BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.
When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.
The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.
BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay- per-view fees (http://bmjopen.bmj.com).
If you have any questions on BMJ Open’s open peer review process please email [email protected]
BMJ Open
Trends of prescribing antimicrobial drugs for urinary tract infections in primary care in the Netherlands
Journal: BMJ Open ManuscriptFor ID peerbmjopen-2018-027221 review only Article Type: Research
Date Submitted by the 12-Oct-2018 Author:
Complete List of Authors: Mulder, Marlies; Erasmus MC, Epidemiology Baan, Esme; Erasmus MC, Verbon, Annelies ; Erasmus MC, Microbiology and Infectious diseases Stricker, Bruno; Ersamus MC, Drug Safety Unit Verhamme, Katia; Erasmus Medical Center
Keywords: Urinary tract infections < UROLOGY, EPIDEMIOLOGY, PRIMARY CARE
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 28 BMJ Open
1 2 3 1 Trends of prescribing antimicrobial drugs for urinary tract infections in primary care in the 4 5 2 Netherlands 6 7 3 Running title: Prescribing of antimicrobial drugs for UTIs by GPs in the Netherlands 8 1,2 3 5 1,2,5 3,6 9 4 M. Mulder , E.J. Baan , A. Verbon , B.H.C. Stricker , K.M.C. Verhamme 10 11 5 1. Department of Epidemiology, Erasmus Medical Centre, PO Box 2040, 3000 CA Rotterdam, The 12 6 Netherlands 13 14 7 2. Health and Youth Care Inspectorate, Postbus 2518, 6401 DA Heerlen, The Netherlands 15 16 8 3. Department of MedicalFor Informatics, peer Erasmus review Medical Centre, PO only Box 2040, 3000 CA Rotterdam, The 17 9 Netherlands 18 19 20 10 4. Department of Medical Microbiology and Infectious Diseases, Erasmus Medical Centre, PO Box 2040, 21 11 3000 CA Rotterdam, The Netherlands 22 23 12 5. Department of Internal Medicine, Erasmus Medical Centre, PO Box 2040, 3000 CA Rotterdam, The 24 13 Netherlands 25 26 14 6. Department of Bioanalysis, Faculty of Pharmaceutical Sciences, University of Ghent, Belgium 27 28 29 15 30 31 16 Keywords: antibiotic stewardship, antimicrobial drug prescribing, urinary tract infections, primary care. 32 33 17 34 35 18 Address for Correspondence: 36 19 Professor Bruno H.Ch. Stricker 37 20 Department of Epidemiology 38 21 Erasmus MC - University Medical Center Rotterdam 39 22 P.O. Box 2040 40 23 3000 CA Rotterdam 41 42 24 The Netherlands 43 25 [email protected] 44 26 45 46 47 48 49 50 51 52 53 54 55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 28
1 2 3 27 Abstract 4 5 6 28 Objective: Urinary tract infections (UTIs) are an important reason to consult a General Practitioner (GP) 7 8 9 29 in the Netherlands. In this study, we describe antimicrobial drug prescribing patterns for UTIs and 10 11 30 investigate adherence of prescribers to Dutch primary care guidelines. 12 13 14 31 Methods: We conducted a population-based cohort study in IPCI, a Dutch GP database encompassing 15 16 32 approximately 2.5 millionFor patients. peer All antimicrobial review drug prescriptions only to patients ≥12 years in 1996- 17 18 33 2014 were extracted. The number of prescriptions and choice of drug type were investigated over 19 20 21 34 calendar time and in different age categories. The choice of antimicrobial drugs classes for UTIs and the 22 23 35 duration of nitrofurantoin use in women were compared to the Dutch primary care guidelines. 24 25 26 36 Results: The source population comprised 1,755,085 patients who received 2,019,335 antimicrobial 27 28 37 drug prescriptions; 401,655 (35.1%) prescriptions were for UTIs (45.2% in women; 12.6% in men). The 29 30 38 proportion of prescriptions for UTI within all prescriptions with an indication code increased from 5.2% 31 32 33 39 in 1996 to 14% in 2014 in men and from 28% in 1996 to 50% in 2014 for women. In men, UTIs were 34 35 40 most frequently treated with fluoroquinolones during the entire study period, whereas use of 36 37 41 fluoroquinolones was only advised as a first-choice in the latest guideline of 2013. In women, UTIs were 38 39 42 increasingly (p<0.05) treated with nitrofuran derivatives with a significant effect of the guideline of 2005 40 41 42 43 (p<0.05). Adherence to the advised duration of nitrofurantoin prescriptions in women has increased 43 44 44 since the guideline of 2005. 45 46 47 45 Conclusions: The antimicrobial drug prescribing for UTIs seem to have increased over time. Adherence 48 49 46 to the UTI guidelines increased with regard to choice and duration of antimicrobial drugs. We showed 50 51 47 that databases like IPCI in which prescription and indication are monitored, can be valuable antibiotic 52 53 54 48 stewardship tools. 55 56 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 28 BMJ Open
1 2 3 49 Strengths and limitations of this study 4 5 6 50 This study was conducted in a large database, including a substantial part (~15%) of the Dutch 7 8 51 population implying that the external validity of our study is high. 9 10 52 In the Netherlands, every person is assigned to a specific GP which limits selection bias. 11 12 53 Improvement of coding over calendar time might bias conclusions on total prescribing of 13 54 antimicrobial drugs. Therefore, we also studied the prescription data for other infections to be 14 15 55 able to study prescriptions of antimicrobial drugs for UTIs in comparison with prescriptions of 16 For peer review only 17 56 antimicrobial drugs for other infections 18 57 We analyzed the adherence of GPs to the Dutch guidelines on treatment of urinary tract 19 20 58 infections and show that database studies can be used as a tool for good antimicrobial 21 22 59 stewardship. 23 24 60 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 28
1 2 3 61 Introduction 4 5 6 62 Urinary tract infections (UTIs) are among the most common infections in humans. UTIs cause a 7
8 1 2 9 63 substantial burden of disease with major economic consequences. In women, UTIs are the most 10 11 64 common reason to consult a general practitioner (GP) in the Netherlands with 232 contacts per 1,000 12 13 65 women in 2014. In men, this number was substantially lower: 37 contacts per 1,000 men.3 14 15 16 66 In most cases UTIs areFor treated withpeer antimicrobial review drugs. The choice only of antimicrobial drug depends on the 17 18 67 antimicrobial sensitivity of pathogens in urinary cultures (if taken), severity of the symptoms and 19 20 21 68 potential co-morbidities of the individual. Based on the resistance patterns of E. coli in the Netherlands, 22 23 69 Dutch guidelines on the treatment of UTIs have changed several times in the last few decades. The latest 24 25 70 Dutch guideline for the treatment of UTIs in primary care was released in 2013 and recommends 26 27 71 nitrofurantoin as first-choice in all patients with cystitis, including men. In case of signs of tissue 28
29 4 30 72 invasion, ciprofloxacin is recommended as first-choice. Additionally, the recommended duration of 31 32 73 treatment has changed. The recommended duration of treatment with nitrofurantoin for cystitis in 33 34 74 women was 3 days in the guidelines of 1989 and 1999, whereas it has been extended to 5 days in the 35 36 75 guidelines of 2005 and 2013 (Table 1).4-7 37 38 39 76 Previous research shows that the use of antimicrobial drugs in the Netherlands in general has increased 40 41 42 77 since 2005 until approximately 2012, most prominently in elderly patients, but has decreased again 43 44 78 since then.8 9 Since UTIs are a major reason to consult a GP, use of antimicrobial drugs for UTIs 45 46 79 contribute significantly to total antimicrobial consumption. Therefore, antibiotic stewardship focusing 47 48 80 on (inappropriate) antimicrobial drug prescribing for UTIs is an important target in the fight against 49 50 51 81 antimicrobial resistance (AMR). This study aims to use a large GP database to investigate the adherence 52 53 82 of GPs to the Dutch UTI guidelines, by studying choice of antimicrobial drug and duration of 54 55 83 nitrofurantoin use over time. 56 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 28 BMJ Open
1 2 3 84 Table 1: Overview of the treatment of UTIs according to the Dutch guidelines 4 5 1989 1999 2005 2013 6 7 Women 8 Cystitis Trimethoprim Nitrofurantoin 1st choice: 1st choice: 9 (J01EA) or (J01XE) Nitrofurantoin Nitrofurantoin 10 Sulfamethizol 3 days or (J01XE) 5 days (J01XE) 5 days 11 (J01EB) or Trimethoprim 2nd choice: 2nd choice: 12 Nitrofurantoin (J01EA) Trimethoprim fosfomycin 13 (J01XE) 3 days (J01EA) (J01XX) 14 3rd choice: 3rd choice: 15 16 For peer review Fosfomycinonly Trimethoprim 17 (J01XX) (J01EA) 18 UTI with tissue invasion Amoxicillin 1st choice: 1st choice: 1st choice: 19 (J01CA) Amoxicillin- Amoxicillin- Ciprofloxacin 20 clavulanic acid clavulanic acid (J01MA) 21 (J01CR) (J01CR) 2nd choice: 22 2nd choice: 2nd choice: Amoxicillin- 23 Sulfamethoxazole- Sulfamethoxazole- clavulanic acid 24 25 trimethoprim trimethoprim (J01CR) rd 26 (J01EE) (J01EE) or 3 choice: 27 Fluoroquinolone Sulfamethoxazole- 28 (J01MA) trimethoprim 29 (J01EE) 30 Men 31 Cystitis UTI in men Nitrofurantoin 1st choice: 1st choice: 32 should always (J01XE) Nitrofurantoin Nitrofurantoin 33 34 be considered or (J01XE) (J01XE) nd nd 35 to be Trimethoprim 2 choice: 2 choice: 36 prostatitis (J01EA) Trimethoprim Trimethoprim 37 (J01EA) (J01EA) 38 UTI with tissue invasion Trimethoprim 1st choice: 1st choice: 1st choice: 39 (J01EA) or Amoxicillin- Amoxicillin- Ciprofloxacin 40 Amoxicillin clavulanic acid clavulanic acid (J01MA) 41 (J01CA) (J01CR) (J01CR) 2nd choice: 42 nd nd 43 2 choice: 2 choice: Amoxicillin- 44 sulfamethoxazole- Sulfamethoxazole- clavulanic acid 45 trimethoprim trimethoprim (J01CR) 46 (J01EE) (J01EE) or 3rd choice: 47 Fluorochinolone Sulfamethoxazole- 48 (J01MA) trimethoprim 49 (J01EE) 50 85 The tabel shows the treatment of UTIs according to the Dutch guidelines of 1989, 1999, 2005 and 2013.4-7 The recommended 51 86 duration of treatment with nitrofurantoin is also described. Please note that information from guidelines on the treatment of 52 87 pregnant women, and also the treatment of risk groups such as diabetes patients or abnormalities of the urinary tract differs is 53 88 not described. Furthermore, the treatment duration of men is in all case longer than in women. 54 55 56 89 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 28
1 2 3 90 Materials and methods 4 5 6 91 Study population 7 8 9 92 This study was conducted using data from the Integrated Primary Care Information database (IPCI), 10 11 12 93 which is a longitudinal observational dynamic database containing the records from more than 450 13 14 94 general practices throughout the Netherlands.10 Briefly, IPCI contains the complete electronic medical 15 16 95 records of ~2,500,000For patients, peercontaining amongst review others data on only diagnoses (coded and free text) and 17 18 96 prescriptions coded according to the ATC (anatomical therapeutical chemical) classification of the 19 20 11 21 97 WHO. The system complies with European Union guidelines on the use of data for medical research 22 23 98 and has been proven valid for pharmaco-epidemiological studies. More detailed information on IPCI has 24 25 99 been described elsewhere.10 The scientific and ethical advisory board of IPCI approved this study (nr. 26 27 100 11/2015). 28 29 30 101 The study cohort comprised patients aged 12 years and older with at least one year of valid database 31 32 st st 33 102 history in the IPCI database. The study period was from the 1 of January 1996 until the 1 of January 34 35 103 2015. Follow-up started from the following, whichever occurred last: start of study period, age of 12 36 37 104 years, and reaching a minimum of 12 months of database history. Follow-up ended when a patient left 38 39 105 the database, died or at the end of the study period, whichever occurred first. Gender and age at time of 40 41 42 106 prescription(s) were assessed for all patients. 43 44 45 107 Prescriptions 46 47 48 108 From the database, we selected all prescriptions for antimicrobial drugs prescribed during the study 49 50 109 period, using an automatic search on the ATC code “J01”, which is the ATC code for antimicrobial drugs. 51 52 110 All prescriptions of antimicrobial drugs were further categorized by ATC drug class e.g. J01AA 53 54 111 (tetracyclines). (Table S1) We analyzed all prescriptions of ATC class J01CA (penicillins with extended 55 56 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 28 BMJ Open
1 2 3 112 spectrum), J01CR (combinations of penicillins, incl. beta-lactamase inhibitors), J01EA (trimethoprim and 4 5 6 113 derivatives), J01EE (combinations of sulfonamides and trimethoprim, incl. derivatives), J01MA 7 8 114 (fluoroquinolones), J01XE (nitrofuran derivatives) and J01XX (other antimicrobials (mainly fosfomycin)). 9 10 115 In addition, the duration of all prescriptions of nitrofurantoin (J01XE01) (which in the Netherlands is only 11 12 116 prescribed for UTIs) was assessed. 13 14 15 117 Indication of use of antimicrobial drugs 16 For peer review only 17 18 118 Indication of use was assessed through an automatic search on disease specific codes. Antimicrobial 19 20 21 119 drug prescriptions were linked to the indication of use through a unique patient identifier linking a 22 23 120 prescription to a diagnosis using the International Classification of Primary Care (ICPC) codes. These ICPC 24 25 121 codes were categorized in the following indications: UTIs, skin infections, respiratory infections, ear 26 27 122 infections or other infections. All prescriptions without an ICPC code were assigned to the group: ‘no 28 29 30 123 code for indication of use’. Urethritis was included in the group of “other” infections and not in the UTI 31 32 124 group because of the distinct pathophysiology. 33 34 35 125 Analyses 36 37 38 126 The total number of prescriptions for UTIs, ‘other infections’ (including respiratory infections, skin 39 40 127 infections, ear infections etc.) and prescriptions without an indication code were calculated for the 41 42 128 complete study time and per year. Additionally, the proportion of UTI prescriptions per calendar year 43 44 45 129 was calculated with as denominator all antimicrobial drug prescriptions with an indication code. Also, 46 47 130 all nitrofurantoin prescriptions were analyzed per year. 48 49 50 131 Because of the dynamic nature of the study cohort, the annual prevalence of antimicrobial drug 51 52 132 prescriptions was calculated by dividing the total number of antimicrobial drug prescriptions by the total 53 54 133 number of person-years (PY). Furthermore, the annual number of users per calendar year was 55 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 28
1 2 3 134 calculated by dividing all users by the total number of PY in that specific calendar year. With regard to 4 5 6 135 the calculation of users, if an individual received prescriptions of more than one antimicrobial drug 7 8 136 classes in one calendar year, the individual contributed data to the different classes. However, if an 9 10 137 individual received 2 or more prescriptions of the same drug in one calendar year, the individual 11 12 138 contributed only once as a user of this specific drug class. The prevalence of prescriptions and the 13 14 139 prevalence of users were studied by age (12-17, 18-25, 26-35, 36-45, 46-55, 56-65, 66-75, 76-85 and ≥86 15 16 For peer review only 17 140 years age categories), gender and calendar year. 18 19 20 141 Finally, all nitrofurantoin prescriptions prescribed to women during the study period were selected and 21 22 142 duration of use was categorized in 3, 5, 7 days or in “other” (when the number of days of use was 23 24 143 unknown or different to 3, 5, or 7 days). The proportion of each category with as denominator the total 25 26 144 number of prescriptions of nitrofurantoin in women was calculated by calendar year. 27 28
29 12 30 145 ARIMA modeling using SPSS 24 was performed according to a previously proposed method. Calendar 31 32 146 time was first univariately added to the analysis in order to study the trends of the different 33 34 147 antimicrobial drug groups over time. Furthermore, a model including calendar time, intervention and 35 36 148 interaction was performed to investigate whether there was a significant difference in slope 37 38 39 149 (interaction) as an effect of the implementation of the guideline of 2005. 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 28 BMJ Open
1 2 3 150 Results 4 5 6 151 The study population comprised 1,755,085 patients of 12 years and older with a mean follow-up time of 7 8 9 152 3.31 year. 671,251 (38.2%) patients were prescribed at least one antimicrobial drug during the study 10 11 153 period; 271,772 (40.5%) men and 399,479 (59.5%) women. In total, they received 2,019,335 12 13 154 antimicrobial drug prescriptions (mean of 3.0 prescriptions per person) during the study period. Of these 14 15 155 prescriptions, 1,144,810 (56.7%) could be linked to an indication, namely 528,464 (46.2%) antimicrobial 16 For peer review only 17 18 156 drug prescriptions for airway infections; 401,655 (35.1%) for UTIs; 157,900 (13.8%) for skin infections; 19 20 157 29,984 (2.6%) for ear infections and 26,807 (2.3%) for “other” infections. 21 22 23 158 The subsequent sections describe the prescribing of antimicrobial drugs per calendar year and age 24 25 159 category for both men and women separately. To study the adherence of GPs to the national guidelines 26 27 160 as a tool in antibiotic stewardship, the results are compared to the recommendations according to the 28
29 4-7 30 161 national GP guidelines (Table 1). 31 32 33 162 Antimicrobial drug prescriptions for UTIs by calendar year 34 35 36 163 For all years combined and only considering those prescriptions with indication of use, prescriptions for 37 38 164 UTIs were 12.6% of the total prescriptions in men and 45.2% in women. The total number of 39 40 165 prescriptions per year for UTIs increased from 9 (men) and 88 (women) prescriptions per 1,000 PY in 41 42 166 1996 to respectively 19 and 153 prescriptions per 1,000 PY in 2014. However, especially at the end of 43 44 45 167 the study period, prescriptions without indication code decreased (Figure 1). To control for 46 47 168 improvement of coding over calendar time, we studied the total number of nitrofurantoin prescriptions, 48 49 169 which in the Netherlands is only prescribed for the indication UTI. In women more clearly than in men, 50 51 170 we see a flattening or even decrease of the number of non-coded nitrofurantoin prescriptions, whereas 52 53 54 171 the number of coded and total nitrofurantoin prescriptions increased (Figure S1). Moreover, we 55 56 172 calculated the proportion of AB prescriptions for UTI with as denominator all prescriptions with an 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 28
1 2 3 173 indication of use. In this analysis, the proportion of AB for UTIs increased from 5.2% in 1996 to 14% in 4 5 6 174 2014 for men and from 28% in 1996 to 50% in 2014 for women (Figure S2). 7 8 9 175 Fluoroquinolones (J01MA) are the most commonly prescribed antimicrobial drugs in men with UTIs with 10 11 176 no clear increase or decrease over time. Other frequently prescribed antimicrobial drugs were 12 13 177 combinations of sulfonamides and trimethoprim (J01EE) and combinations of penicillins, incl. beta- 14 15 178 lactamase inhibitors (J01CR), the last group increased significantly over time until 2013. Also, in men, the 16 For peer review only 17 18 179 number of nitrofuran derivatives prescriptions significantly increased from 0.4 prescriptions per 1,000 19 20 180 PY in 1996 to 6.2 prescriptions per 1,000 PY in 2014. In women, nitrofuran derivatives (J01XE) were 21 22 181 clearly the most frequently prescribed drugs since 1999, with a strong and significant increase in the last 23 24 182 years from 52 prescriptions per 1,000 PY in 2008 to 98 prescriptions per 1,000 PY in 2014. They 25 26 183 increasingly replaced the prescriptions of trimethoprim and derivatives (J01EA) and fluoroquinolones 27 28 29 184 (J01MA), which were also commonly prescribed in women. Also, the prescriptions of combinations of 30 31 185 penicillins, incl. beta-lactamase inhibitors (J01CR) and fosfomycin significantly increased, mainly in most 32 33 186 recent years (Figure 2). 34 35 36 187 Adherence to guidelines 37 38 39 188 The implementation of the Dutch guideline on the treatment of UTI of 2005 was associated in women 40 41 42 189 with a significant decrease of the slope of prescriptions of trimethoprim and derivatives (J01EA) and an 43 44 190 increase in nitrofuran derivatives (J01XE) prescriptions. Unfortunately, we did not have sufficient data 45 46 191 before 1999 and after 2013 to study the implementation of the other guidelines. For men, no significant 47 48 192 effects were found of the implementation of the guideline on the slope of any prescriptions. 49 50 51 193 Prescriptions of antimicrobial drugs for UTIs by age groups 52 53 54 55 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 28 BMJ Open
1 2 3 194 Both in men and women, an increase in total number of UTI prescriptions was observed by age category. 4 5 6 195 In women the number of prescriptions in all age groups is higher than in men, with nitrofuran 7 8 196 derivatives (J01XE) as the most prescribed antimicrobial drug in all age groups, followed by 9 10 197 fluoroquinolones (J01MA) and trimethoprim and derivatives (J01EA). In men, fluoroquinolones (J01MA) 11 12 198 are the most frequently prescribed antimicrobial drugs in most age groups, followed by nitrofuran 13 14 199 derivatives (J01XE) and combinations of penicillins, incl. beta-lactamase inhibitors (J01CR) (Figure 3). 15 16 For peer review only 17 18 200 Analyzing the total number of prescriptions for UTIs (J01 – antibacterials for systemic use) per age 19 20 201 category, we observed an increase in the last years for all age categories, but most prominent in the 21 22 202 higher age categories (Figure S3). 23 24 25 203 Changes in duration of nitrofurantoin prescriptions 26 27 28 204 In total, 215,531 prescriptions of nitrofurantoin with an ICPC code for UTIs were prescribed to women in 29 30 205 the study period. In line with a change in recommended duration from 3 to 5 days in 2005, the 31 32 33 206 proportion of prescriptions with a duration of 5 days increased strongly from 2005 onwards, while 34 35 207 proportion of prescriptions with a duration of 3 days became scarce. The proportion of prescriptions 36 37 208 with a duration of 7 days decreased from 33% in 1996 to 18% in 2014. The proportion of prescriptions 38 39 209 with a duration other than 3, 5 or 7 days (defined as “other duration”) remained stable over time, with a 40 41 42 210 proportion ranging between 6% and 10% (Figure 4). 43 44 45 211 46 47 48 49 50 51 52 53 54 55 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 28
1 2 3 212 Discussion 4 5 6 213 In this study we investigated the prescribing of antimicrobial drugs for UTIs in primary care during the 7
8 st st 9 214 study period January 1 1996 until January 1 2015. To assess compliance to Dutch guidelines on the 10 11 215 treatment of UTI in primary care, the choice of antimicrobial drug was compared to the 12 13 216 recommendations in these guidelines. In men, the number of nitrofurantoin prescriptions increased, but 14 15 217 fluoroquinolones were the most commonly prescribed antimicrobial drug in all years, although they are 16 For peer review only 17 18 218 only recommended as first-choice option in the most recent guideline. Women were most often 19 20 219 prescribed nitrofuran derivatives (J01XE), which increased over the years, followed by trimethoprim and 21 22 220 derivatives (J01EA) and fluoroquinolones (J01MA), which both decreased. Furthermore, compliance with 23 24 221 respect to the duration of nitrofurantoin use in women has increased since the guideline of 2005. 25 26 27 222 Nitrofurantoin was shown to be the most prescribed drugs for the treatment of UTIs, especially in 28 29 30 223 women, and increasingly in men. The Dutch guidelines of 1989 indicated nitrofurantoin as one of the 31 32 224 options for women, whereas it was suggested as option for both men and women in the guideline of 33 34 225 1999 and as first-choice drug for cystitis in the 2005 and 2013 guideline. This is reflected in our findings; 35 36 226 we observed in men and women an increase in nitrofurantoin prescriptions since 1999, with a steep and 37 38 39 227 significant increase since 2005. In this period nitrofurantoin replaced trimethoprim in women, which 40 41 228 significantly decreased after 2005. In the guidelines of 1999 and 2005, amoxicillin-clavulanic acid was 42 43 229 recommended for UTIs with tissue invasion, whereas amoxicillin (or trimethoprim in men) was 44 45 230 recommended for UTIs with tissue invasion in the guidelines prior to 1999. In men and women, a 46 47 231 significant increase of combinations of penicillins, incl. beta-lactamase inhibitors such as amoxicillin- 48 49 50 232 clavulanic acid, was shown during the study period. In men, a flattening was seen from 2012 on, just 51 52 233 before the guideline of 2013, which recommended ciprofloxacin as first-choice. 53 54 55 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 28 BMJ Open
1 2 3 234 A deviation from the UTI guidelines was observed with regard to the prescribing of fluoroquinolones. 4 5 6 235 Fluoroquinolones, in particular ciprofloxacin, were only advised as first-choice drug for UTIs with tissue 7 nd 8 236 invasion from 2013 on (in 2005 it was 2 choice and in 1989 use was discouraged), whereas in men 9 10 237 these drugs were the most frequently prescribed antimicrobial drug during the complete study period. 11 12 238 In women, we saw a significant decrease over time suggesting an increase of compliance of GPs over 13 14 239 time. We also studied the adherence to guidelines with regard to the duration of nitrofurantoin 15 16 For peer review only 17 240 prescribing in women. The recommended duration increased from 3 to 5 days in 2005, which is reflected 18 19 241 in our findings, indicating an increase of adherence to the guidelines by GPs, mainly since the guidelines 20 21 242 of 2005 and 2013.4-7 22 23 24 243 A study in another Dutch GP database investigated antimicrobial prescribing for several infections 25 26 244 during 2001 and showed that approximately 75% of the prescriptions for cystitis were first-choice drugs 27 28 13 29 245 (nitrofurantoin or trimethoprim) according to guidelines. The percentage of first-choice drug 30 31 246 prescriptions in our study in 2001 is comparable (approximately 67% for women). A smaller study from 32 33 247 2009 in 970 non-pregnant women of 11 years and older with cystitis determined that 66% were 34 35 248 prescribed nitrofurantoin, whereas we found a proportion of 57% in all women in 2009.14 Our 36 37 38 249 proportion is somewhat smaller, possibly because we did not distinguish between cystitis and 39 40 250 pyelonephritis, since a proportion of the prescriptions had an indication code for both. However, the 41 42 251 sample size of the other study is much smaller resulting in more uncertainty around the true prevalence 43 44 252 of antimicrobial drug use. 45 46 47 253 As for all observational studies, our study has strengths and limitations. The strength is the use of a large 48 49 50 254 population-based cohort from a database with detailed information on prescriptions and where 51 52 255 information is collected as part of routine clinical care reducing the risk of selection and information 53 54 256 bias. However, as part of new health quality standards, GPs are encouraged to improve disease coding, a 55 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 28
1 2 3 257 phenomenon also observed in other countries. This might in part explain the increase of total 4 5 6 258 prescriptions for UTIs which is underlined by our finding that the number of prescriptions for 7 8 259 nitrofurantoin without an indication code seemed to decrease, especially in women since approximately 9 10 260 2007. To control for this bias, we analyzed UTI prescriptions as a proportion of the total number of 11 12 261 antimicrobial drug prescriptions with an indication code (thus excluding prescriptions without a disease 13 14 262 code). Yet, we observed that the increase of prescriptions was larger for UTI’s than for other indications. 15 16 For peer review only 17 263 This increase in antimicrobial drug prescribing for the treatment of UTI is supported by other literature. 18 19 264 According to the NIVEL (Netherlands institute for health services research) registration, the number of 20 21 265 contacts with a GP for UTIs increased (especially in women) from 185 per 1,000 in 2012 to 295 per 1,000 22 23 266 in 2015.3 This was confirmed by another large Dutch GP database, showing an increase of GP visits for 24 25 15 16 26 267 UTIs and AB prescriptions for UTI in the period 2007-2010. This increase of use of antimicrobial drugs 27 28 268 for UTIs might partly be explained by ageing of the population. Indeed, we showed that antimicrobial 29 30 269 drugs were especially prescribed to patients in older age categories. Additionally, a Dutch study from 31 32 270 2012 also showed an increase of antimicrobial drug prescribing by age, and an increase of total 33 34 35 271 prescriptions of antimicrobial drugs to elderly: in 2000, 9% of patients above 80 years had at least one 36 37 272 prescription, which increased to 22% in 2009.8 However, ageing of the population might not solely 38 39 273 explain the increase, since we also showed an increase in prescriptions of antimicrobial drugs in younger 40 41 274 age groups. Possible explanations might be that more patients suffer from recurrent UTIs, that patients 42 43 44 275 more frequently visit their GP in case of a UTI or that GPs more easily prescribe antimicrobial drugs for 45 46 276 UTIs, but this has to be elucidated in future studies. 47 48 49 277 Reducing the use of antimicrobial drugs is important, since the drawback of treating infections with 50 51 278 antimicrobial drugs is the development of AMR.17 Patients suffering from infections caused by resistant 52 53 279 bacteria have an increased risk of a worse clinical outcome and death. 17 Moreover, AMR leads to an 54
55 18 19 56 280 increase in healthcare costs, which has been estimated to be 1.5 billion euros in EU every year. 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 28 BMJ Open
1 2 3 281 Misuse and overuse of antimicrobial drugs are the most important risk factors of antimicrobial 4
5 20 6 282 resistance worldwide. Therefore, antimicrobial stewardship programs have been developed, aiming to 7 8 283 optimize the prescribing of antimicrobial drugs, minimize misuse and overuse in order to minimize 9 10 284 AMR.21 Our study demonstrates that databases with antimicrobial drug prescriptions and indications are 11 12 285 useful for surveillance of antimicrobial drug prescriptions over time. In future, information gathered in 13 14 286 such databases should be provided to GP practices as part of antibiotic stewardship in primary care. 15 16 For peer review only 17 18 287 19 20 288 Funding 21 22 289 The authors have no funding to report. 23 24 290 25 26 291 Competing interests statement 27 28 29 292 KV works for a department who in the past received unconditional grants from: Pfizer/BI, Yamanouchi, 30 31 293 GSK, Novartis. None of which are related to the content of this paper. The other authors have nothing to 32 33 294 declare. 34 35 295 36 37 38 296 Author’s contributions 39 40 297 MM, AV, BS and KV designed the study. MM, EB and KV analyzed the data and interpreted these with 41 42 298 the aid of AV and BS. MM wrote the manuscript with critical input from all co-authors. 43 44 299 45 46 47 300 Data sharing statement 48 49 301 The aggregated data are available with the corresponding author upon reasonable request. 50 51 302 52 53 303 Patients and public involvement 54 55 56 304 No patients were involved in any stage of this study. 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 28
1 2 3 305 References 4 5 306 6 307 1. Foxman B. Epidemiology of urinary tract infections: incidence, morbidity, and economic costs. Am J 7 308 Med 2002;113 Suppl 1A:5S-13S. 8 309 2. Foxman B. The epidemiology of urinary tract infection. Nat Rev Urol 2010;7(12):653-60. 9 10 310 3. NIVEL Zorgregistraties eerste lijn 2015: Nederlands instituut voor onderzoek van de gezondheidszorg; 11 311 [updated 02-02-2016. Available from: https://www.nivel.nl/nl/NZR/huisarts-top-20-diagnoses- 12 312 bij-contacten-naar-geslacht. 13 313 4. van Pinxteren BK, B. J.; Geerlings, S. E.; Visser, H. S., Klinkhamer, S.; van der Weele, G. M.; Verduijn, M. 14 314 M.; Opstelten, W.; Burgers, J. S.; van Asselt, K. M. NHG-Standaard Urineweginfecties. Huisarts en 15 315 Wetenschap 2013;56(6):270-80. 16 316 5. Timmermans AEB, ForP. J. A. M.; peer Winkens, R. A.review G.; Arets, H.; Wiersma, only T. . NHG-Standaard 17 317 Urineweginfecties. Huisarts en Wetenschap 1999;42(13):613-22. 18 19 318 6. van Balen FAMB, P. J. A. M.; van Pienbroek, E.; Winkens, R.A.G. NHG-Standaard Urineweginfecties 20 319 Huisarts en Wetenschap 1989;32(11):439-43. 21 320 7. van Haaren KAMV, H. S.; van Vliet, S.; Timmermans, A. E.; Yadava, R.; Geerlings, S. E.; ter Riet, G.; van 22 321 Pinxteren, B. NHG-Standaard Urineweginfecties Huisarts en Wetenschap 2005;48(7):341-52. 23 322 8. Haeseker MB, Dukers-Muijrers NH, Hoebe CJ, et al. Trends in antibiotic prescribing in adults in Dutch 24 323 general practice. PLoS One 2012;7(12):e51860. 25 324 9. NethMap. Consumption of antimicrobial agents and antimicrobial resistance among medically 26 325 important bacteria in The Netherlands in 2015: SWAB/RIVM, 2016. 27 326 10. Vlug AE, van der Lei J, Mosseveld BM, et al. Postmarketing surveillance based on electronic patient 28 29 327 records: the IPCI project. Methods Inf Med 1999;38(4-5):339-44. 30 328 11. WCCfDSM. WHO Collaborating Centre for Drug Statistics Methodology; [Available from: 31 329 https://www.whocc.no/. 32 330 12. (EPOC) CEPaOoC. Interrupted time series (ITS) analyses: EPOC Resources for review authors; 2017 33 331 [Available from: http://epoc.cochrane.org/epoc-specific-resources-review-authors. 34 332 13. Ong DS, Kuyvenhoven MM, van Dijk L, et al. Antibiotics for respiratory, ear and urinary tract 35 333 disorders and consistency among GPs. J Antimicrob Chemother 2008;62(3):587-92. 36 334 14. den Heijer DD, GA. Maes, J. Stobberingh, EE. Antibiotica bij ongecompliceerde urineweginfecties: 37 38 335 geen toename van resistentie in de afgelopen 5 jaar. Nederlands Tijdschrift voor Geneeskunde 39 336 2011;155. 40 337 15. Willems CS, van den Broek D'Obrenan J, Numans ME, et al. Cystitis: antibiotic prescribing, 41 338 consultation, attitudes and opinions. Fam Pract 2014;31(2):149-55. 42 339 16. van den Broek d'Obrenan J, Verheij TJ, Numans ME, et al. Antibiotic use in Dutch primary care: 43 340 relation between diagnosis, consultation and treatment. J Antimicrob Chemother 44 341 2014;69(6):1701-7. 45 342 17. Bergman M, Nyberg ST, Huovinen P, et al. Association between antimicrobial consumption and 46 343 resistance in Escherichia coli. Antimicrob Agents Chemother 2009;53(3):912-7. 47 48 344 18. Cosgrove SE. The relationship between antimicrobial resistance and patient outcomes: mortality, 49 345 length of hospital stay, and health care costs. Clin Infect Dis 2006;42 Suppl 2:S82-9. 50 346 19. The bacterial challenge: time to react. A call to narrow the gap between multidrug-resistant bacteria 51 347 in the EU and the development of new antibacterial agents: ECDC/EMEA Joint technical report, 52 348 2009. 53 349 20. Ventola CL. The antibiotic resistance crisis: part 1: causes and threats. P T 2015;40(4):277-83. 54 350 21. Ventola CL. The antibiotic resistance crisis: part 2: management strategies and new agents. P T 55 351 2015;40(5):344-52. 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 28 BMJ Open
1 2 3 353 Figure legends 4 5 354 Figure 1: The number of users and prescriptions of antimicrobial drugs for UTIs, other infections and 6 7 355 infections without indication code in the period 1996 – 2014. Users show the number of patients who 8 356 were prescribed at least 1 antimicrobial drug of the drug group of interest in that specific year per 1,000 9 357 PY. Prescriptions show the total number of prescriptions who were prescribed in that specific year per 10 11 358 1,000 PY. Other infections included all users/prescriptions for infections with an indication code other 12 359 than UTIs, such as respiratory infections, ear infections and skin infections. 13 14 360 Figure 2: The number of users and prescriptions of antimicrobial drugs for urinary tract infections in 15 361 the period 1996 – 2014. Users men and women show the number of patients who were prescribed at 16 For peer review only 17 362 least 1 antimicrobial drug of the drug group of interest in that specific year per 1,000 PY. Prescriptions 18 363 men and women show the total number of prescriptions who were prescribed in that specific year per 19 364 1,000 PY. The figure only shows the most prevalent drug groups J01CA (penicillins with extended 20 365 spectrum), J01CR (combinations of penicillins, incl. beta-lactamase inhibitors), J01EA (trimethoprim and 21 22 366 derivatives), J01EE (combinations of sulfonamides and trimethoprim, incl. derivatives), J01MA 23 367 (fluoroquinolones), J01XE (nitrofuran derivatives), J01XX (other antimicrobial drugs). The prevalence of 24 368 other antimicrobial drug classes was analyzed but was because of low numbers not described in the 25 26 369 results section. Note that the scale of the y-axis differs between men and women. 27 28 370 Figure 3: The number of users and prescriptions of antimicrobial drugs for urinary tract infections per 29 371 age group. Users men and women show the number of patients who got prescribed at least 1 30 372 antimicrobial drug of the drug group in that specific age group during the total study time per 1,000 PY. 31 32 373 Prescriptions men and women show the total number of prescriptions that were prescribed in that 33 374 specific age group during the total study time per 1,000 PY. The figure only shows the most prevalent 34 375 drug groups J01CA (penicillins with extended spectrum), J01CR (combinations of penicillins, incl. beta- 35 376 lactamase inhibitors), J01EA (trimethoprim and derivatives), J01EE (combinations of sulfonamides and 36 37 377 trimethoprim, incl. derivatives), J01MA (fluoroquinolones), J01XE (nitrofuran derivatives), J01XX (other 38 378 antimicrobial drugs). Note that the scale of the y-axis differs between men and women. 39 40 379 Figure 4: The proportion of the different durations of nitrofurantoin prescriptions in women in the 41 380 period 1996 – 2014. The proportion of prescriptions of nitrofurantoin prescriptions (J01XXE01) with a 42 43 381 duration of 3, 5, 7 days or “other” duration (unequal to 3, 5, 7 days or unknown) with in de denominator 44 382 all these prescriptions. Until 2005 the duration recommended by the guideline was 3 days, whereas 45 383 since guideline of 2005 the advised duration was 5 days. 46 47 384 48 49 385 50 51 52 53 54 55 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 28
1 2 250 UTI men 3 Users 4 UTI women 5 200 other men 6 7 other women 8 no code men 9 150 no code women 10 11 12 100 13 14 15 Users (per1,000 PY) 16 50 17 18 For peer review only 19 0 20 21 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 22 Year 23 24 250 25 Prescriptions 26 27 200
28 29 30 31 150 32 33 34 100 35 36 37 38 Prescriptions (per1,000 PY) 50 39 40 41 42 0 43 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 44 Year 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 28 BMJ Open
1 2 20 160 3 Users men Users women 18 4 140 5 16 6 120 7 14
J01CA penicillins with
8 100 extended spectrum 9 12 10 10 80 11 J01CR combinations of 12 8 For peer60 review only penicillins, incl. 13 enzyme inhibitors 14 6 Users (per1,000 PY) Users (per1,000 PY) 40 15 J01EA trimethoprim 4 16 and derivatives 20 17 2 18 19 0 0 J01EE combinations of 20 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 sulfonamides and 21 Year Year trimethoprim 22 20 160 23 Prescriptions men Prescriptions women J01MA 24 18 140 fluoroquinolones 25
16 26 120 J01XE nitrofuran 27 14 derivatives 28 100 29 12 30 10 80 J01XX other 31 antimicrobial drugs 32 8 60 [fosfomycin] 33 6 34 40 J01 any antimicrobial Prescriptions (per1,000 PY) 35 Prescriptions (per1,000 PY) 4 drug 36 20 37 2 38 0 0 39 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 40 Year Year 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 20 of 28
1 2 160 450 3 400 4 140 5 350 6 120
7 300 J01CA penicillins with 8 100 extended spectrum 9 250 10 80 11 200 J01CR combinations of 12 60 penicillins, incl. For peer150 review only 13 PY) 1,000 (per Users enzyme inhibitors 14 Users (per1,000 PY) 40 100 15 J01EA trimethoprim and derivatives 16 20 50 17 18 0 0 19 12-17 18-25 26-35 36-45 46-55 56-65 66-75 76-85 86+ 12-17 18-25 26-35 36-45 46-55 56-65 66-75 76-85 86+ J01EE combinations of 20 YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS sulfonamides and 21 Age category Age category trimethoprim 22 160 450 J01MA 23 24 140 400 fluoroquinolones 25
350 26 120 J01XE nitrofuran 27 300 derivatives 28 100 29 250 80 30 J01XX other 200 31 antimicrobial drugs 32 60 150 [fosfomycin] 33 40 34 100 J01 any antimicrobial Prescriptions (per1,000 PY) 35 Prescriptions (per1,000 PY) drug 36 20 50 37 38 0 0 39 12-17 18-25 26-35 36-45 46-55 56-65 66-75 76-85 86+ 12-17 18-25 26-35 36-45 46-55 56-65 66-75 76-85 86+ YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS 40 Age category Age category 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 21 of 28 BMJ Open
1 2 100% 3 4 90% 5 6 80% 7 8 70% 9 10 60% 11 3 days 50% 12 For peer review only5 days 13 14 40% 7 days 15 other 16 30% 17 20% 18 Percentage of total prescriptions 19 20 10% 21 0% 22 23 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 24 Year 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 22 of 28
1 2 3 1 Supplementary table 1: Antimicrobial drug classes 4 2 5 ATC code Drug class 6 7 J01AA* Tetracyclines 8 J01CA Penicillins with extended spectrum 9 10 J01CE* Beta-lactamase sensitive penicillins 11 J01CF* Beta-lactamase resistant penicillins 12 13 J01CR Combinations of penicillins, incl. beta-lactamase inhibitors 14 J01DB* First-generations cephalosporins 15 J01DC* Second-generation cephalosporins 16 For peer review only 17 J01DD* Third-generation cephalosporins 18 J01EA Trimethoprim and derivatives 19 20 J01EB* Short-acting sulfonamides 21 J01EE Combinations of sulfonamides, and trimethoprim, incl. derivatives 22 23 J01FA* Macrolides 24 J01FF* Lincosamides 25 26 J01GB* Other aminoglycosides 27 J01MA Fluoroquinolones 28 29 J01MB* Other quinolones 30 J01XC* Steroid antibacterials 31 J01XE Nitrofuran derivatives 32 33 J01XX Other antibacterials 34 3 All antimicrobial drug classes that were prescribed for UTIs in the study population. The number of prescriptions of drugs with * 35 4 was very low and are not shown in the figures. 36 5 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 28 BMJ Open
1 2 3 16 4 5 Men 6 14 7
8 12 9 10 10 11 12 8 13 14 6 15 16 4
Prescriptions (per1,000 PY) For peer review only 17 18 19 2 20 21 0 22 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 23 6 Year 24 25 160 26 Women 27 140 28
29 120 30 31 100 32 33 34 80 35 36 60 37
38 Prescriptions per1,000 PY 40 39 40 41 20 42 43 0 44 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 45 7 Year 46 8 Supplementary figure 1: The number of prescriptions of nitrofurantoin with and without an indication code for urinary tract 47 9 infections for men and women. Note that the scale of the y-axis differs between men and women. 48 49 10 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 28
1 2 3 100 4 5 90 Users 6 80 7 8 70 9
60 10 11 50 12 40 code(%) 13 30 14 15 20 16 10 For peer review only
17 UTI userswithin users with indication
18 0 19 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 20 11 Year 21 100 22 Prescriptions 23 90 24 80
25 70 26 27 60 28 50 29 30 40 31 30 32 33 with indication code(%) 20 34 10 UTI UTI prescriptionswithin prescriptions
35 36 0 37 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Year 38 12 39 40 13 Supplementary figure 2: The percentage of users/prescriptions with an indication code for UTIs within all users/prescriptions 41 14 with an indication code (thus including indication codes for UTIs, airway infections, ear infections, skin infections etc.). 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 28 BMJ Open
1 2 3 120 450 4 Users men 5 400 Users women 100
6 350 7 80 300 8 9 250 60 10 200 11 40 150
12 perPY 1,000 ( Users For peer PY) 1,000 (per Users review only 100 13 20 14 50 15 0 0 16 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 17 Year Year 18 19 120 450 20 Prescriptions men Prescriptions women
400 21 100 22 350 23 80 300 24 250 25 60 26 200 27 40 150 28 29 100
Prescriptions PY) 1,000 (per Prescriptions 20 Prescriptions PY) 1,000 (per Prescriptions 30 50 31 0 32 0 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 33 Year Year 34 Supplementary figure 3: The number of users and prescriptions of antimicrobial drugs for urinary tract infections for different age groups per year. Users men and women show the 35 number of patients who got prescribed at least 1 antimicrobial drug in that specific age group per year per 1,000 PY. Prescriptions men and women show the total number of antimicrobial 36 drug prescriptions that were prescribed in that specific age group per year per 1,000 PY. Note that the scale of the y-axis differs between men and women. 37 38 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 26 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 28 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 28
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open
Trends of prescribing antimicrobial drugs for urinary tract infections in primary care in the Netherlands: a population- based cohort study ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2018-027221.R1
Article Type: Research
Date Submitted by the 13-Feb-2019 Author:
Complete List of Authors: Mulder, Marlies; Erasmus MC, Department of Epidemiology; Health and Youth Care Inspectorate Baan, Esme; Erasmus MC, Department of Medical Informatics Verbon, Annelies ; Erasmus MC, Medical Microbiology and Infectious diseases Stricker, Bruno; Ersamus MC, Department of Epidemiology; Health and Youth Care Inspectorate Verhamme, Katia; Erasmus Medical Center, Department of Medical Informatics; Universiteit Gent, Department of Bioanalysis
Primary Subject Pharmacology and therapeutics Heading:
Secondary Subject Heading: Infectious diseases, Epidemiology, General practice / Family practice
Urinary tract infections < UROLOGY, EPIDEMIOLOGY, PRIMARY CARE, Keywords: BACTERIOLOGY, Quality in health care < HEALTH SERVICES ADMINISTRATION & MANAGEMENT
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 34 BMJ Open
1 2 3 1 Trends of prescribing antimicrobial drugs for urinary tract infections in primary care in the 4 5 2 Netherlands: a population-based cohort study 6 7 3 Running title: Prescribing of antimicrobial drugs for UTIs by GPs in the Netherlands 8 1,2 3 4 1,2,5 3,6 9 4 M. Mulder , E.J. Baan , A. Verbon , B.H.C. Stricker , K.M.C. Verhamme 10 11 5 1. Department of Epidemiology, Erasmus Medical Centre, PO Box 2040, 3000 CA Rotterdam, The 12 6 Netherlands 13 14 7 2. Health and Youth Care Inspectorate, Postbus 2518, 6401 DA Heerlen, The Netherlands 15 16 8 3. Department of MedicalFor Informatics, peer Erasmus review Medical Centre, PO only Box 2040, 3000 CA Rotterdam, The 17 9 Netherlands 18 19 20 10 4. Department of Medical Microbiology and Infectious Diseases, Erasmus Medical Centre, PO Box 2040, 21 11 3000 CA Rotterdam, The Netherlands 22 23 12 5. Department of Internal Medicine, Erasmus Medical Centre, PO Box 2040, 3000 CA Rotterdam, The 24 13 Netherlands 25 26 14 6. Department of Bioanalysis, Faculty of Pharmaceutical Sciences, University of Ghent, Belgium 27 28 29 15 30 31 16 Keywords: antibiotic stewardship, antimicrobial drug prescribing, urinary tract infections, primary care. 32 33 17 34 35 18 Address for Correspondence: 36 19 Professor Bruno H.Ch. Stricker 37 20 Department of Epidemiology 38 21 Erasmus MC - University Medical Center Rotterdam 39 22 P.O. Box 2040 40 23 3000 CA Rotterdam 41 42 24 The Netherlands 43 25 [email protected] 44 26 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 34
1 2 3 27 Abstract 4 5 6 28 Objective: Urinary tract infections (UTIs) are an important reason to consult a General Practitioner (GP) 7 8 9 29 Here, we describe antimicrobial drug prescribing patterns for UTIs by GPs in relation to the Dutch 10 11 30 primary care guidelines. 12 13 14 31 Methods: We conducted a population-based cohort study in the Dutch Integrated Primary Care 15 16 32 Information databaseFor (IPCI), which peer encompasses review approximately 2.5 only million patients. All patients ≥12 17 18 33 years with at least one year of follow-up from 1996 to 2014 were extracted from the database. The 19 20 21 34 number of prescriptions and choice of drug type were investigated over time and in different age 22 23 35 categories. The choice of antimicrobial drugs classes for UTIs and the duration of nitrofurantoin use in 24 25 36 women were compared to the Dutch primary care guidelines of 1989, 1999, 2005 and 2013. 26 27 28 37 Results: The source population comprised 1,755,085 patients who received 2,019,335 antimicrobial 29 30 38 drug prescriptions; 401,655 (35.1%) prescriptions were for UTIs (45.2% in women; 12.6% in men). The 31 32 33 39 proportion of prescriptions for UTI within all prescriptions with an indication code increased from 5.2% 34 35 40 in 1996 to 14% in 2014 in men and from 28% in 1996 to 50% in 2014 for women. In men, UTIs were 36 37 41 most frequently treated with fluoroquinolones during the entire study period, whereas fluoroquinolones 38 39 42 were only advised as first-choice in the latest guideline of 2013. In women, UTIs were increasingly 40 41 42 43 (p<0.05) treated with nitrofuran derivatives with a statistically significant difference after 43 44 44 implementation of the guideline of 2005. Compliance to the advised duration of nitrofurantoin 45 46 45 prescriptions in women has increased since the guideline of 2005. 47 48 49 46 Conclusions: The antimicrobial drug prescribing for UTI seemed to have increased over time. Prescribing 50 51 47 in line to the UTI guidelines increased with regards to choice and duration of antimicrobial drugs. We 52 53 54 48 showed that databases like IPCI in which prescription and indication are monitored, can be valuable 55 56 49 antibiotic stewardship tools. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 34 BMJ Open
1 2 3 50 Strengths and limitations of this study 4 5 6 51 This study was conducted in a large database, including a substantial part (~15%) of the Dutch 7 8 52 population, implying that the external validity of our study is high. 9 10 53 Every individual in the Netherlands is assigned to a single GP and antimicrobial drugs are not 11 54 sold over the counter. This limits information bias, since we had access to all antimicrobial drug 12 13 55 prescriptions prescribed in primary care. 14 15 56 Improvement of coding discipline in more recent years might bias conclusions on total 16 For peer review only 17 57 prescribing of antimicrobial drugs, but we assume that this bias would be similar for all 18 58 indications. Furthermore, we compared the antimicrobial drug prescription data for UTIs with 19 20 59 antimicrobial drug prescription data for other infections. 21 22 60 We analyzed the choice of antimicrobial drugs prescribed by GPs in relation to the Dutch 23 61 guidelines on treatment of urinary tract infections and show that database studies can be used 24 25 62 as a tool for good antimicrobial stewardship. 26 27 28 63 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 34
1 2 3 64 Introduction 4 5 6 65 Urinary tract infections (UTIs) are among the most common infections in humans. UTIs cause a 7
8 1 2 9 66 substantial burden of disease with major economic consequences. In women, UTIs are the most 10 11 67 common reason to consult a general practitioner (GP) in the Netherlands with 232 contacts per 1,000 12 13 68 women in 2014. In men, this number was substantially lower: 37 contacts per 1,000 men.3 14 15 16 69 In most cases UTIs areFor treated withpeer antimicrobial review drugs. The choice only of antimicrobial drug depends on the 17 18 70 antimicrobial sensitivity of pathogens in urinary cultures (if taken), severity of the symptoms and 19 20 21 71 potential co-morbidities of the individual. Based on the resistance patterns of E. coli in the Netherlands, 22 23 72 Dutch guidelines on the treatment of UTI have changed several times in the last few decades. The latest 24 25 73 Dutch guideline for the treatment of urinary UTIs in primary care was released in 2013 and recommends 26 27 74 nitrofurantoin as first-choice in all patients with cystitis, including men. In case of signs of tissue 28
29 4 30 75 invasion, ciprofloxacin is recommended as first-choice. Additionally, the recommended duration of 31 32 76 treatment has changed. The recommended duration of treatment with nitrofurantoin for cystitis in 33 34 77 women was 3 days in the guidelines of 1989 and 1999, whereas it has been extended to 5 days in the 35 36 78 guidelines of 2005 and 2013 (Table 1).4-7 37 38 39 79 Table 1: Overview of the treatment of UTIs according to the Dutch guidelines for GPs 40 41 1989 1999 2005 2013 42 43 Women 44 Cystitis Trimethoprim Nitrofurantoin 1st choice: 1st choice: 45 (J01EA) or (J01XE) Nitrofurantoin Nitrofurantoin 46 Sulfamethizol 3 days or (J01XE) 5 days (J01XE) 5 days 47 (J01EB) or Trimethoprim 2nd choice: 2nd choice: 48 Nitrofurantoin (J01EA) Trimethoprim fosfomycin 49 (J01XE) 3 days (J01EA) (J01XX) 50 rd rd 51 3 choice: 3 choice: 52 Fosfomycin Trimethoprim 53 (J01XX) (J01EA) 54 UTI with tissue invasion Amoxicillin 1st choice: 1st choice: 1st choice: 55 (J01CA) Amoxicillin- Amoxicillin- Ciprofloxacin 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 34 BMJ Open
1 2 3 clavulanic acid clavulanic acid (J01MA) 4 (J01CR) (J01CR) 2nd choice: 5 nd nd 6 2 choice: 2 choice: Amoxicillin- 7 Sulfamethoxazole- Sulfamethoxazole- clavulanic acid 8 trimethoprim trimethoprim (J01CR) 9 (J01EE) (J01EE) or 3rd choice: 10 Fluoroquinolone Sulfamethoxazole- 11 (J01MA) trimethoprim 12 (J01EE) 13 Men 14 st st 15 Cystitis UTI in men Nitrofurantoin 1 choice: 1 choice: 16 Forshould peer always (J01XE) review Nitrofurantoinonly Nitrofurantoin 17 be considered or (J01XE) (J01XE) 18 to be Trimethoprim 2nd choice: 2nd choice: 19 prostatitis (J01EA) Trimethoprim Trimethoprim 20 (J01EA) (J01EA) 21 UTI with tissue invasion Trimethoprim 1st choice: 1st choice: 1st choice: 22 (J01EA) or Amoxicillin- Amoxicillin- Ciprofloxacin 23 24 Amoxicillin clavulanic acid clavulanic acid (J01MA) nd 25 (J01CA) (J01CR) (J01CR) 2 choice: 26 2nd choice: 2nd choice: Amoxicillin- 27 sulfamethoxazole- Sulfamethoxazole- clavulanic acid 28 trimethoprim trimethoprim (J01CR) 29 (J01EE) (J01EE) or 3rd choice: 30 Fluorochinolone Sulfamethoxazole- 31 (J01MA) trimethoprim 32 33 (J01EE) 4-7 34 80 The table shows the treatment of UTIs according to the Dutch guidelines of 1989, 1999, 2005 and 2013. The recommended 35 81 duration of treatment with nitrofurantoin is also described. Please note that information from guidelines on the treatment of 36 82 pregnant women, and also the treatment of risk groups such as diabetes patients or abnormalities of the urinary tract differs is 37 83 not described. Furthermore, the treatment duration of men is in all case longer than in women. 38 39 84 Previous research shows that the use of antimicrobial drugs in the Netherlands in general has increased 40 41 85 since 2005 until approximately 2012, most prominently in elderly patients, but has decreased again 42 43 86 since then.8 9 Since UTIs are a major reason to consult a GP, use of antimicrobial drugs for UTIs 44 45 87 contribute significantly to total antimicrobial consumption. Therefore, antibiotic stewardship focusing 46 47 48 88 on (inappropriate) antimicrobial drug prescribing for UTIs is an important target in the fight against 49 50 89 antimicrobial resistance (AMR). This study aims to study the choice of antimicrobial drugs prescribed for 51 52 90 UTIs by GPs and the duration of nitrofurantoin use over time in relation to the Dutch national guidelines 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 34
1 2 3 91 (of 1989, 1999, 2005 and 2013) for the treatment of UTIs in primary care using a large electronic primary 4 5 6 92 care database. 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 34 BMJ Open
1 2 3 93 Materials and methods 4 5 6 94 Data source 7 8 9 95 This study was conducted using data from the Integrated Primary Care Information database (IPCI), 10 11 12 96 which is a longitudinal observational dynamic database containing the records from more than 450 13 14 97 general practices throughout the Netherlands.10 Briefly, IPCI contains the complete electronic medical 15 16 98 records of ~2,500,000For patients, peercontaining amongst review others data on only diagnoses (coded and free text) and 17 18 99 prescriptions coded according to the ATC (anatomical therapeutical chemical) classification of the 19 20 11 21 100 WHO. The system complies with European Union guidelines on the use of data for medical research 22 23 101 and has been proven valid for pharmaco-epidemiological studies. More detailed information on IPCI has 24 25 102 been described elsewhere.10 The scientific and ethical advisory board of IPCI approved this study (nr. 26 27 103 11/2015). 28 29 30 104 Study population 31 32 33 105 The study cohort comprised patients aged 12 years and older with at least one year of valid database 34 35 st st 36 106 history in the IPCI database. The study period was from the 1 of January 1996 until the 1 of January 37 38 107 2015. Follow-up started from the following, whichever occurred last: start of study period, age of 12 39 40 108 years, and reaching a minimum of 12 months of database history. Follow-up ended when a patient left 41 42 109 the database, died or at the end of the study period, whichever occurred first. Gender, age at time of 43 44 45 110 prescription(s) and follow-up time (time of each patient in the database) were assessed for all patients. 46 47 48 111 Prescriptions 49 50 51 112 From the database, we selected all prescriptions for antimicrobial drugs prescribed during the study 52 53 113 period, using an automatic search on the ATC code “J01”, which is the ATC code for antimicrobial drugs. 54 55 114 All prescriptions of antimicrobial drugs were further categorized by ATC drug class e.g. J01AA 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 34
1 2 3 115 (tetracyclines). (Table S1) We analyzed all prescriptions of ATC class J01CA (penicillins with extended 4 5 6 116 spectrum), J01CR (combinations of penicillins, incl. beta-lactamase inhibitors), J01EA (trimethoprim and 7 8 117 derivatives), J01EE (combinations of sulfonamides and trimethoprim, incl. derivatives), J01MA 9 10 118 (fluoroquinolones), J01XE (nitrofuran derivatives) and J01XX (other antimicrobials (mainly fosfomycin)). 11 12 119 In addition, the duration of all prescriptions of nitrofurantoin (J01XE01) (which in the Netherlands is only 13 14 120 prescribed for UTIs) was assessed. 15 16 For peer review only 17 18 121 Indication of use of antimicrobial drugs 19 20 21 122 Indication of use was assessed through an automatic search on disease specific codes. Antimicrobial 22 23 123 drug prescriptions were linked to the indication of use through a unique patient identifier linking a 24 25 124 prescription to a diagnosis using the International Classification of Primary Care (ICPC-1) codes (version 26 27 125 5). These ICPC codes were categorized in the following indications: UTIs, skin infections, airway 28 29 30 126 infections, ear infections or other infections (Table S2). All prescriptions without an ICPC code were 31 32 127 assigned to the group: ‘no code for indication of use’. Urethritis was included in the group of “other” 33 34 128 infections and not in the UTI group because of the distinct pathophysiology. 35 36 37 129 Analyses 38 39 40 130 The total number of prescriptions for UTIs, ‘other infections’ (including respiratory infections, skin 41 42 131 infections, ear infections etc.) and prescriptions without an indication code were calculated for the 43 44 45 132 complete study time and per year. Additionally, the proportion of UTI prescriptions per calendar year 46 47 133 was calculated with as denominator all antimicrobial drug prescriptions with an indication code. Also, 48 49 134 all nitrofurantoin prescriptions were analyzed per year. 50 51 52 135 Since both the number of users, but also the total number of antimicrobial drug prescriptions are 53 54 136 interesting with regard to the study of antimicrobial resistance, we studied both. Because of the 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 34 BMJ Open
1 2 3 137 dynamic nature of the study cohort, the annual frequency of antimicrobial drug prescriptions was 4 5 6 138 calculated by dividing the total number of antimicrobial drug prescriptions by the total number of 7 8 139 person-years (PY). The annual number of users per calendar year was calculated by dividing all users by 9 10 140 the total number of PY in that specific year. With regard to the calculation of users, if an individual 11 12 141 received prescriptions of more than one antimicrobial drug classes in one year, the individual 13 14 142 contributed data to the different classes. However, if an individual received 2 or more prescriptions of 15 16 For peer review only 17 143 the same drug in one year, the individual contributed only once as a user of this specific drug class. The 18 19 144 frequency of prescriptions and users were studied by age (12-17, 18-25, 26-35, 36-45, 46-55, 56-65, 66- 20 21 145 75, 76-85 and ≥86 years age categories), gender and calendar year. The prescribing of antimicrobial 22 23 146 drugs by GPs was compared to the recommendations according to the national guidelines (Table 1).4-7 24 25 26 147 We intended to investigate if improved coding over time has influenced the results of the frequency of 27 28 29 148 prescribing antimicrobial drug prescriptions for UTIs in general, independent of the drug class. 30 31 149 Therefore, we also studied the frequency of nitrofurantoin prescriptions over time (since nitrofurantoin 32 33 150 was only prescribed for UTIs) and we studied the proportion of prescriptions for UTIs within total 34 35 151 prescriptions with an indication code (thus including all antimicrobial drug prescriptions with 36 37 38 152 information on indication of use). 39 40 41 153 Furthermore, all nitrofurantoin prescriptions prescribed to women during the study period were 42 43 154 selected and duration of use was categorized in 3, 5, 7 days or in “other” (when the number of days of 44 45 155 use was unknown or different to 3, 5, or 7 days). The proportion of each category with as denominator 46 47 156 the total number of prescriptions of nitrofurantoin in women was calculated by year. 48 49 50 51 157 Finally, we used a time series ARIMA model (in SPSS 24) in order to determine the effect of the 52 53 158 implementation of the guideline of 2005.12 The calendar year was first univariably added to the analysis 54 55 159 to study differences of antimicrobial drug prescribing over time. Next, we ran a model including calendar 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 34
1 2 3 160 year and the intervention (implementation of the guideline). When the p-value of the interaction term 4 5 6 161 was < 0.05, the difference in slope before and after the implementation of the guideline was significant, 7 8 162 implying a significant effect of the implementation of this guideline. 9 10 11 163 Patients and public involvement 12 13 164 No patients were involved in any stage of this study. 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 34 BMJ Open
1 2 3 165 Results 4 5 6 166 The study population comprised 1,755,085 patients of 12 years and older with a mean follow-up time of 7 8 9 167 3.31 year. 671,251 (38.2%) patients were prescribed at least one antimicrobial drug during the study 10 11 168 period; 271,772 (40.5%) men and 399,479 (59.5%) women. In total, they received 2,019,335 12 13 169 antimicrobial drug prescriptions (mean of 3.0 prescriptions per person) during the study period. Of these 14 15 170 prescriptions, 1,144,810 (56.7%) could be linked to an indication, namely 528,464 (46.2%) antimicrobial 16 For peer review only 17 18 171 drug prescriptions for airway infections; 401,655 (35.1%) for UTIs; 157,900 (13.8%) for skin infections; 19 20 172 29,984 (2.6%) for ear infections and 26,807 (2.3%) for “other” infections. 21 22 23 173 Antimicrobial drug prescriptions for UTIs by calendar year 24 25 26 174 For all years combined and only considering those prescriptions with indication of use, prescriptions for 27 28 175 UTIs were 12.6% of the total prescriptions in men and 45.2% in women. The total number of 29 30 176 prescriptions per year for UTIs increased from 9 (men) and 88 (women) prescriptions per 1,000 PY in 31 32 33 177 1996 to respectively 19 and 153 prescriptions per 1,000 PY in 2014. However, especially at the end of 34 35 178 the study period, prescriptions without indication code decreased (Figure 1). To control for 36 37 179 improvement of coding over time, we studied the total number of nitrofurantoin prescriptions, which in 38 39 180 the Netherlands is only prescribed for the indication UTI. In women more clearly than in men, we see a 40 41 42 181 flattening or even decrease of the number of non-coded nitrofurantoin prescriptions, whereas the 43 44 182 number of coded and total nitrofurantoin prescriptions increased (Figure S1). Moreover, we calculated 45 46 183 the proportion of antimicrobial drug prescriptions for UTI with as denominator all antimicrobial drug 47 48 184 prescriptions (for all indications) with an indication code. In this analysis, the proportion of antimicrobial 49 50 51 185 drugs for UTIs increased from 5.2% in 1996 to 14% in 2014 for men and from 28% in 1996 to 50% in 52 53 186 2014 for women (Figure S2). 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 34
1 2 3 187 Fluoroquinolones (J01MA) are the most commonly prescribed antimicrobial drugs in men with UTIs with 4 5 6 188 no clear increase or decrease over time. Other frequently prescribed antimicrobial drugs were 7 8 189 combinations of sulfonamides and trimethoprim (J01EE) and combinations of penicillins, incl. beta- 9 10 190 lactamase inhibitors (J01CR), the last group increased significantly over time until 2013. Also, in men, the 11 12 191 number of nitrofuran derivatives prescriptions significantly increased from 0.4 prescriptions per 1,000 13 14 192 PY in 1996 to 6.2 prescriptions per 1,000 PY in 2014. In women, nitrofuran derivatives (J01XE) were 15 16 For peer review only 17 193 clearly the most frequently prescribed drugs since 1999, with a strong and significant increase in the last 18 19 194 years from 52 prescriptions per 1,000 PY in 2008 to 98 prescriptions per 1,000 PY in 2014. They 20 21 195 increasingly replaced the prescriptions of trimethoprim and derivatives (J01EA) and fluoroquinolones 22 23 196 (J01MA), which were also commonly prescribed in women. Also, the prescriptions of combinations of 24 25 26 197 penicillins, incl. beta-lactamase inhibitors (J01CR) and fosfomycin significantly increased, mainly in most 27 28 198 recent years (Figure 2). 29 30 31 199 Choice of antimicrobial drug prescription in relation to the recommendations of the national primary 32 33 200 care guidelines 34 35 36 201 The implementation of the Dutch guideline on the treatment of UTI of 2005 was associated in women 37 38 39 202 with a significant decrease of the slope of prescriptions of trimethoprim and derivatives (J01EA) and an 40 41 203 increase in nitrofuran derivatives (J01XE) prescriptions. Unfortunately, we did not have sufficient data 42 43 204 before 1999 and after 2013 to study the implementation of the other guidelines. For men, no significant 44 45 205 effects were found of the implementation of the guidelines on the slope of any prescriptions. 46 47 48 206 Prescriptions of antimicrobial drugs for UTIs by age groups 49 50 51 207 Both in men and women, an increase in total number of UTI prescriptions was observed by age category. 52 53 54 208 In women the number of prescriptions in all age groups is higher than in men, with nitrofuran 55 56 209 derivatives (J01XE) as the most prescribed antimicrobial drug in all age groups, followed by 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 34 BMJ Open
1 2 3 210 fluoroquinolones (J01MA) and trimethoprim and derivatives (J01EA). In men, fluoroquinolones (J01MA) 4 5 6 211 were the most frequently prescribed antimicrobial drugs in most age groups, followed by nitrofuran 7 8 212 derivatives (J01XE) and combinations of penicillins, incl. beta-lactamase inhibitors (J01CR) (Figure 3). 9 10 11 213 Analyzing the total number of prescriptions for UTIs (J01 – antibacterials for systemic use) per age 12 13 214 category, we observed an increase in the last years for all age categories, but most prominent in the 14 15 215 higher age categories (Figure S3). 16 For peer review only 17 18 216 Changes in duration of nitrofurantoin prescriptions 19 20 21 217 In total, 215,531 prescriptions of nitrofurantoin with an ICPC code for UTIs were prescribed to women in 22 23 24 218 the study period. In line with a change in recommended duration from 3 to 5 days in 2005, the 25 26 219 proportion of prescriptions with a duration of 5 days increased strongly from 2005 onwards, while 27 28 220 proportion of prescriptions with a duration of 3 days became scarce. The proportion of prescriptions 29 30 221 with a duration of 7 days decreased from 33% in 1996 to 18% in 2014. The proportion of prescriptions 31 32 33 222 with a duration other than 3, 5 or 7 days (defined as “other duration”) remained stable over time, with a 34 35 223 proportion ranging between 6% and 10% (Figure 4). 36 37 38 224 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 34
1 2 3 225 Discussion 4 5 6 226 In this study, we investigated the prescribing of antimicrobial drugs for UTIs in primary care during the 7
8 st st 9 227 study period January 1 1996 until January 1 2015 and compare this with the recommendations in the 10 11 228 Dutch primary care guidelines. These guidelines are actively distributed by the Dutch College of General 12 13 229 Practitioners. 14 15 16 230 During the study period,For nitrofurantoin peer was shown review to be the most only prescribed drugs for the treatment of 17 18 231 UTIs, especially in women, and increasingly in men. The Dutch guidelines of 1989 indicated 19 20 21 232 nitrofurantoin as one of the options for women, whereas it was suggested as option for both men and 22 23 233 women in the guideline of 1999 and as first-choice drug for cystitis in the 2005 and 2013 guideline. This 24 25 234 is reflected in our findings; we observed in men and women an increase in nitrofurantoin prescriptions 26 27 235 since 1999, with a steep and significant increase since 2005. In this period nitrofurantoin replaced 28 29 30 236 trimethoprim in women, which significantly decreased after 2005. Furthermore, in the guidelines of 31 32 237 1999 and 2005, amoxicillin-clavulanic acid was recommended for UTIs with tissue invasion, whereas 33 34 238 amoxicillin (or trimethoprim in men) was recommended for UTIs with tissue invasion in the guidelines 35 36 239 prior to 1999. In men and women, a significant increase of combinations of penicillins, incl. beta- 37 38 39 240 lactamase inhibitors such as amoxicillin-clavulanic acid, was shown during the study period. In men, a 40 41 241 flattening was seen from 2012 on, just before the guideline of 2013, which recommended ciprofloxacin 42 43 242 as first-choice. 44 45 46 243 We also studied the compliance of GPs to guidelines with regard to the duration of 47 48 244 nitrofurantoin prescribing in women. The recommended duration increased from 3 to 5 days in 2005, 49 50 51 245 which is reflected in our findings, indicating an increase of adherence to the guidelines by GPs, mainly 52 53 246 since the guidelines of 2005 and 2013.4-7 A deviation from the UTI guidelines was observed with regard 54 55 247 to the prescribing of fluoroquinolones. Fluoroquinolones, in particular ciprofloxacin, were only advised 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 34 BMJ Open
1 2 3 248 as first-choice drug for UTIs with tissue invasion from 2013 on (in 2005 it was 2nd choice and in 1989 use 4 5 6 249 was discouraged), whereas in men these drugs were the most frequently described antimicrobial drug 7 8 250 during the complete study period. In women, we saw a significant decrease over time suggesting an 9 10 251 increase of compliance of GPs over time. 11 12 13 252 A study in another Dutch GP database investigated antimicrobial prescribing for several infections 14 15 253 during 2001 and showed that approximately 75% of the prescriptions for cystitis were first-choice drugs 16 For peer review only 17 13 18 254 (nitrofurantoin or trimethoprim) according to guidelines. The percentage of first-choice drug 19 20 255 prescriptions in our study in 2001 is comparable (approximately 67% for women). A smaller study from 21 22 256 2009 in 970 non-pregnant women of 11 years and older with cystitis determined that 66% were 23 24 257 prescribed nitrofurantoin, whereas we found a proportion of 57% in all women in 2009.14 Our 25 26 258 proportion is somewhat smaller, possibly because we did not distinguish between cystitis and 27 28 29 259 pyelonephritis, since a proportion of the prescriptions had an indication code for both. However, the 30 31 260 sample size of the other study is much smaller resulting in more uncertainty around the true prevalence 32 33 261 of antimicrobial drug use. 34 35 36 262 As part of new health quality standards, GPs are encouraged to improve disease coding, a phenomenon 37 38 39 263 also observed in other countries. This might in part explain the increase of total prescriptions for UTIs 40 41 264 which is underlined by our finding that the number of prescriptions for nitrofurantoin without an 42 43 265 indication code seemed to decrease, especially in women since approximately 2007. To control for this 44 45 266 bias, we analyzed UTI prescriptions as a proportion of the total number of antimicrobial drug 46 47 267 prescriptions with an indication code (thus excluding prescriptions without a disease code). Yet, we 48 49 50 268 observed that the increase of prescriptions was larger for UTI’s than for other indications. This increase 51 52 269 in antimicrobial drug prescribing for the treatment of UTI is supported by other literature. According to 53 54 270 the NIVEL (Netherlands institute for health services research) registration, the number of contacts with a 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 34
1 2 3 271 GP for UTIs increased (especially in women) from 185 per 1,000 in 2012 to 295 per 1,000 in 2015.3 This 4 5 6 272 was confirmed by another large Dutch GP database, showing an increase of GP visits for UTIs and 7 15 16 8 273 antimicrobial drug prescriptions for UTI in the period 2007-2010. This increase of use of antimicrobial 9 10 274 drugs for UTIs might partly be explained by ageing of the population. Indeed, we showed that 11 12 275 antimicrobial drugs were especially prescribed to patients in older age categories. Additionally, a Dutch 13 14 276 study from 2012 also showed an increase of antimicrobial drug prescribing by age, and an increase of 15 16 For peer review only 17 277 total prescriptions of antimicrobial drugs to elderly: in 2000, 9% of patients above 80 years had at least 18 19 278 one prescription, which increased to 22% in 2009.8 However, ageing of the population cannot solely 20 21 279 explain the increase, since we also showed an increase in prescriptions of antimicrobial drugs in younger 22 23 280 age groups. Possible explanations might be that more patients suffer from recurrent UTIs, that patients 24 25 26 281 more frequently visit their GP in case of a UTI or that GPs more easily prescribe antimicrobial drugs for 27 28 282 UTIs, but this has to be elucidated in future studies. 29 30 31 283 One of the strengths of this study is that we use a large population-based cohort from a database with 32 33 284 detailed information on prescriptions and where information is collected as part of routine clinical care 34 35 285 reducing the risk of selection and information bias. We used these data to show that they can be of 36 37 38 286 importance in studying antibiotic stewardship. Unfortunately, because of confidentiality, we could not 39 40 287 study differences in prescribing patterns by region nor by GP characteristics such as age and/or gender. 41 42 288 Additionally, although we found that antimicrobial drugs were increasingly prescribed according to the 43 44 289 guidelines, the study design did not allow us to investigate the reasons behind deviations from the 45 46 47 290 guidelines. This is of interest and should be studied further, eventually by qualitative prospective 48 49 291 research to distinguish between forced deviations caused by resistant pathogens and other deviations, 50 51 292 which should be diminished to promote good antimicrobial stewardship. 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 34 BMJ Open
1 2 3 293 Reducing the use of antimicrobial drugs is important, since the drawback of treating infections with 4
5 17 6 294 antimicrobial drugs is the development of AMR. Patients suffering from infections caused by resistant 7 17 8 295 bacteria have an increased risk of a worse clinical outcome and death. Moreover, AMR leads to an 9 10 296 increase in healthcare costs, which has been estimated to be 1.5 billion euros in EU every year.18 19 11 12 297 Misuse and overuse of antimicrobial drugs are the most important risk factors of antimicrobial 13 14 298 resistance worldwide.20 Therefore, antimicrobial stewardship programs have been developed, aiming to 15 16 For peer review only 17 299 optimize the prescribing of antimicrobial drugs, minimize misuse and overuse in order to minimize 18 19 300 AMR.21 Our study demonstrates that databases with antimicrobial drug prescriptions and indications are 20 21 301 useful for surveillance of antimicrobial drug prescriptions over time. In future, information gathered in 22 23 302 such databases should be provided to GP practices as part of antibiotic stewardship in primary care. 24 25 26 303 27 28 29 304 Funding 30 31 305 The authors have no funding to report. 32 33 306 34 35 307 Competing interests statement 36 37 38 308 KV works for a department who in the past received unconditional grants from: Pfizer/BI, Yamanouchi, 39 40 309 GSK, Novartis. None of which are related to the content of this paper. The other authors have nothing to 41 42 310 declare. 43 44 311 45 46 47 312 Author’s contributions 48 49 313 MM, AV, BS and KV designed the study. MM, EB and KV analyzed the data and interpreted these with 50 51 314 the aid of AV and BS. MM wrote the manuscript with critical input from all co-authors. 52 53 315 54 55 56 316 Data sharing statement 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 34
1 2 3 317 The aggregated data are available with the corresponding author upon reasonable request. 4 5 6 318 7 8 319 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 34 BMJ Open
1 2 3 320 References 4 5 321 6 322 1. Foxman B. Epidemiology of urinary tract infections: incidence, morbidity, and economic costs. Am J 7 323 Med 2002;113 Suppl 1A:5S-13S. 8 324 2. Foxman B. The epidemiology of urinary tract infection. Nat Rev Urol 2010;7(12):653-60. 9 10 325 3. NIVEL Zorgregistraties eerste lijn 2015: Nederlands instituut voor onderzoek van de gezondheidszorg; 11 326 [updated 02-02-2016. Available from: https://www.nivel.nl/nl/NZR/huisarts-top-20-diagnoses- 12 327 bij-contacten-naar-geslacht]. 13 328 4. van Pinxteren BK, B. J.; Geerlings, S. E.; Visser, H. S., Klinkhamer, S.; van der Weele, G. M.; Verduijn, M. 14 329 M.; Opstelten, W.; Burgers, J. S.; van Asselt, K. M. NHG-Standaard Urineweginfecties. Huisarts en 15 330 Wetenschap 2013;56(6):270-80. 16 331 5. Timmermans AE, BaselierFor PJAM, peer Winkens RAG,review et al. NHG-Standaard only Urineweginfecties. Huisarts en 17 332 Wetenschap 1999;42(13):613-22. 18 19 333 6. van Balen FAM, Baselier PJAM, van Pienbroek E, et al. NHG-Standaard Urineweginfecties Huisarts en 20 334 Wetenschap 1989;32(11):439-43. 21 335 7. van Haaren KAMV, H. S.; van Vliet, S.; Timmermans, A. E.; Yadava, R.; Geerlings, S. E.; ter Riet, G.; van 22 336 Pinxteren, B. NHG-Standaard Urineweginfecties Huisarts en Wetenschap 2005;48(7):341-52. 23 337 8. Haeseker MB, Dukers-Muijrers NH, Hoebe CJ, et al. Trends in antibiotic prescribing in adults in Dutch 24 338 general practice. PLoS One 2012;7(12):e51860. 25 339 9. NethMap. Consumption of antimicrobial agents and antimicrobial resistance among medically 26 340 important bacteria in The Netherlands in 2015: SWAB/RIVM, 2016. 27 341 10. Vlug AE, van der Lei J, Mosseveld BM, et al. Postmarketing surveillance based on electronic patient 28 29 342 records: the IPCI project. Methods Inf Med 1999;38(4-5):339-44. 30 343 11. WHO Collaborating Centre for Drug Statistics Methodology. [Available from: 31 344 https://www.whocc.no/]. 32 345 12. Cochrane Effective Practice and Organisation of Care. Interrupted time series (ITS) analyses: EPOC 33 346 Resources for review authors; 2017 [Available from: http://epoc.cochrane.org/epoc-specific- 34 347 resources-review-authors]. 35 348 13. Ong DS, Kuyvenhoven MM, van Dijk L, et al. Antibiotics for respiratory, ear and urinary tract 36 349 disorders and consistency among GPs. J Antimicrob Chemother 2008;62(3):587-92. 37 38 350 14. den Heijer DD, GA. Maes, J. Stobberingh, EE. Antibiotica bij ongecompliceerde urineweginfecties: 39 351 geen toename van resistentie in de afgelopen 5 jaar. Nederlands Tijdschrift voor Geneeskunde 40 352 2011;155. 41 353 15. Willems CS, van den Broek D'Obrenan J, Numans ME, et al. Cystitis: antibiotic prescribing, 42 354 consultation, attitudes and opinions. Fam Pract 2014;31(2):149-55. 43 355 16. van den Broek d'Obrenan J, Verheij TJ, Numans ME, et al. Antibiotic use in Dutch primary care: 44 356 relation between diagnosis, consultation and treatment. J Antimicrob Chemother 45 357 2014;69(6):1701-7. 46 358 17. Bergman M, Nyberg ST, Huovinen P, et al. Association between antimicrobial consumption and 47 48 359 resistance in Escherichia coli. Antimicrob Agents Chemother 2009;53(3):912-7. 49 360 18. Cosgrove SE. The relationship between antimicrobial resistance and patient outcomes: mortality, 50 361 length of hospital stay, and health care costs. Clin Infect Dis 2006;42 Suppl 2:S82-9. 51 362 19. The bacterial challenge: time to react. A call to narrow the gap between multidrug-resistant bacteria 52 363 in the EU and the development of new antibacterial agents: ECDC/EMEA Joint technical report, 53 364 2009. 54 365 20. Ventola CL. The antibiotic resistance crisis: part 1: causes and threats. P T 2015;40(4):277-83. 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 34
1 2 3 366 21. Ventola CL. The antibiotic resistance crisis: part 2: management strategies and new agents. P T 4 367 2015;40(5):344-52. 5 6 368 7 8 369 9 370 Figure 1: The number of users and prescriptions of antimicrobial drugs for UTIs, other infections and infections without 10 371 indication code in the period 1996 – 2014. 11 12 372 13 14 373 Figure 2: The number of users and prescriptions of antimicrobial drugs for urinary tract infections in the period 1996 – 2014. 15 374 Note that the scale of the y-axis differs between men and women. 16 For peer review only 17 375 18 19 376 Figure 3: The number of users and prescriptions of antimicrobial drugs for urinary tract infections per age group. Note that 20 377 the scale of the y-axis differs between men and women. 21 378 22 379 Figure 4: The proportion of the different durations of nitrofurantoin prescriptions in women in the period 1996 – 2014. The 23 380 proportion of prescriptions of nitrofurantoin prescriptions (J01XXE01) with a duration of 3, 5, 7 days or “other” duration 24 381 (unequal to 3, 5, 7 days or unknown) with in de denominator all these prescriptions. Until 2005 the duration recommended by 25 382 the guideline was 3 days, whereas since guideline of 2005 the advised duration was 5 days. 26 383 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 34 BMJ Open
1 2 3 384 4 5 6 7 8 9 10 11 12 For peer review only 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 22 of 34
1 2 3 250 4 Users 5 6 200 7 8 9 150 10 11 12 100 13
14 Users (per 1,000 PY) 50 15 16 17 0 18 1996 1998 For2000 peer2002 review2004 2006 2008only2010 2012 2014 19 Year
20 250 21 Prescriptions 22 23 200 24 25 26 150 27 28 29 100 30 31 50 32 Prescriptions (per 1,000 PY) 33 34 0 35 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 36 Year 37 Figure 1: The number of users and prescriptions of antimicrobial drugs for UTIs, other infections and infections without 38 indication code in the period 1996 – 2014. 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 34 BMJ Open
1 2 20 160 3 Users men Users women 18 4 140 5 16 6 120 14 7 J01CA penicillins with 8 12 100 extended spectrum 9 10 10 80 11 J01CR combinations 12 8 For peer60 review only of penicillins, incl. 13 6 enzyme inhibitors Users (per 1,000 PY) 1,000 (per Users 14 PY) 1,000 (per Users 40 15 4 J01EA trimethoprim 16 20 and derivatives 17 2 18 0 0 19 J01EE combinations 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 of sulfonamides and 20 Year Year trimethoprim 21 20 160 22 Prescriptionsmen Prescriptions women J01MA 23 18 140 fluoroquinolones 24 25 16 120 26 14 J01XE nitrofuran 27 100 derivatives 28 12 29 10 80 30 J01XX other 31 antimicrobial drugs 8 60 32 [fosfomycin] 33 6 40 J01 any antimicrobial 34 Prescriptions (per 1,000 PY) Prescriptions (per 1,000 PY) 4 drug 35 36 2 20 37 38 0 0 39 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Year Year 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 24 of 34
1 2 160 450 3 4 140 400 5 350 6 120
7 300 J01CA penicillins with 8 100 extended spectrum 9 250 10 80 11 200 J01CR combinations 12 60 of penicillins, incl. For peer150 review only 13 PY) 1,000 (per Users enzyme inhibitors Users (per 1,000 PY) 14 40 100 15 J01EA trimethoprim and derivatives 16 20 50 17 18 0 0 19 12-17 18-25 26-35 36-45 46-55 56-65 66-75 76-85 86+ 12-17 18-25 26-35 36-45 46-55 56-65 66-75 76-85 86+ J01EE combinations 20 YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS of sulfonamides and 21 Age category Age category trimethoprim 22 160 450 23 J01MA 24 140 400 fluoroquinolones 25
350 26 120 J01XE nitrofuran 27 300 derivatives 28 100 29 250 80 30 J01XX other 200 31 antimicrobial drugs 32 60 150 [fosfomycin] 33 40 34 100 J01 any antimicrobial Prescriptions (per 1,000 PY) 35 Prescriptions (per 1,000 PY) drug 36 20 50 37 38 0 0 39 12-17 18-25 26-35 36-45 46-55 56-65 66-75 76-85 86+ 12-17 18-25 26-35 36-45 46-55 56-65 66-75 76-85 86+ 40 YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS YEARS Age category Age category 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 25 of 34 BMJ Open
1 2 3 4 100% 5 6 90% 7 8 80% 9 10 70% 11 12 60% For peer review only 3 days 13 50% 14 5 days 15 16 40% 7 days 17 18 30% other 19 20 Percentage of total prescriptions 20% 21 22 10% 23 24 0% 25 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 26 Year 27 28 Figure 4: The proportion of the different durations of nitrofurantoin prescriptions in women in the period 1996 – 2014. The proportion of prescriptions of nitrofurantoin prescriptions (J01XXE01) with a duration of 3, 5, 7 days or “other” duration (unequal to 3, 5, 7 days or unknown) with in de denominator all these prescriptions. Until 2005 the duration recommended by 29 the guideline was 3 days, whereas since guideline of 2005 the advised duration was 5 days. 30
31 32 33 34 35 36 37 38 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 26 of 34
1 2 3 1 Supplementary table 1: ATC codes of antimicrobial drugs prescribed by GPs to treat urinary tract 4 2 infections 5 3 6 ATC code Drug class 7 8 J01AA* Tetracyclines 9 J01CA Penicillins with extended spectrum 10 11 J01CE* Beta-lactamase sensitive penicillins 12 J01CF* Beta-lactamase resistant penicillins 13 14 J01CR Combinations of penicillins, incl. beta-lactamase inhibitors 15 J01DB* First-generations cephalosporins 16 For peer review only 17 J01DC* Second-generation cephalosporins 18 J01DD* Third-generation cephalosporins 19 J01EA Trimethoprim and derivatives 20 21 J01EB* Short-acting sulfonamides 22 J01EE Combinations of sulfonamides, and trimethoprim, incl. derivatives 23 24 J01FA* Macrolides 25 J01FF* Lincosamides 26 27 J01GB* Other aminoglycosides 28 J01MA Fluoroquinolones 29 30 J01MB* Other quinolones 31 J01XC* Steroid antibacterials 32 33 J01XE Nitrofuran derivatives 34 J01XX Other antibacterials 35 4 All antimicrobial drug classes that were prescribed for UTIs in the study population. The number of prescriptions of drugs with * 36 5 was very low and are not shown in the figures. 37 6 38 7 39 8 40 9 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 34 BMJ Open
1 2 3 10 Supplementary table 2: ICPC codes 4 11 5 ICPC code Indication 6 7 A03 Fever 8 A78 Lyme disease 9 10 A99 Other generalized/not specified diseases 11 B70 Acute lympadenitis 12 13 D70 Infectious diarrhea 14 D73 Supposititious gastroenteritis 15 F70 Infectious conjunctivitis 16 For peer review only 17 F72 Blepharitis 18 F73 Other infection/infection eye/adnexa 19 20 H04 Excretion from ear 21 H70 Otitis externa 22 23 H71 Acute otitis media 24 H72 Otits media with effusion 25 26 H74 Chronic otitis media/mastoiditis/other infection 27 R02 Dyspnoea attributed to airways 28 29 R05 Cough 30 R09 Symptoms/complaints sinuses 31 R21 Sore throat 32 33 R22 Symptoms/complaints tonsils 34 R70 Tuberculosis (airways) 35 36 R71 Pertussis 37 R72 Scarlet fever 38 39 R74 Cold/pharyngitis 40 R75 Acute/chronic rhinosinusitis 41 42 R76 Acute tonsillitis/peritonsillar abscess 43 R77 Acute laryngitis/tracheitis 44 45 R78 Acute bronchitis 46 R81 Pneumoniae 47 48 R83 Other respiratory infections 49 R90 Hypertrophy/chronic infection tonsils 50 S06 Local redness/erythema skin 51 52 S09 Local infection finger/toe/paronychia 53 S10 Furuncle/karbunkel/local cellulitis 54 55 S11 Folliculitis 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 34
1 2 3 S12 Tick bite 4 5 S13 Bite human/animal 6 S18 Cut 7 8 S76 Other skin infection 9 S84 Impetigo 10 11 S87 Constitutional eczema 12 S88 Other eczema 13 S96 Acne 14 15 U01 Painful micturition* 16 U02 For peerFrequent review micturition* only 17 18 U70 Acute pyelonephritis* 19 U71 Cystitis* 20 21 U72 Urethritis 22 U88 Glomerulonephritis 23 24 X84 Vaginitis 25 X99 Other disease of female reproduction organs 26 27 W94 Mastitis 28 Y74 Orchitis/Epididymitis 29 30 Y75 Balanitis 31 12 All indications that were used in the study as possible indications for which antimicrobial drugs were prescribed. Antimicrobial drugs prescribed 32 13 for the indications with * were assigned as antimicrobial drugs prescribed for UTIs. 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 34 BMJ Open
1 2 3 16 4 5 Men 6 14 7 8 12 9 10 10 11 12 13 8 14 15 6 16 For peer review only
17 Prescriptions (per1,000 PY) 4 18 19 2 20 21 22 0 23 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 24 14 Year 25 160 26 Women 27 28 140 29
30 120 31 32 100 33 34 35 80 36 37 60 38
39 Prescriptions per1,000 PY 40 40 41 20 42 43 44 0 45 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 46 15 Year 47 16 Supplementary figure 1: The number of prescriptions of nitrofurantoin with and without an indication code for urinary tract 48 17 infections for men and women. Note that the scale of the y-axis differs between men and women. 49 50 18 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 34
1 2 3 100 4 Users 5 90 6 7 80 8 9 70 10 11 60 12 50 13 14 40 15 16 30 For peer review only 17 18 20 19 20 10 UTI UTI userswithin users with indication code(%) 21 0 22 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 23 Year 24 19 25 100 26 Prescriptions 27 90 28 29 80 30 31 70 32 60 33 34 50 35 36 40
37 indication code(%) 38 30 39 40 20
41 UTI prescriptionswithin prescriptions with 10 42 43 0 44 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 45 Year 46 20 47 48 21 Supplementary figure 2: The percentage of users/prescriptions with an indication code for UTIs within all users/prescriptions 49 22 with an indication code (thus including indication codes for UTIs, airway infections, ear infections, skin infections etc.). 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 34 BMJ Open
1 2 3 120 450 4 Users men 5 400 Users women 100
6 350 7 80 300 8 9 250 60 10 200 11 40 150
12 perPY 1,000 ( Users For peer PY) 1,000 (per Users review only 100 13 20 14 50 15 0 0 16 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 17 Year Year 18 19 120 450 20 Prescriptions men Prescriptions women 400
21 100 22 350 23 80 300 24 25 250 60 26 200 27 40 28 150 29 100 Prescriptions PY) 1,000 (per Prescriptions 30 20 PY) 1,000 (per Prescriptions 50 31 32 0 0 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 33 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 34 Year Year 35 Supplementary figure 3: The number of users and prescriptions of antimicrobial drugs for urinary tract infections for different age groups per year. Note that the scale of the y-axis differs 36 between men and women. 37 38 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 32 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 34 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 34
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml