mre & Romsdal prescription study. Acta Paediatr 1998; 87(2): 218224. 11. Holstein BE, Hansen EH, Due P, Almarsdottir AB. Self-reported medicine use among 11- to 15-year-old girls and boys in Denmark 19881998. Scand J Public Health 2003; 31(5): 334341. 12. Cazzato T, Pandolni C, Campi R, Bonati M. Drug prescribing in outpatient children in Southern Italy. Eur J Clin Pharmacol 2001; 57(8): 611616. 13. Sanz E, Hernandez MA, Ratchina S, et al. Drug utilization in outpatient children. A comparison among Tenerife, Valencia, and Barcelona (Spain), Toulouse (France), Soa (Bulgaria), Bratislava (Slovakia) and Smolensk (Russia). Eur J Clin Pharmacol 2004; 60(2): 127 134. 14. Borgnolo G, Simon G, Francescutti C, Lattuada L, Zanier L. Antibiotic prescription in Italian children: a population-based study in Friuli Venezia Giulia, north-east Italy. Acta Paediatr 2001; 90(11): 1316 1320. 15. Gagliotti C, Morsillo F, Resi D, Milandri M, Moro ML. A three-year population-based study of antibiotic treatments for children. Acta Paediatr 2005; 94(10): 15021504. 16. Melander E, Merlo J, Lithman T, et al. Utilisation of antibiotics in young children: opposite relationships to adult educational levels in Danish and Swedish counties. Eur J Clin Pharmacol 2003; 59(4): 331 335. 17. Hansen EH, Holstein BE, Due P, Currie CE. International survey of self-reported medicine use among adolescents. Ann Pharmacother 2003; 37(3): 361366. 18. Dengler R, Roberts H. Adolescents use of prescribed and over-thecounter preparations. J Pub Health Med 1996; 18(4): 437442. 19. Chambers CT, Reid GJ, McGrath PJ, Finley A. Self-administration of over-the-counter medication for pain among adolescents. Arch Pediatr Adolesc Med 1997; 151(5): 449455. 20. Stoelben S, Krappweis J, Rossler G, Kirch W. Adolescents drug use and drug knowledge. Eur J Pediatr 2000; 159(8): 608614. 21. Wong ICK, Siang Chua S, Edmondson H. Childrens over-the-counter medicines pharmacoepidemiological (COPE) study. Int J Pharm Pract 2007; 15(1): 1722. 22. Klaukka T, Martikainen J, Kalimo E. Drug Utilization in Finland 19641987. Publications of the Social Insurance Institution: Helsinki, 1990; 71. 23. Tolonen H. Towards the High Quality of Population Health Surveys. Standardization and Quality Control. Publications of the National Public Health Institute, A27/2005: Edita Oy: Helsinki, Finland, 2006. 24. World Health Organization (WHO). The Anatomical Therapeutic Chemical Classication System with Dened Daily Doses (ATC/ DDD). World Health Organization, c2008. http://www.who.int/classications/atcddd/en/. [accessed 8 January 2009]. 25. Hosmer DW, Lemeshow S. Applied Logistic Regression. Wiley & Sons: New York, 1989; 6871. 26. Bush PJ, Iannotti RJ, Davidson FR. A longitudinal study of children and medicines. In Topics in Pharmaceutical Sciences, Bremer DD, Speiser P (eds). Elsevier: New York, NY, 1985; 391403. 27. Headley J, Northstone K. Medication administered to children from 0 to 7.5 years in the Avon Longitudinal Study of Parents and Children (ALSPAC). Eur J Clin Pharmacol 2007; 63(2): 189195.

Copyright # 2009 John Wiley & Sons, Ltd.

Pharmacoepidemiology and Drug Safety, 2010; 19: 400407 DOI: 10.1002/pds