This article has Open Peer Review reports available.
Use of antibiotics in rural and urban regions in the Netherlands: an observational drug utilization study
© de Jong et al.; licensee BioMed Central Ltd. 2014
Received: 10 February 2014
Accepted: 27 June 2014
Published: 3 July 2014
Large livestock farms might increase the infection risk for the nearby human population because of an increased risk for disease outbreaks and because antibiotic-resistant bacteria are more likely to be present. We hypothesized that populations residing in rural areas have more contact with cattle compared with populations in urban areas, and will use more antibiotics or more frequently require a new course of antibiotics.
Using data from the prescription database IADB.nl, we compared antibiotic use by patients living in rural areas to the use by patients living in urban areas. We also followed cohorts of antibiotic users and determined the patients who required a second antibiotic within 14 days after beginning the first antibiotic.
The yearly prevalence of antibiotic use was greater in rural areas compared with urban areas (2009: 23.6% versus 20.2% (p < 0.001), especially in the younger age groups. More adult patients residing in rural areas required a second course of antibiotic treatment within 14 days after starting the first treatment.
Individuals use more antibiotics, and adults more frequently require a second antibiotic prescription within 14 days, in rural areas compared with urban areas. Although the differences were small and the risks for the general rural population were not high, this difference should be investigated further.
Individuals residing in rural areas in the Netherlands used more antibiotics, especially young people < 45 years of age.
Adults in rural areas were more likely to require a second course of antibiotics within 14 days after starting the first course.
The findings of this study might result from greater exposure to resistant bacteria that originate from cattle farms.
Concerns about the health of individuals who work at or live near livestock farms have been increasing in the Netherlands. Over the last 20 years, the number of farms have decreased while the numbers of animals have increased. This increase has been especially significant for animals (e.g., poultry, pigs), which are kept in large numbers in concentrated growout facilities . Concentrations of large numbers of animals could contribute to an increased risk of outbreaks of diseases, such as avian and swine influenzas, foot-and-mouth-disease, and Q fever. Increased within-population transmission of bacteria, development of resistance via the increased use of veterinary antibiotics, and the concentration of fine dust near livestock farms are consequences of this intensive animal production.
In 2009, the nearby human population was affected by an outbreak of Q fever (Coxiella burnettii) in goats in the southern Netherlands . The discovery of a livestock-associated multi-drug resistant Staphylococcus aureus (MRSA) [3, 4] resulted in revision of the guidelines for individuals who work with cattle and become hospitalized . Highly resistant Escherichia coli strains from dairy farms have caused serious infections in humans in Belgium and the United States [6, 7].
In addition to these specific outbreaks, there may be general effects on the health of populations living in areas near livestock farms. There is little information about whether, compared with people in urban areas and others who are not typically in close contact with cattle, infections are more common, antibiotic use is greater, and antibiotics are more likely to be ineffective in populations that reside in rural areas.
Using a pharmacy prescription database, we investigated the differences in antibiotic use between urban and rural areas. We used the need for a new antibiotic prescription within 14 days after starting a course of treatment as a signal for therapeutic failure.
The IADB.nl database
Information on drug use was obtained from the IADB.nl database, which contains pharmacy-dispensing data from community pharmacies in the Netherlands. Dutch patients usually register at a single community pharmacy, so a single pharmacy provided an almost complete listing of each subject’s prescribed drugs . The pharmacy data included information on the name of the drug dispensed, the Anatomical Therapeutic Chemical (ATC) classification, a prescription date, the number of days the drug was prescribed, and the number of defined daily doses. The characteristics of these data were based on World Health Organization definitions . Over-the-counter drugs and in-hospital prescriptions were not included. The database contains the prescriptions of a population of 500,000 individuals.
The data in the IADB.nl database were strictly anonymous. No identification of individuals was possible. According to Dutch regulations, ethical approval was not required for the study .
Degree of urbanization
To perform this study, we required additional information on the environments in which the subjects resided. We used degree of urbanization (DU) information maintained by Statistics Netherlands (Centraal Bureau voor de Statistiek, Den Haag/Heerlen, the Netherlands) . In the Statistics Netherlands database, the DU value varied from 1 (extremely urbanized; 2,500 or more addresses per square kilometer) to 5 (not urbanized; fewer than 500 addresses per square kilometer). The DU value was available for every neighborhood within a community.
Every prescription in the IADB.nl database was linked to the first four numbers of each subject’s postal code (Postal Code-4). We selected all Postal Code-4 areas in which ≥60% of the population was included in the IADB.nl database.
The DU was linked to neighborhoods in the Statistics Netherlands database, not to postal code areas. There was only partial correspondence, and in most cases differences and overlaps were present. To resolve this problem, we determined the neighborhood with the lowest DU value (i.e., the most urbanized) in each Postal Code-4 area. The Postal Code-4 area was then assigned this DU value. To maintain anonymity of the data, the postal codes were deleted from the data set after the DU values were assigned.
To find an optimal effect, we selected the Postal Code-4 areas with a DU of 1 (n = 24) (‘urban’) and with a DU of 5 (n = 28) (‘rural’). The rural areas were identified as agricultural areas using the software application Google Earth (©2011 Google Inc., Mountain View, CA, USA). We selected all antibiotic drug prescriptions that were dispensed in these areas (ATC-code starting with J01) between 1998 and 2009.
Analysis of the number of prescriptions and types of antibiotics
The number of prescriptions per person residing in both the rural and the urban populations was determined for each year (1998 to 2009). We specifically examined the prescriptions for 1999, 2004, and 2009, and calculated the prevalence of the population using at least one course of antibiotics per year, stratified by age and sex. For these 3 years, we also calculated the average number of prescriptions per antibiotic user. We also determined the proportions of the prescriptions of the different types of antibiotics for the rural and urban populations.
Therapeutic failure was investigated by determining whether a second course of antibiotics was prescribed within 14 days of the start of a new course of antibiotics. A course of antibiotics was defined as new if the patient was not prescribed antibiotics during the preceding 30 days. To determine therapeutic failure, two cohorts of patients (one rural and one urban) who started an antibiotic course between 1 January 1999 and 31 December 2009 were selected. We compared the rural cohort with the urban cohort in an age-stratified analysis, and estimated the relative risks (RR) of needing a new prescription within 14 days after beginning the new treatment.
The Chi-square test was used to compare percentages. The Mann-Whitney U test was used to compare the numbers of prescriptions per person. Relative risks (RRs) with 95% confidence intervals (CI) were used to compare therapeutic failure. Microsoft® Office Excel 2010 (Microsoft® Corp., Redmond, WA, USA), SPSS 16.0 for Windows (IBM Corp., Armonk, NY, USA), and R version 2.13.0 (Free Software, Free Software Foundation, Boston, MA, USA) software were used for the data analysis. A p-value < 0.05 was considered to be statistically significant.
Characteristics of the prescriptions and of the study population, 2009
Rural (DU = 5)
Urban (DU = 1)
Number of prescriptions
Number of patients in the population (2009)
Age distribution of patients (2009)
71 years or older
Males in the population (2009)
Prevalence (%) of persons using antibiotics, stratified by 3 separate years and by age and sex
71 and older
Average number of prescriptions per year, per user of antibiotics, stratified by 3 separate years and by age and sex
71 and older
Proportions of the different antibiotic groups (number of prescriptions)
Rural (n = 183 947)
Urban (n = 626 584)
Number of patients in the cohort of subjects who were prescribed a second course of antibiotics within 14 days after the initial course
Nr new antibiotic
Total nr patients
Nr new antibiotic
Total nr patients
Relative risk (Confidence intervals)
In 14 days
In 14 days
We found that the use of antibiotics per year was higher in rural areas compared with urban areas (Figure 1). This difference mainly applied to individuals <46 years of age. No significant differences were found in the older age groups. In fact, in urban areas the antibiotic use was even greater (Tables 2 and 3). In rural areas, it was more customary to begin another course of antibiotics within 14 days after starting the initial antibiotic (Table 5), but this difference did not apply to the youngest group. There was also a difference in the type of antibiotic drugs prescribed. ‘Reserved’ antibiotics were prescribed more often in urban areas. This result did not support our hypothesis that these types of drugs would be more likely to be prescribed in rural areas (Table 4).
The empirical probability of medication use, especially antibiotics, is also associated with socioeconomic status [11, 12]. In many countries, the socioeconomic status in rural areas is lower compared with urban areas. In the Netherlands, populations with lower socioeconomic status tend to be present in the most urbanized areas . Therefore, socioeconomic differences cannot explain the higher number of antibiotics prescriptions in rural areas. The age distribution (Table 1) results indicate that there were more individuals aged 20 through 45 years living in urban areas. This result could be attributed to the fact that the largest city in the IADB-population is Groningen, which is a university town. A young population in the urban area could be a factor that contributed to the lower use of antibiotics, but the results in Tables 2 and 3 indicate that the difference between the two areas mostly occurs in the younger groups.
The use of cephalosporins and fluoroquinolones was more frequent in urban areas, which was not the result we expected. The close proximity of hospitals and the availability of specialized care could have contributed to this result.
No differences were found in children for the RR of needing another antibiotic, but the RR significantly exceeded 1 in the older age groups. Therapeutic failure of antibiotics seems to occur more often in older individuals living in rural areas, compared with those living in urban areas. The RR values were very close to 1, which implies that the measured effect was small. The effect might have been diluted by rural patients who did not have regular contact with cattle.
No other studies have been published for populations from the Netherlands that compare the use of antibiotics in urban and rural areas. Because urban and rural areas are characteristically very dissimilar in different countries, the results of studies performed in other countries are difficult to compare with the results of our study.
Two studies that investigated health symptoms associated with rural areas were performed in the Netherlands. Investigators that examined a population in the southern Netherlands compared the health of individuals residing near intensive livestock farms with that of individuals in other regions . They reported that there were no relevant differences. The results of this previous study suggest that compared with the reference group, fewer people near intensive livestock farms suffer from asthma, but respiratory infections in humans with asthma are more frequent. Pneumonia occurs more often near farms, which the authors related to the Q fever epidemic that occurred in that area. Children with eczema were also more likely to reside near farms. In the second study, which investigated rural-urban health differences, rural areas had a higher prevalence of infections. Working in close contact with animals was proposed as a cause for the difference . Similar to our results, both studies found only small differences between groups. The health risk for becoming infected that results from residing near livestock farms is probably not large in the general population in rural areas. The effect we found could have been diluted, because we had no specific information about individual contact with animals. These details should be investigated further in groups where more information is available about contact with animals and distance of a residence from farms. It is possible that for certain groups (e.g., individuals with asthma or the elderly) the risk could increase over time, as resistance against antibiotics disseminates in susceptible populations. It is important to note that in our study population children and younger people used more antibiotics, but the older age groups required a second course of antibiotic treatment more often.
Two advantages of our study were that we used objective pharmacy data and large population sizes. One study limitation was that some specific population data was missing. The database consisted of pharmacy prescription data, but did not include information on indications. The study was also limited by its assumption that inhabitants of rural areas were more likely to reside in closer proximity to cattle. We assumed that because most farmers reside on their farms, the chance that a patient worked with animals was greater compared with urban residents.
Compared with young people residing in urban areas, young people that resided in rural areas were more likely to use antibiotic drugs. Adults in rural areas more frequently required a second prescription of antibiotics within 14 days after beginning the first prescription. This result suggests that therapeutic failure due to antibiotic resistance may be a factor in this group. The differences between groups were quite small, probably because the effects were diluted. The risk for the general rural population is probably not significant. However, the factors that may have contributed to these results should be investigated in greater detail.
The data in the IADB.nl database are strictly anonymous. No derivation to any recognizable person is possible. Ethical approval was not required, according to Dutch regulations.
This study was supported by the pharmacy Apotheek Hardegarijp, where the lead author was allowed to use working time for research. There was no specific grant funding provided from any funding agency in the public, commercial, or non-profit sectors.
- van der Meulen H, de Bont K, Agricola H, van Horne P, Hoste R, van der Knijff A, Leenstra F, van der Meer R, de Smet A: Schaalvergroting in de land- en tuinbouw. Effecten bij veehouderij en glastuinbouw. 2011Google Scholar
- Smit LA, van der Sman-de BF, Opstal-van Winden AW, Hooiveld M, Beekhuizen J, Wouters IM, Yzermans J, Heederik D: Q fever and pneumonia in an area with a high livestock density: a large population-based study. PLoS One. 2012, 7 (6): e38843-View ArticlePubMedPubMed CentralGoogle Scholar
- van Cleef BA, Verkade EJ, Wulf MW, Buiting AG, Voss A, Huijsdens XW, van Pelt W, Mulders MN, Kluytmans JA: Prevalence of livestock-associated MRSA in communities with high pig-densities in The Netherlands. PLoS One. 2010, 5 (2): e9385-View ArticlePubMedPubMed CentralGoogle Scholar
- Mulders MN, Haenen AP, Geenen PL, Vesseur PC, Poldervaart ES, Bosch T, Huijsdens XW, Hengeveld PD, Dam-Deisz WD, Great EA, Mevius D, Voss A, Van De Giessen AW: Prevalence of livestock-associated MRSA in broiler flocks and risk factors for slaughterhouse personnel in The Netherlands. Epidemiol Infect. 2010, 138 (5): 743-755.View ArticlePubMedGoogle Scholar
- Rijks Instituut voor Volksgezondheid en Milieu (National Institute for Public Health and Environment): WIP-richtlijnen MRSA, ziekenhuis. 2012, Available at: http://www.rivm.nl/dsresource?objectid=rivmp:46603&type=org&disposition=inline. Accessed 09/18, 2012Google Scholar
- De Schrijver K, Buvens G, Posse B, Van den Branden D, Oosterlynck O, De Zutter L, Eilers K, Dierick K, Van Damme-Lombaerts R, Lauwers C, Jacobs R: Outbreak of verocytotoxin-producing E. coli O145 and O26 infections associated with the consumption of ice cream produced at a farm, Belgium, 2007. Euro Surveill. 2008, 13 (7):Google Scholar
- Guh A, Phan Q, Nelson R, Purviance K, Milardo E, Kinney S, Mshar P, Kasacek W, Cartter M: Outbreak of Escherichia coli O157 associated with raw milk, Connecticut, 2008. Clin Infect Dis. 2010, 51 (12): 1411-1417.View ArticlePubMedGoogle Scholar
- Visser ST, Schuiling-Veninga CC, Bos JH, de Jong-van den Berg LT, Postma MJ: The population-based prescription database IADB.nl: its development, usefulness in outcomes research and challenges. Expert Rev Pharmacoecon Outcomes Res. 2013, 13 (3): 285-292.View ArticlePubMedGoogle Scholar
- WHO Collaborating Centre for Drug Statistic Methodology: Guidelines for ATC classification and DDD assignment. 2012, Available at: http://www.whocc.no/atc_ddd_index/. Accessed 05/30, 2012Google Scholar
- Statistics Netherlands: Explanation of degree of urbanisation. Available at: http://www.cbs.nl/en-GB/menu/methoden/begrippen/default.htm?Languageswitch=on&ConceptID=658. Accessed 05-30, 2012
- Kozyrskyj AL, Dahl ME, Chateau DG, Mazowita GB, Klassen TP, Law BJ: Evidence-based prescribing of antibiotics for children: role of socioeconomic status and physician characteristics. CMAJ. 2004, 171 (2): 139-145.View ArticlePubMedPubMed CentralGoogle Scholar
- Mangrio E, Wremp A, Moghaddassi M, Merlo J, Bramhagen AC, Rosvall M: Antibiotic use among 8-month-old children in Malmo, Sweden–in relation to child characteristics and parental sociodemographic, psychosocial and lifestyle factors. BMC Pediatr. 2009, 9: 31-View ArticlePubMedPubMed CentralGoogle Scholar
- Mulder M: Sociaaleconomische status 2010. Volksgezondheid Toekomst Verkenning, Dutch National Atlas of Public Health. 2012, Available at: http://www.zorgatlas.nl/beinvloedende-factoren/sociale-omgeving/ses/sociaaleconomische-status/. Accessed 06/13, 2012Google Scholar
- Kroneman M, Verheij R, Tacken M, van der Zee J: Urban-rural health differences: primary care data and self reported data render different results. Health Place. 2010, 16 (5): 893-902.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2458/14/677/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.