BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from
Medication Incidents in Primary Care Medicine: a Prospective Study in the Swiss Sentinel Surveillance Network (Sentinella)
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2016-013658
Article Type: Research
Date Submitted by the Author: 29-Jul-2016
Complete List of Authors: Gnädinger, Markus; University of Zurich, Institute for General Medicine Conen, Dieter; Swiss Patient Safety, Herzig, Lilli; University of Lausanne, Institute of General Medicine Puhan, Milo; University of Zurich, Institute of Epidemiology, Biostatistics & Prevention Staehelin, Alfred; Sentinel Surveillance Network, Swiss Federal Office of Public Health, Zoller, Marco; Zurich University Hospital, Instiute for General Medicine Ceschi, Alessandro; National Poisons Centre, Tox Info Suisse, Associated Institute of the University of Zurich, Division of Science, Head of ; University Hospital Zurich , Dept. Clinical Pharmacology & Toxicology http://bmjopen.bmj.com/ Primary Subject http://bmjopen.bmj.com/ General practice / Family practice Heading:
Secondary Subject Heading: Pharmacology and therapeutics, Paediatrics
CLINICAL PHARMACOLOGY, Health & safety < HEALTH SERVICES Keywords: ADMINISTRATION & MANAGEMENT, Adverse events < THERAPEUTICS
on September 28, 2021 by guest. Protected copyright. on September 28, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Medication Incidents in Primary Care Medicine 5 6 7 8 A Prospective Study in the Swiss Sentinel Surveil 9 10 lance Network (Sentinella) 11 12 1Markus Gnädinger (corresponding author, [email protected]), 2Dieter Conen, 3,4Lilli 13 5 1,4 1 6 14 Herzig, Milo Puhan, Alfred Staehelin, Marco Zoller, Alessandro Ceschi 15 For peer review only 16 1Institute of Primary Care, University of Zurich, 2Patientensicherheit Schweiz, Zurich 3Policlinique 17 4 18 Médicale, University of Lausanne, Sentinel Surveillance Network, Swiss Federal Office of Public 19 Health, Bern (Sentinella), 5Epidemiology, Biostatistics, and Prevention Institute, University of Zur 20 6 21 ich, Division of Clinical Pharmacology and Toxicology, Department of Internal Medicine, Ente 22 Ospedaliero Cantonale, Lugano, Switzerland and Department of Clinical Pharmacology and Toxi 23 24 cology, University Hospital Zurich, Zurich, Switzerland and National Poisons Centre, Tox Info 25 Suisse, Associated Institute of the University of Zurich, Zurich, Switzerland 26 27 28 (Correspondence: Markus Gnädinger, Dr. med. Facharzt für Innere Medizin, Birkenweg 8, 9323 Steinach, 29 0041 71 446 04 64) 30 31 32
33 Objectives: To describe the type, frequency, seasonal and regional distribution of medication http://bmjopen.bmj.com/ 34 incidents in primary care in Switzerland and to elucidate possible risk factors for medication inci 35 36 dents. 37 38 Design: Prospective surveillance study. 39 40
41 Setting: Swiss primary health care, Swiss Sentinel Surveillance Network. on September 28, 2021 by guest. Protected copyright. 42 43 Participants: Patients with drug treatment who experienced any erroneous event related to the 44 45 medication process and interfering with normal treatment course, as judged by their physician. 46 The 180 physicians in the study were general practitioners or pediatricians participating in the 47 48 Swiss Federal Sentinel reporting system in 2015. 49 50 Outcomes: Primary: medication incidents; secondary: potential risk factors like age, gender, 51 52 poly medication, morbidity, care dependency, previous hospitalization. 53 54 55 Results: The mean rates of detected medication incidents were 2.07 per general practitioner and 56 year (46.5 per 100,000 contacts) and 0.15 per pediatrician and year (2.8 per 100,000 contacts), 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 respectively. The following factors were associated with medication incidents (OR, 95% CI): high 4 er age 1.004 per year (1.001;1.006), care dependency 1.458 (1.025;2.073) for care by community 5 6 nurse, and 1.802 (1.399;2.323) for care by an institution, chronic conditions 1.052 per condition 7 (1.029;1.075), medications 1.052 per medication (1.030;1.074), as well as Thurgau Morbidity In 8 9 dex for stage 4 1.292 (1.004;1.662), 5 1.420 (1.078;1.868), and 6 1.680 (1.178;2.396), respective 10 11 ly. Most cases were linked to an incorrect dosage for a given patient, while prescription of an er 12 roneous medication was the second most common error. 13 14 15 Conclusions:For Medication peer incidents are commonreview in general medicine only whereas they rarely occur in 16 pediatrics. Reasons for medication incidents are diverse but often seem to be linked to communi 17 18 cation problems. Older and multimorbid patients are at a particularly high risk for medication inci 19 dents. 20 21 22 Trial registration: www.clinicaltrials.gov, NCT0229537 23 24 Keywords: Patient safety, pharmaceutical preparations, medication errors. 25 26 27 Strength and limitations 28 This is the first Swiss prospective and systematic collection of incident data in primary care. 29 30 It covers three linguistic regions and two drug distribution systems. 31 It was conducted by experienced physicians and with high response rates. 32
33 http://bmjopen.bmj.com/ 34 There was – as expected – bias from selective and underreporting or non detection of medication 35 incidents. 36 37 38 39 40 Proposed reviewers:
41 on September 28, 2021 by guest. Protected copyright. 42 Prof. Tobias Dreischulte, Population Health Sciences, School of Medicine 43 44 The Mackenzie Building, Kirsty Semple Way, Dundee DD2 4BF, 45 46 [email protected] 47 Prof. Meredith A B Makeham, MPH(Hons), FRACGP, Lecturer and NHMRC Scholar, Dis 48 49 cipline of General Practice, The University of Sydney, Sydney, NSW. make 50 [email protected] 51 52 53 54 55 56 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Introduction 4 5 6 Patient safety is a major concern in healthcare systems worldwide. Although most safety research 7 has been conducted in the in patient setting [1], evidence indicates that medical errors and ad 8 9 verse events pose a serious threat for patients in the primary care setting as well, since most pa 10 tients receive ambulatory care [2 4]. The rationale of this project has been published in our study 11 12 protocol [5]. The aim of the project was to describe the type, incidence, seasonal and regional dis 13 tribution of medication incidents in primary care in Switzerland and to elucidate risk factors for 14 15 medication incident.For peer review only 16 17 18 19 20 Method 21 22 Study design 23 24 25 We conducted a prospective surveillance study among primary care patients during 2015 to identi 26 27 fy cases of medication incidents. 28 29 Study population 30 31 32 The study population was any person undergoing drug treatment in general internal or pediatric
33 practices participating the Sentinella network. The latter covers a representative sample of patients http://bmjopen.bmj.com/ 34 35 in primary care for Switzerland [6, included to this manuscript]. Founded in 1986, it was mainly 36 designed to survey transmissible diseases. Later, it also assessed other health problems of public 37 38 interest. It generates daily to weekly current data and covers the entire geographic and linguistic 39 regions of our country. Children, the mentally handicapped or the elderly were also included, all of 40 whom might be at increased risk for medication errors.
41 on September 28, 2021 by guest. Protected copyright. 42 43 Medication incidents 44 45 We defined medication incidents as any erroneous event (as defined by the physician) related to 46 47 the medication process and interfering with normal treatment course (e.g. administration of an er 48 roneous medication). We did not include lack of treatment effect, adverse drug reactions or drug 49 50 drug or drug disease interactions without detectable treatment error. Nor did we consider medica 51 tion incidents if patients refused to have them reported to the Sentinella system. 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 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Data sources 4 5 6 The study physicians recorded the patient’s year of birth and gender on their weekly reporting 7 form. After a maximum of four weeks they had to fill in a detailed incident questionnaire (Appendix 8 9 A). It was comprised of their Sentinella number and the calendar week of notification. Concerning 10 the patients, they reported the living situation, several supposed risk factors for an incident, as well 11 12 as the following variables: hospitalization during the previous year, care dependency, number of 13 14 drugs used chronically, number of chronic conditions, and the Thurgau Morbidity Index (TMI), to be 15 compared withFor a denominator peer analysis (below)review [6]. We further receivedonly a detailed description of the 16 17 incident and proposals to avoid future incidents. 18 19 We got the annual number of patient to physician contacts (PPC) per practice from the Sentinella 20 21 administration, as well as morbidity data from a fortnight cross sectional denominator analysis of all 22 patients consulting a Sentinella practice during weeks 11 or 12 [6]. 23 24 25 We received the anonymized list of participating physicians, their specialty, as well as the commu 26 nity size (Swiss Federal Statistical Office) and the linguistic region from the Sentinella administra 27 28 tion. Information on Swiss medication sales in 2015 by ATC groups was derived from Interpharma 29 Switzerland (Appendix F). 30 31 32 The questionnaires were completed either electronically or on a paper/pencil version, the former as
33 online SurveyMonkey™ questionnaires, the latter as sealed envelopes sent from the Sentinella http://bmjopen.bmj.com/ 34 35 administration. We had three study questionnaires: a detailed incident questionnaire, an initial one, 36 and a final one (Appendices A, B, and C). The initial questionnaire served to describe the physi 37 38 cian’s practices in terms of e.g. number of physicians, availability of electronic patient history, elec 39 tronic drug drug interaction control system, and drug distribution system (Appendix B). The final 40 questionnaire investigated non reporting and difficulties with coding the morbidity variables (Ap 41 on September 28, 2021 by guest. Protected copyright. 42 pendix C). 43 44 45 We calculated the variable “incident relevance” from the items “disturbance” and “endangering” of 46 the patients; if any of them was graded with “medium” or higher, the variable relevance was set to 47 48 “more”, otherwise it was set to “less”. Because of question ambiguity, we set coding of the item 49 50 care dependency to “missing” for patients younger than 20 years of age. Evans’ Index – a prog 51 nostic index – was calculated by simple addition of the number of chronic drug treatments with the 52 53 number of chronic conditions [7]. 54 55 56 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 The following free text variables were manually coded: relationship of the incident to the suspected 4 medication, preventability of the incident, reactions to the incident, and proposals to avoid further 5 6 incidents. 7 8 Statistical methods 9 10 Values are given as frequencies, mean ± SD or median [interquartile range (IQR)], depending on 11 non normal distribution or non interval scaled data level. To assess the association of medication 12 13 incidents with potential risk factors we used the GENLINMIXED procedure. Clustering of patients 14 was addressed by using a mixed binary logistic regression with the fixed factors of gender, year of 15 For peer review only 16 birth, care dependency, number of chronic drug treatments, number of chronic conditions, and TMI 17 18 as well as the physicians’ practice number as a random factor; if one item was missing, the whole 19 record was excluded from the analysis. We used IBM SPSS 23. 20 21 22 23 Results 24 25 The Sentinella system 26 27 During the year 2015, 149 practices were enrolled to the Sentinella system. Of them, 144 practices 28 were known to report regularly; their properties are listed in Table 1. The Sentinella physicians are 29 30 representative for the overall Swiss physician population [5,6]. Drugs were auto distributed by 42% 31 of the study practices. Approx. half of the physicians had electronic and the other paper based 32
33 medical records. Systematic drug drug interaction control systems were installed only by a minority http://bmjopen.bmj.com/ 34 (36.8%) of the physicians. During the year 2015 (which included unusually for calendar adaptation, 35 36 53 instead of 52 reporting weeks), the general practitioners (GPs) had 4,456±2,137 PPC; for the 37 pediatricians (PEDs), these were 5,297±2,715. 38 39 40 Study flow
41 on September 28, 2021 by guest. Protected copyright. 42 During the year 2015, we received 216 incident notifications (Figure 1). In 11 cases, we did not 43 receive the detailed notification form. Eight cases had to be removed from the database because 44 45 they fulfilled the exclusion criterion (i.e. “adverse drug reactions without detectable error”), and one 46 case had to be removed because of double data entry. This led to 197 cases which could be ana 47 48 lyzed. The distribution of the monthly incident notifications throughout 2015 is depicted in Figure e1 49 50 (Appendix D), and the distribution of the numbers of cases reported by each practice in Figure 2. 51 52 53 Description of patients 54 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Table 2 lists age, gender and geographic distribution as well as the physician to patient relation 4 ship and the observer of the incident. Only three cases evolved in PED practices. No statistically 5 6 significant differences were found between the two relevance classes of the incidents. 7
8 9 Number of incidents, non-reporting 10 In the GPs 194 incidents, 148 physicians, 4,456 yearly PPC led to 1.31 incidents per physician and 11 12 year or 29.4 per 100,000 PPC; in PEDs 3 incidents, 32 physicians, 5,297 yearly PPC led to 0.1 13 14 incident per physician and year or 1.8 per 100,000 PPC. 15 For peer review only 16 17 To evaluate the non-reporting of incidents, we asked the physicians in the final study question 18 naire. Out of 180 actively reporting physicians we received 145 questionnaires (80.6% response 19 20 rate). To our question: “Did you not report medication incidents that you had noticed during the last 21 year?”, they answered: “Never or almost” 110 (75.9%), “yes, but seldom” 22 (15.2%), “yes, fre 22 23 quently” 9 (6.2%), “always, or almost” 4 (2.8%). Reasons for not reporting incidents were “lack of 24 time” or “forgetfulness”. If we speculate that these answers represent reporting rates of 76 100, 51 25 26 75, 26 50, or 0 25% respectively, we could divide the observed rates by the middle of the reporting 27 classes: 0.875, 0.625, 0.375 and 0.125. By doing so, we calculated a rate of 50% of underreport 28 29 ing; if we furthermore consider the 5% of the incidents where the questionnaires were not sent, this 30 rate increases to 58%. We therefore have to multiply the observed rates with a factor of 1.58 re 31 32 sulting in the following rates of detected incidents: GP 2.07 per physician and year, 46.5 per
33 http://bmjopen.bmj.com/ 100,000 PPC and PED 0.15 per physician and year, 2.8 per 100,000 PPC. 34 35 36 Types and causes of incidents, organ systems involved, preventability 37 38 The types of error are listed in Table e1 (Appendix D) and Figure 3; in 26 cases more than one 39 40 was mentioned. Most cases were linked to an incorrect dosage for a given patient, while prescrip
41 tion of an erroneous medication was the second most common error. on September 28, 2021 by guest. Protected copyright. 42 43 44 Most errors concerned orally applied medication. Parenterally applied drugs led to fewer error re 45 porting; in our study they comprised insulin and vaccinations. There were cases of an incorrect 46 47 (influenza vs. anti tetanus) or incomplete (Boostrix® vs. Boostrix Polio®) vaccination and of undue 48 vaccination (a third anti HPV vaccine). The first case was due to a communication problem within 49 50 the practice staff, the other to a vaccination in absence of the vaccination card. 51 52 In 89 of the 195 cases, organ system damage was reported, in 13 cases more than one organ sys 53 54 tem was involved, most frequently the central nervous system (Figure 4) (Table e2, Appendix D). 55 Possible triggers of the incident were reported in 194 of the 197 cases; in 45 cases there was more 56 57 than one single reason, most frequently lacking alertness of the reporting physicians or their staff 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 (Table e3). When asked who might be the “responsible” for the incident, 191 replied. The most 4 common person or institution possibly responsible for the medication incident was: the reporting 5 6 physician 41 (21.5%), followed by the practice nurse 26 (13.6%), the institution where the patient 7 lives 33 (17.3%), the pharmacy 7 (3.7%), the hospital 12 (6.3%), the community nurse 4 (2.1%), 8 9 the patients or their proxies 15 (7.9%), and the manufacturer 2 (1.0%); in 9 cases (4.6%) this was 10 unclear, in 37 cases (19.4%), there was more than source one to blame. Preventability of the inci 11 12 dents was classified (by our study board): unlikely in 6 cases (3.0%), possible in 58 (29.4%), prob 13 14 able in 114 (57.9%), and definite in 19 (9.6%). 15 For peer review only 16 17 Endangering and disturbances 18 In 192 of the 197 cases we received information about patients’ endangering as estimated by the 19 20 physicians. In 39 cases (20.3%) there was no endangering, in 83 (43.2%) light, in 51 (26.6%) 21 moderate, in 19 (9.9%) severe. The disturbances caused by the incident are listed in Table 3. Out 22 23 of the 197 incidents, 74 (37.5%) were classified as “more” relevant (Table 2). 24 25 26 Interface problems, orientation, predictability, repeat incidents 27 The presence of interface problems was reported in 64 of 197 cases (32.4%); these were: with a 28 29 hospital in 28 cases (43.8%), with an institution in 14 (21.9%), with a community nurse in 6 (9.4%), 30 with a pharmacist in 9 (14.0%), with a specialist in 3 (4.7%), and with others in 4 (6.3%). In 184 31 32 cases, we received information about orienting patients about the incident; in 98 cases, the patient
33 http://bmjopen.bmj.com/ was oriented by the reporting physician or his staff (50.3%), non orientation was stated: because 34 35 the patient was not able to understand (children, demented) in 26 cases (14.1%), because the 36 problem had already been solved and the notification would have unduly disturbed the confidence 37 38 in 18 cases (9.8%), because the patient or the proxies had observed themselves in 23 cases 39 40 (12.5%), because the patient had already been oriented by others in 7 cases (3.8%) or because of
41 another reason in 12 cases (6.5%). As for predictability of the incident, we received 183 valid an on September 28, 2021 by guest. Protected copyright. 42 43 swers; of them 82 (44.8%) stated “yes, in the given constellation, the incident was to be expected”. 44 When asked whether they had already reported a similar incident in the study, 29 out of 196 45 46 (14.8%) answered in the affirmative. 47 48 49 Risk factors to undergo a medication incident 50 To detect patient risk factors to undergo an incident, we compared the incident data with those of a 51 52 fortnight denominator analysis [6]; the following univariate factors accumulated preferentially in 53 incident patients: higher age, care dependency, higher numbers of chronic conditions or medica 54 55 tions (or higher Evans’ Index), as well as higher TMI (table 4). In the multivariate analysis only in 56 patient care by an institution remained a significant factor. Other suspected risk factors are listed in 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Table e4 (Appendix D); these items were not included in the denominator study and therefore lack 4 a comparator. When discerning patients with a more from them with a less relevant incident, only 5 6 psychiatric illness reached a significantly increased proportion. We did not detect major differences 7 between the two drug distribution systems. 8 9 10 Relationship between incidents and drug class, ATC-group 11 We assessed semi quantitatively whether the class of the suspected drug was causally linked to 12 13 the emergence of the incident; the frequencies were: none 49 (25.1%), unlikely 31 (15.9%), possi 14 ble 77 (39.5%), probable 34 (17.4%), definite 4 (2.1%), did not apply 2. ATC codes of the suspect 15 For peer review only 16 ed medications are listed in Table 5. Most naming concerned the groups C “cardiovascular” 17 18 (23.2%) and N “nervous system” (22.1%). Among the medication classes judged to be related to 19 the incident (n=37) the most frequently named group were oral anticoagulants: rivaroxaban 9, 20 21 phenprocoumon 7, and acenocoumarol 2 cases; this explains the 7 fold increased relative risk of 22 ATC group B as compared to sales. As an example for errors without relation to the drug class we 23 24 must mention the 17 cases of institutionalized patients ingesting medications scheduled for other 25 residents, in most cases this was person eating at the same table, but in other cases the person in 26 27 care mixed up patient names. 28 29 30 Reactions to the incident and suggestions to prevent further ones 31 In 141 of the 197 cases, the physicians reported to have changed something after the incident, in 32
33 13 cases this was more than one reaction, mostly often named were communication with other http://bmjopen.bmj.com/ 34 caregivers or better instruction of patients (Table e5, Appendix D). The respondents were asked for 35 36 proposals how to prevent future incidents of the reported type; 125 of them made a total of 243 37 suggestions (Table e6, Appendix D). Of these, 37 (15.2%) were related to accurate medication 38 39 lists, 10 (4.1%) to better patient instructions, 38 (15.6%) to organization of regular follow up con 40 trols, and 55 (22.6%) involved organizational changes within the practice and its staff.
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 Discussion 45 46 In a representative group of primary care physicians [8], we found approximately one case of a 47 48 medication incident per GP and year. 49 50 51 Incident rates 52 We calculated the rates of detected medication incidents as follows: GP 2.07 per physician and 53 54 year, 46.5 per 100,000 PPC and PED 0.15 per physician and year, 2.8 per 100,000 PPC. Medica 55 tion incidents may make up a proportion of approximately one third of all incidents [9]; the rates for 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 all safety incidents may amount 6.20 per physician and year or 139.4 per 100,000 PPC in GPs and 4 0.45 per physician and year or 8.4 per 100,000 PPC in PEDs. 5 6 In Australian primary care patients an incident rate of 4.98 per GP and year was reported; this is 7 close to our estimated rate of 6.20; hence, there was no sub typing of the incidents [10]. In a three 8 9 year study, O’Beirne and colleagues reported a rate of 1.8 safety incidents per year and physician 10 [9]. In a literature review of western countries’ publications, Sandars and Esmail described a rate of 11 12 5 to 80 errors per 100,000 consultations [2], a rate somewhat lower than the 139.4 cases per 13 14 100,000 consultations estimated in our study. Kuo and colleagues reported a proportion of 15% of 15 all incidents toFor be related peer to medication [11];review since we estimated only a proportion of 33% this would 16 17 give rise to even higher rates of all safety incidents when calculated from our study database. The 18 13 fold higher incident rate in GPs as compared to PEDs is not surprising given the lower medica 19 20 tion rates in children. A recent British study confirmed this much lower rate of incidents in children; 21 out of 46,902 family practice safety reports, 1,788 concerned children (26 times less than adults) 22 23 [12]. 24 25 Definition of incidents and reliability of reporting 26 27 On the other hand, incident rates may be influenced by their definition. Gandhi and colleagues 28 coined the term “avoidable adverse drug reaction” [13]. In our study, we explicitly excluded adverse 29 30 drug reactions without detectable error. Runciman and colleagues defined a patient safety incident 31 as follows: “An event or a circumstance that could have resulted or did result in unnecessary harm 32
33 to a patient” [14]. An intuitive definition of a medical error was given by Makeham et al: “That was a http://bmjopen.bmj.com/ 34 threat to patient wellbeing and should not happen. I don’t want it to happen again.” [15]. As shown 35 36 in Figure e1 (Appendix D), reporting frequency was higher at the beginning of the study as com 37 pared to the later course of it. This could reflect some loss of interest or forgetfulness by the report 38 39 ing physicians. As calculated from our final questionnaire after the study, the reporting physicians 40 failed to report about one in three cases of the detected medication errors. Non detection of inci
41 on September 28, 2021 by guest. Protected copyright. 42 dents may even be more frequent than non reporting of observed incidents. A missed possible 43 44 drug drug interaction may be detected by chart review, a documentation error would probably have 45 been found only in 1:1 supervision, which is very time consuming, costly and may additionally in 46 47 fluence performance of the observed physician. It is therefore virtually impossible to decipher the 48 real rate of non detected incidents. The problems in detection of incidents were the reason to pos 49 50 tulate a “mix of methods” as needed to identify adverse events in general practice [16]. 51 52 53 54 Other approaches 55 In a retrospective, semi quantitative analysis, Gehring and colleagues investigated safety incidents 56 57 in Swiss primary care [17]; among 23 predefined classes of safety incidents, the respondents ad 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 mitted 15 to have occurred at least yearly – four of them being linked to drug treatment. Another 4 approach to incidents is to ask patients as performed by Mira and colleagues [18]; they interviewed 5 6 patients (> 65 years with five or more chronic drug treatments) and found that 75% of the patients 7 reported to have been affected by at least one medication error during the previous twelve months. 8 9 We did not collect data on appropriateness of treatment; hence medication inappropriate for some 10 groups of patients (elderly or patients with impaired renal function) may provoke incidents [19]. A 11 12 recent Scottish study demonstrated an impressive reduction in high risk prescriptions (nonsteroidal 13 14 anti inflammatory drugs, antiplatelet, and anticoagulation) as well as hospital admissions for gas 15 trointestinal ulcersFor or heart peer failure by a combinationreview of educational only measures, informatics and fi 16 17 nancial incentives [20]. 18 19 20 21 Type, consequences, causes and preventability of error 22 Most cases involved application of an erroneous dosing or of a wrong medication; non application 23 24 of necessary medication was also frequent. The prototype of wrong medication was confounding of 25 prepared medication in home residents. The classic case of non application of a necessary medi 26 27 cation was the missed re uptake of anticoagulation after an operation. Non application of neces 28 sary drugs seems to be a relevant source of unnecessary harm to patients [21]. The repartition of 29 30 the incidents was similar as reported by others [11,17,18,22]. More than half of the patients did not 31 have any disturbances after the incident; otherwise in most cases the nervous system was affect 32
33 ed. No fatalities were reported, but seven patients (3.5%) needed stationary care. In 2004 Piro http://bmjopen.bmj.com/ 34 hamed and colleagues published a study on adverse drug reactions as a cause for admission to 35 36 hospital; they found that 6.5% of all hospitalizations and 4.0% of all hospital stays were caused by 37 adverse drug reactions, 72% of them preventable and 2% leading to death [23], hence, this study 38 39 included all cases, and therefore a majority of cases of adverse drug reactions without a detectable 40 error. An older Swiss study was published by Livio and colleagues; out of 3,195 hospitalizations,
41 on September 28, 2021 by guest. Protected copyright. 42 she identified 229 cases (7.2%) as probably caused by adverse drug reactions [24]. In that study, 43 44 32% of the events were classified as being preventable (which sensu strictu does not mean that an 45 error had caused the incident), and in 6% of the cases, fatalities were reported. Concerning cir 46 47 cumstances, most naming was – as found in our patients – lacking alertness and communication 48 problems within practice staff, but there was a large variety of other topics. In inpatients, one in ten 49 50 drug administrations was described to be erroneous [25]. Medical errors seem to be the “third lead 51 ing cause of death in the US” [26]. 52 53 54 55 Risk factors 56 57 When looking for propensity factors, the univariate analysis comparing a fortnight denominator with 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 our data revealed all factors investigated (except gender) as significantly accumulated in the inci 4 dent group; these were higher age, care dependency, higher numbers of chronic conditions or 5 6 medications (and higher Evans’ Index), as well as higher TMI. When performing a multivariate 7 analysis, only those living in a home for the elderly or handicapped remained a significant risk fac 8 9 tor, but the small number of observations precluded possible less important risk factors to be de 10 tected. The only significant risk factor for undergoing an incident of higher relevance was psychiat 11 12 ric disease. In the literature there are several factors described to correlate with proneness to un 13 14 dergo a medication incident, which quite well reflects the results of our study. Mainly, this is evi 15 dently the numberFor of drugs peer ingested [13,27]. review Higher or young ageonly [28], or morbidity [29,30] were 16 17 also described. Most incidents concerned ATC groups N “nervous system” or C “cardiovascular”, 18 which were also of the mostly sold drugs; an exception was group B with anticoagulants and a 19 20 sevenfold increased relative risk as compared to Swiss sales in 2015. It seems wise to be alert to 21 avoid errors when prescribing medication of these groups. The prevailing position of anticoagulants 22 23 (18.5% of cases) was described also by Field and colleagues [30]. Otherwise the repartition of our 24 suspected drugs was similar to other primary care studies [11,30], a literature review [31] or a theo 25 26 retical paper [22]. 27 28 Prospects 29 30 First, incidents were endured like stormy weather, and worse culprits were blamed. Thereafter – 31 inspired by flight companies critical incident analysis (safety I) was adopted in medicine, preferen 32
33 tially in anesthesia. In anonymous systems, incidents were analyzed and ideas come up with to http://bmjopen.bmj.com/ 34 avoid further emergence of similar incidents. Perhaps this is now the moment to think about safety 35 36 II, meaning that systems should be organized in a resilient way insensitive to perturbations and 37 tolerant to time pressure, misunderstandings and mistakes [32,33]. 38 39 40
41 Conclusion on September 28, 2021 by guest. Protected copyright. 42 43 Medication incidents are common in general medicine whereas they rarely occur in pediatrics. 44 45 Reasons for medication incidents are diverse but often seem to be linked to communication prob 46 lems. Older and multimorbid patients are at a particularly high risk for medication incidents. 47 48 49 50 51 52 Abbreviations 53 54 GP general practitioner 55 56 57 PED pediatrician 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 PPC patient to physician contacts 4 5 TMI Thurgau Morbidity Index 6 7 8 Author affiliations 9 10 MG Institute of Primary Care, University of Zurich, Switzerland. 11 12 AC Division of Clinical Pharmacology and Toxicology, Department of Internal Medicine, Ente 13 14 Ospedaliero Cantonale, Lugano, and Department of Clinical Pharmacology and Toxicology, 15 UniversityFor Hospital Zurich,peer Zurich, andreview National Poisons Centre,only Tox Info Suisse, Associated 16 17 Institute of the University of Zurich, Zurich, Switzerland 18 DC Patientensicherheit Schweiz, Zurich, Switzerland. 19 20 LH Policlinique Médicale, University of Lausanne, Switzerland and Sentinel Surveillance Network, 21 Swiss Federal Office of Public Health, Bern, Switzerland. 22 23 MP Epidemiology, Biostatistics, and Prevention Institute, University of Zurich, Switzerland. 24 AS Sentinel Surveillance Network, Swiss Federal Office of Public Health, Bern, Switzerland. 25 26 MZ Institute of General Medicine, University of Zurich, Switzerland. 27 28 29 Acknowledgements 30 31 32 We are grateful to Lee Wennerberg for the English language corrections, Dr Sven Staender,
33 http://bmjopen.bmj.com/ Männedorf for the helpful comments, and Simon Gnädinger, Zurich for the manual coding of free 34 35 text items. We thank the Sentinella program commission for their support, the reporting physicians 36 of Sentinella for their unflagging enthusiastic collection of data, and the Federal Office of Public 37 38 Health for providing data and translating the questionnaires into French. 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 Contributorship statement 44 45 MG lead the study, did the pilot study (questionnaire development, data entering and processing) 46 47 wrote all documents, did all the contacts with the Sentinella administration, ethics committee, and 48 others, programmed the electronic questionnaires, entered hand written questionnaires into the 49 50 database, did the data processing and wrote the publication after data collection. 51 AC is an expert on clinical pharmacology and drug safety. 52 53 DC is an expert on patient safety. 54 LH is French speaking and helped to interpret the French questionnaires. She is an expert on mul 55 56 timorbidity. She is a member of the Sentinella program commission. 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 MP is head of Epidemiology, Biostatistics & Prevention Institute. He is responsible for the sound 4 methodology. 5 6 AS had the idea for the study. He is vice president of the Sentinella program commission. 7 MZ is expert on electronic data exchange in primary care. 8 9 All of them have seen all the study documents and have contributed intellectually to their elabora 10 tion. All have contributed to revise the draft of this publication and approve the submitted version of 11 12 this publication. 13 14 15 Funding For peer review only 16 17 The study was funded by Bangerter Rhyner Foundation, Basel. 18 19 20 Competing interests 21 22 23 The authors declare that they have no financial interest conflicts with this study. 24 25 Patient consent 26 27 28 Not necessary for anonymous data. 29 30 31 Ethics approval 32
33 The ethical committee of Canton Zurich decided that our study did not need formal approval be http://bmjopen.bmj.com/ 34 35 cause the data are completely anonymous (KEK ZH 2014 0400). The study was recorded in 36 www.ClinicalTrials.gov: NCT02295371, as well as in our national study registry (www.kofam.ch; 37 38 SNCTP000001207). 39 40 Data sharing statement
41 on September 28, 2021 by guest. Protected copyright. 42 43 No additional data are available. 44 45 Strobe statement 46 47 48 Whenever possible, the guidelines of the Strobe statement were followed (Appendix E). 49 50
51 52 References 53 54 55 1. Manias E: Detection of medication related problems in hospital practice: a review. Br J Clin 56 Pharmacol 2012; 76 (1): 7 20. 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 2. Sandars J, Esmail A: The frequency and nature of medical error in primary care: under 4 standing the diversity across studies. Fam Pract 20 (3): 231 6, 2003. 5 6 3. Miller GC, Britth HC, Valenti L. Adverse drug events in general practice patients in Aus 7 tralia. Med J Aust. 2006;184(7):321 324. 8 9 4. Thomsen LA, Winterstein AG, Søndergaard B, Haugbølle LS, Melander A. Systematic 10 review of the incidence and characteristics of preventable adverse drug events in ambulato 11 12 ry care. Ann Pharmacother. 2007; 41(9):1411 1426. 13 14 5. Gnädinger M, Ceschi A, Conen D, et al: Medication incidents in primary care medicine: 15 For peer review only 16 protocol of a study by the Swiss Federal Sentinel Reporting System. BMJ Open 04/2015; 17 18 5(4): e007773. DOI:10.1136/bmjopen 2015 007773. 19 20 6. Gnädinger M, Herzig L, Ceschi A et al: The Burden Related to Chronic Conditions and 21 Multimorbidity in the Swiss Primary Care Population: A Prospective Study in the Swiss Sen 22 23 tinel Surveillance Network (Sentinella), 2016, to be submitted. That manuscript is includ- 24 ed in the present submission. 25 26 7. Evans DC, Cook CH, Christy JM et al: Comorbidtiy polypharmacy scoring facilitates out 27 come prediction in older trauma patients. J Am Geriatr Soc 60: 1465 70, 2012. 28 29 30 8. Cohidon C, Cornuz C, Senn N: Primary care in Switzerland: evolution of physicians‘ pro 31 file and activities in twenty years (1993 2012). BMC Fam Pract 16: 107, 2015. 32
33 9. O’Beirne M, Sterling PD, Zwicker K et al: Safety incidents in family medicine. BMJ Qual http://bmjopen.bmj.com/ 34 35 Saf 20: 1005 10, 2011. 36 10. Makeham MAB, Kidd MR, Saltman DC et al: The threats to Australian patient safety 37 38 (TAPS) study: incidents of reported errors in general practice. Med J Aus 185: 95 98, 2006. 39 11. Kuo GM, Phillips RL Graham D et al: Medication errors reported by US family physicians 40 and their office staff. Qual Saf Health Care 17: 286 90, 2008. 41 on September 28, 2021 by guest. Protected copyright. 42 12. Rees P, Edwards A, Panesar S et al: Safety incidents in the primary care office setting. 43 44 Pediatrics 135 (6): 1027 35, 2015. 45 46 13. Gandhi TK, Weingart SN, Borus BA, et al: Adverse drug events in ambulatory care. N 47 48 Engl J Med 348: 1556 64, 2003. 49 50 14. Runciman W, Hibbert P, Thomoson R et al: Towards an international classification ofr 51 52 patient safety: key concepts and terms. Int J Qual Health Care 21 (1): 18 26, 2009. 53 15. Makeham MAB, Dovey SM, County M et al: An international taxonomy for errors in gen 54 55 eral practice: a pilot study. Med J Aus 177:68 72; 2002. 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 16. Wetzels R, Wolters R, van Weel C et al: Mix of methods is needed to identify adverse 4 events in general practice: a prospective observational study. BMC Fam Pract 2008; 9:35. 5 6 7 17. Gehring K, Schwappach DLB, Battaglia M et al: Frequency of and harm associated with 8 primary care safety incidents. Am J Managed Care 18 (9): e323 7, 2012. 9 10 18. Mira JJ, Orozco-Beltran D, Perez-Jover V et al: Physician patient communication failure 11 12 facilitates medication errors in older polymedicated patients with multiple comorbidities. 13 Fam Pract 30: 56 63, 2013. 14 15 For peer review only 16 19. Cooper JA, Moriarty F, Ryan C et al: Potentially inappropriate prescribing in two popula 17 tions with differing socio economic profiles: a cross sectional database study using the 18 19 PROMPT criteria. Eur J Clin Pharmacol 72: 583 91, 2016. 20 21 20. Dreischulte T, Donnan P, Grant A et al: Safer prescribing – a trial of education, informat 22 ics, and financial incentives. New Engl J Med 374: 1053 64, 2016. 23 24 21. O’Grady I, Gerrett D: Minimising harm from missed drug doses. Nursing Times 111 (44): 25 26 12 15, 2015. 27 22. Huges RG, Ortiz E: Medication errors. Why they happen, and how they can be prevented. 28 29 Am J Nursing 205 (3): 14 24, 2005. 30 23. Pirmohamed M, James S, Meakin S, et al: Adverse drug reactions as cause of admission 31 32 to hospital: prospective analysis of 18 820 patients. BMJ 2004; 329:15–19.
33 24. Livio F, Buclin T, Yersin B et al: Hospitalisations pour effet indésirable médicamenteux. http://bmjopen.bmj.com/ 34 35 Recensement prospectif dans un service d’urgences médicales [French]. Raisons de Santé 36 23, 1998. 37 38 39 25. Berdot S, Gillaizeau F, Caruba T et al: Drug administration errors in hospital inpatients: a 40 systematic review. PLoS ONE 8 (6): e68856, 2013; doi: 10.1371/journal.pone.0068856.
41 on September 28, 2021 by guest. Protected copyright. 42 26. Makary M, Daniel M : Medical error – the third leading cause of detha in the US. BMJ 43 44 2016 ; 353: i2139. 45 27. Nobili A, Pasina L, Tettamanti M et al: Potentially severe drug interactions in elderly out 46 47 patients: results of an observational study of an administrative prescription database. J Clin 48 Pharm Ther 34: 377 86, 2009. 49 50 51 28. Avery AJ, Ghaleb M, Barber N, et al: The prevalence and nature of prescribing and moni 52 toring errors in English general practice: a retrospective case note review. Br. J Gen Pract 53 54 e543, 2013. 55 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 29. Field TS, Gurwitz JH, Avorn J et al: Risk factors for adverse drug events among nursing 4 home residents. Arch Intern Med 161: 1629 34, 2001. 5 6 30. Field TS, Mazor KM, Briesacher B, et al: Adverse drug events resulting from patient er 7 rors in older adults. J Am Geriatr Soc 55: 271 6, 2007. 8 9 31. Saedder EA, Brock B, Nielsen LP et al: Identifying high risk medication: a systematic lit 10 11 erature review. Eur J Clin Pharmacol 70: 637 45, 2014. 12 32. Staender S, Kaufmann M. Sicherheitsmanagement 2015: von “Safety I” zu “Safety II”. 13 14 Schweiz Ärztezeitung 96 (5): 154 7, 2015. 15 For peer review only 16 17 33. Gandhi TK, Lee TH: Patient safety beyond the hospital. New Engl J Med 363 (11): 1001 3, 18 2010. 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 List of figures and tables 4 5
6 7 8 Figure 1: Study flow chart 9 10 Figure 2: Distribution of the number of cases reported by practice 11 12 13 Figure 3: Type of error 14 15 Figure 4: OrganFor system involvedpeer review only 16 17 18 Table 1: Characteristics of the reporting physicians in 2015 19 20 Table 2: General description of the cases 21 22 23 Table 3: Disturbances after the incident 24 25 Table 4: Possible risk factors for incident as compared to a denominator analysis during calendar 26 27 weeks 11 and 12 [Gnädinger M 2016] 28 29 Table 5: ACT Groups of suspected medications 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 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 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 Figure 1. Study flow chart 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 Figure 2. Distribution of the number of cases reported by practice 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 Figure 3. Type of error (n=197 questionnaires). For more detailed information, see Ta 44 ble e1 (Appendix D). 45 46 47 48 49 50 51 52 53 54 55 56 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 Figure 4. Organ system involved (n=91 of 197 questionnaires). For more detailed in 35 36 formation, see Table e2 (Appendix D). 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Physicians’ gender 5 male 128 (71.1%) 6 female 52 (28.9%) 7 Physicians’ age class 8 < 40 12 (6.7%) 40 49 44 (24.4%) 9 50 59 66 (36.7%) 10 60 and over 58 (32.2%) 11 Specialty 12 GP 148 (82.2%) 13 PED 32 (17.8%) 14 Number of physicians in practice (n=144) 15 1 For peer review only69 (47.8%) 16 2 39 (27.1%) 17 3 17 (11.8%) 18 4 to 5 8 (5.6%) 19 6 to 9 7 (4.9%) 20 10 and over 4 (2.8%) Number of physicians per practice reporting to Sentinella (n=144) 21 1 119 (82.6%) 22 2 19 (13.2%) 23 3 5 (3.5%) 24 8 1 (0.7%) 25 Linguistic region 26 German 122 (67.8%) 27 French 44 (24.4%) 28 Italian 14 (7.8%) 29 Urbanity of the practice 30 urban 93 (51.7%) 31 agglomeration 60 (33.3%) 32 rural 27 (15.0%) Workload per week 33 http://bmjopen.bmj.com/ < 15 h 9 (5.0%) 34 15 30 h 36 (20.0%) 35 > 30 h 135 (75.0%) 36 Drug distribution system 37 dispensing by physician 73 (42.2%) 38 mixed system 19 (10.6%) 39 dispensing by pharmacy 85 (47.2%) 40 Electronic documentation yes 89 (49.4%)
41 on September 28, 2021 by guest. Protected copyright. 42 no 91 (50.6%) 43 Electronic interaction control 44 yes 65 (36.1%) 45 no 115 (63.9%) 46 Electronic prescription yes, with thesaurus 62 (34.4%) 47 yes, but without thesaurus 24 (13.3%) 48 none 94 (52.2%) 49 Certification of the practice 50 yes 46 (25.6%) 51 none 134 (74.4%) 52 Staff meetings 53 yes, at least monthly 69 (38.3%) 54 yes, but less frequently 70 (38.9%) 55 none 41 (22.8%) 56 Quality circle participation 57 yes, at least monthly 134 (74.4%) 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 yes, but less frequently 23 (12.8%) 4 none 23 (12.8%) 5 6 Table 1. Characteristics of the reporting physicians in 2015 (144 practices, 180 physicians) 7 . 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Relevance All 5 6 7 less more 8 9 10 Number of cases 124 73 197 11 12 13 Patients age 69.2 ± 20.6 69.4 ± 21.2 69.3 ± 20.8 14 15 For peer review only 16 Patients gender, 17 % males 40.3 32.9 37.6 18 19 20 Physicians specialty 21 % pediatricians 1.6 1.4 1.5 22 23 24 Linguistic region, % 25 26 German 75.0 68.5 72.6 27 French 18.5 28.8 22.3 28 Italian 6.5 2.7 5.1 29 30 31 Physician to patient relationship, % 32 own family physician 86.3 78.1 83.2 33 http://bmjopen.bmj.com/ 34 urgency / holiday replacing 0.8 4.1 2.0 35 institution physician 11.3 17.8 13.7 36 37 other 1.6 0.0 1.0 38 39 40 Observer of the incident; % physician / practice staff 50.0 50.7 50.3 41 on September 28, 2021 by guest. Protected copyright. 42 patient / proxies 21.8 23.3 22.3 43 44 community nurse 1.6 4.1 2.5 45 institution (where patient lives) 15.3 16.4 15.7 46 hospital 0.8 1.4 1.0 47 48 other physicians 2.4 0.0 1.5 49 pharmacist 7.3 4.1 6.1 50 51 other 0.8 0.0 0.5 52 53 54 Table 2. General description of the cases. Differences between “less” and “more” categories are 55 not statistically significant. We calculated the variable “incident relevance” from the variables “dis 56 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 turbance” and “endangering” of the patients; if any of them was graded with “medium” or higher, 4 the variable relevance was set to “more”, otherwise to “less”. 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Item N percent 4 5 6 Severity of disturbance 7 8 no symptoms but pathological laboratory tests 15 8.0 9 light 44 23.4 10 11 moderate 22 11.7 12 severe 10 5.3 13 14 fatality 0 0.0 15 Subtotal (thisFor is the base peerof the next rows) review only 91 48.4 16 17 no symptoms, normal (or no) laboratory tests 97 51.6 18 19 Total 188 100.0 20 Missing data 9 n.a. 21 22 All patients 197 n.a. 23 24 25 Time until recovery 26 hours 26 28.5 27 28 days 41 45.1 29 weeks 15 16.5 30 31 not yet known or missing information 9 9.9 32 All patients with disturbances 91 100.0
33 http://bmjopen.bmj.com/ 34 35 Recovering 36 37 without sequels 78 85.8 38 with light to moderate sequels 2 2.2 39 40 with severe sequels or fatality* 5 5.4
41 not yet known or missing information 6 6.6 on September 28, 2021 by guest. Protected copyright. 42 43 All patients with disturbances 91 100.0 44 45 46 Treatment / surveillance 47 not needed 48 52.7 48 49 ambulatory care 33 36.3 50 hospital care** 7 7.7 51 52 missing information 3 3.3 53 All patients with disturbances 91 100.0 54 55 56 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Table 3. Disturbances after the incident. * In one case, there was a reduced kidney 4 5 function, the other cases remained unclear, and no fatalities were reported. ** Two cases 6 had to be surveilled in the emergency room, the hospital stays were: intoxications with 7 8 thiethylperazine, with fenoterol plus ipratropium bromide, with zolpidem, further a derailed 9 10 diabetes type 2 (after missed treatment with metformin), and a gastro intestinal hemor 11 rhage in a patient where antithrombotic treatment with rivaroxaban was not communicated 12 13 to the physician. 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Item Patient group OR (95% CI) OR (95% CI) 5 (crude) (adjusted) 6 7 Denominator Incidents 8 9 10 Number of observations 26,852 197 / / 11 12 Age (mean±SD), years 46.7 ± 27.5 69.3 ± 20.8 1.004 (1.001;1.006)**a 1.001 (0.996;1.005) 13 14 15 Gender, For peer review only 16 ● male 47.0 37.6 1 1 17 ● female 53.0 62.4 1.048 (0.916;1.197) 1.055 (0.878;1.196) 18 19 20 Care dependency, 21 d 22 number (%)* 23 ● none 16,335 (85.5%) 96 (51.6%) 1 1 24 ● yes, by proxies 954 (5.0%) 12 (6.5%) 1.121 (0.789;1.594) 0.979 (0.674;1.423) 25 a ● yes, by community nurse 723 (3.8%) 22 (11.8%) 1.458 (1.025;2.073) 1.201 (0.821;1.758) 26 1,099 (5.8%) 56 (30.1%) 1.802 (1.399;2.323)c 1.528 (1.141;2.046) a 27 ● yes, by institution 28 29 30 c Number of conditions 2 (0;4) 5 (3;7) 1.052 (1.029;1.075)** 1.030 (0.994;1.067)** 31 32 (median, IQR)
33 http://bmjopen.bmj.com/ 34 Number of chronic active 1 (0;4) 6 (3;9) 1.052 (1.030;1.074)**c 1.030 (0.995;1.067)** 35 36 treatments (median, IQR) 37 38 Evans’ Index 3 (0;8) 11 (6;17) 1.009 (1.005;1.013)** c n.a.*** 39 40 (median, IQR)
41 on September 28, 2021 by guest. Protected copyright. 42 Thurgau Morbidity Index d 43 d 44 value (%) 45 ● 0 8,463 (31.5%) 24 (12.2%) 1 1 46 ● 1 3,611 (13.4%) 8 (4.1%) 0.989 (0.787;1.242) 0.908 (0.694;1.190) 47 ● 2 4,102 (15.3%) 23 (11.7%) 1.049 (0.847;1.300) 0.898 (0.685;1.169) 48 3,877 (14.4%) 39 (19.8%) 1.131 (0.914;1.399) 0.830 (0.611;1.127) 49 ● 3 a 50 ● 4 2,119 (7.9%) 39 (19.8%) 1.292 (1.004;1.662) 0.901 (0.643;1.265) b 51 ● 5 1,539 (5.7%) 38 (19.3%) 1.420 (1.078;1.868) 0.823 (0.547;1.239) 52 709 (2.6%) 26 (13.2%) 1.680 (1.178;2.396)c 0.866 (0.523;1.436) 53 ● 6 18 0 54 missing values 55 56 57 58 28 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Table 4. Possible risk factors for incident as compared to a denominator analysis during 4 calendar weeks 11 and 12 [Gnädinger M 2016]. * Because of question ambiguity this analysis 5 6 was restricted to adult patients (age > 19 years); this led to 19,812 valid observations in the de 7 nominator and 183 in the incident groups. ** Per one conditions, medication, year or index point. 8 9 *** Because Evans’ Index is a composite of condition and medication numbers; it was not included 10 in the multiple regression analysis. Significance levels: a p<0.05, b p<0.01, c p<0.001. d Because 11 12 GENLIN procedure was not able to process ordinal scaled variables, correlations between study 13 14 group and TMI or care dependency were tested with Spearman’s rho: the correlation coefficients 15 were +0.075For or +0.094, respectively,peer p<0.001. review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 29 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 ATC Class All incidents Incidents with prob- Relative Risk Swiss 2015 sales* 4 able or definite (percentages left (number of packages) 5 relationship with by right column) 6 7 medication 8 9 A Alimentary tract and metabolism 28 (14.6%) 6 (15.8%) 1.06 31,455,252 (14.9%) 10 11 12 B Blood and blood forming organs 23 (12.0%) 7 (18.4%) 7.08 5,507,624 (2.6%) 13 14 C Cardiovascular system 44 (22.9%) 1 (2.6%) 0.35 16,027,143 (7.5%) 15 For peer review only 16 17 D Dermatologics 1 (0.5%) 0 (0.0%) 0.00 17,314,810 (8.2%) 18 19 G Genito urinary system and sex 1 (0.5%) 0 (0.0%) 0.00 7,936,641 (3.7%) 20 hormones 21 22 23 H Systemic hormonal preparations 7 (3.6%) 1 (2.6%) 2.00 2,875,760 (1.3%) 24 (excluding sex hormones and 25 insulins) 26 27 28 J Anti infectives for systemic use 23 (12.0%) 6 (15.8%) 3.95 8,444,623 (4.0%) 29
30 K Infusion liquids 0 (0.0%) 0 (0.0%) 0.00 24,158,749 (11.5%) 31 32
33 L Anti neoplastic and immunomodu 5 (2.6%) 1 (2.6%) 2.89 1,934,950 (0.9%) http://bmjopen.bmj.com/ 34 lating agents 35 36 37 M Musculo skeletal system 4 (2.1%) 2 (5.3%) 0.76 14,787,413 (7.0%) 38 39 N Nervous system 43 (22.4%) 10 (26.3%) 1.30 42,690,195 (20.2%) 40
41 on September 28, 2021 by guest. Protected copyright. 42 P Anti parasitic products, insecti 1 (0.5%) 0 (0.0%) 0.00 462,559 (0.2%) 43 cides and repellents 44 45 R Respiratory system 7 (3.6%) 3 (7.9%) 0.57 28,837,468 (13.7%) 46 47 48 S Sensory organs 2 (1.0%) 0 (0.0%) 0.00 7,658,311 (3.6%) 49 50 51 T Diagnostic use 0 (0.0%) 0 (0.0%) 0.00 43,184 (0.0%) 52 53 V Various 3 (1.6%) 1 (2.6%) 6.50 855,707 (0.4%) 54 55 56 Total 192 (100.0%) 38 (100%) 1.0 210,990,389 (100.0%) 57 58 30 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 does not apply 5 0 n.a. n.a. 4 5 6 Table 5. ACT-Groups of suspected medications. Relative risk of drugs with probable or definite 7 relationship with the incident as compared to sales proportions. * Information by Interpharma Swit 8 9 zerland (Appendix F). 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 31 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Appendices 4 5 A) Incident questionnaire 6 B) Initial physicians’ questionnaire 7 C) Final physicians’ questionnaire 8 D) Electronic figures and tables 9 10 E) STROBE statement 11 F) Swiss drug sales 2015 by ATC groups 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 32 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 Appendix A Incident questionnaire 7 8 9 Medication Incidents in Primary Care (MIPC) 10 (Version 1.0, January 25th 2015) (Internet version not available in English) 11 12 13 14 15 Incident reportingFor peer form review only 16 17 18 19 Administrative Information 20 21 22 23 24 1. Sentinella identification number: 2. Week of reporting: 25 26 27 28 29 30 The Patient 31 32
33 http://bmjopen.bmj.com/ 34 35 3. Year of birth: 4. Gender: m f 36 37 38 39 5. What was your relationship to the patient when the incident happened? Were you the 40
41 on September 28, 2021 by guest. Protected copyright. 42 family physician emergency / substitute physician institution physician other 43 if other what kind . 44 45 46 47 6. What is the patient’s living situation? 48 49 50 with partner / family alone institution unknown 51 52 53 54 7. Are there social problems? 55 56 yes no unknown 57 58 33 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 8. Is the patient demented or otherwise mentally handicapped? 6 7 8 yes no unknown 9 10 11 12 9. Does he suffer from psychological problems? 13 14 yes no unknown 15 For peer review only 16 17 18 19 10. Does he take psychotropic drugs (esp. antidepressants, neuroleptics, benzodiazepines, 20 opiates)? 21 22 yes no unknown 23 24 25 26 11. Are there linguistic problems? 27 28 yes no unknown 29 30
31 32 12. Does he or she smoke? 33 http://bmjopen.bmj.com/ 34 35 yes no unknown 36 37 38 39 13. Is there substance abuse (other than nicotine)? 40
41 yes no unknown , if yes, what substance? .. on September 28, 2021 by guest. Protected copyright. 42 43 44 45 14. Does the patient have uncorrected / uncorrectable visual impairment? 46 47 48 yes no unknown 49 50 51 52 15. Does the patient have uncorrected / uncorrectable hearing impairment? 53 54 55 yes no unknown 56 57 58 34 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 16. Does the patient have uncorrected / uncorrectable mobility impairment? 4 yes no unknown 5 6 7 8 2 9 17. Is there renal insufficiency (GFR: <60 ml/min/1.73 m )? 10 11 yes no unknown 12 13 14 15 For peer review only 16 18. Is there hepatic insufficiency or liver cirrhosis? 17 18 yes no unknown 19 20 21 22 23 19. Was the patient hospitalized in the past 12 months? 24 yes no unknown 25 26 27 28 20. Is the patient taken care of by others? (only one answer permitted)? 29 30 yes, family / proxies yes, community nurse yes, institution no unknown 31 32
33 http://bmjopen.bmj.com/ 34 21. Number of regularly applied active substances (including non-daily applied ones, see 35 guidelines)? 36 37 38 unknown 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 22. Number of chronic conditions (see guidelines)? 43 44 45 unknown 46 47 48 49 23. Scale value of „Thurgau Morbidity Index” (chronic part, see guidelines)? 50 51 unknown 52 53 54 55 56 57 58 35 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Details of the incident 4 5 24. Please, give a short description of the incident (in block letters): 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 25. Who observed the incident? (multiple answers possible)? 22 Physician / staff patient / relatives community nurse home / institution hospi 23 tal 24 25 other physicians pharmacist other unknown 26 for „other” please specify: 27 28 29 30 31 26. What happened (multiple answers possible)? 32 dosage too high 33 http://bmjopen.bmj.com/ 34 dosage too low 35 application too short 36 37 application too long 38 wrong administration route 39 40 wrong medication
41 indicated medication not received on September 28, 2021 by guest. Protected copyright. 42 43 expired / defective medication 44 problems with insurance reimbursement 45 46 47 unclear / undefined 48 other (please specify): 49 50 51 52 27. Please state the trade name of the medication used in the incident: 53 54 .. 55 56 57 58 36 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 28. Please note other medication names, presuming they are relevant to the case. 4 5 6 7 8 9 none unknown 10 11 12 29. How would you judge the degree of hazard to the patient during the incident? 13 14 15 mild mediumFor severepeer none review does not apply only unknown 16 17 18 19 30. How intense was the impairment (as judged by the patient)? 20 21 mild medium severe fatal no symptoms, but pathological lab values 22 23 no impairment does not apply unknown 24 If there wasn’t any impairment, please skip to question 34. 25 26 27 28 31. How long did the impairment last? 29 30 hours days weeks longer unknown 31 32
33 http://bmjopen.bmj.com/ 34 32. How was the recovery? 35 without residues with mild residues with severe residues / fatal unknown 36 37 38 39 40 33. Which organ system was affected (multiple answers possible)?
41 cardiovascular on September 28, 2021 by guest. Protected copyright. 42 43 central nervous 44 gastro enteral 45 46 kidneys 47 liver 48 49 lung 50 skin 51 52 other, please specify . 53 54 55 56 57 58 37 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 34. Did you have to apply a specific surveillance or treatment for the incident? 4 yes, ambulatory yes, hospital no unknown 5 If yes, please specify which? .. 6 7 8 9 10 35. What factors contributed causally to the emergence of the incident (multiple answers 11 possible)? 12 13 off duty hours 14 15 communicationFor failure within peer practice review only 16 17 generic substitution by pharmacist 18 19 hand written prescription incorrectly interpreted 20 21 conflicting multiple prescriptions 22 lack of alertness within practice 23 24 lack of documentation 25 insufficient patient instruction 26 27 lack of aids (e.g. Dosette®) 28 lack of cooperation by patient / relatives 29 misleading package leaflet information 30 31 patient’s internet search 32 administrative problems 33 http://bmjopen.bmj.com/ 34 manufacturer related (defective medication) 35 distributer related (out of stock) 36 37 lack of maintenance (e.g. first aid kit) 38 other, please specify .. 39 40 unknown
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 36. Was there an interface problem? If yes, which (multiple answers possible)? 45 46 47 yes, with hospital 48 yes, with institution 49 50 yes, with community nurse 51 yes, with pharmacist 52 53 54 yes, with specialist physician 55 yes, with other, please specify? 56 57 58 38 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 no 4 5 If no, please skip to question 38. 6 7 8 37. Was there an explicit comparison of prescription lists with the institution / person? 9 10 11 yes, verbal / by phone yes, written / by fax no unknown 12 13 14 38/39. Was the patient informed about the incident? 15 For peer review only 16 yes, by myself / practice staff 17 18 no, because he was not able to understand the message (children, demented) 19 no, because the problem was solved and communication would have impaired confidence 20 21 no, because the patient had moved or was deceased 22 no, this was not needed because patient / relatives themselves had observed the incident 23 24 no, because others had already informed him 25 26 no, because: . 27 unknown 28 29 30 31 If yes, what was the patient’s reaction? 32
33 http://bmjopen.bmj.com/ 34 40. What did you do as a result to prevent similar incidents in the future (multiple answers 35 36 possible)? 37 change standard operations procedures 38 better instruction of patients 39 40 communication with institution(s)
41 notification of manufacturer on September 28, 2021 by guest. Protected copyright. 42 43 notification of liability insurer 44 notification of drug authority („yellow leaflet“) 45 46 notification of the “critical incident reporting system” 47 other, please specify? .. 48 49 nothing 50 51 52 41. Who or what was ultimately responsible for the occurrence of the incident? 53 . 54 55 56 57 58 39 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 42. In the given situation, do you think one could have anticipated the event? 4 yes no 5 6 7 8 43. Did you already report an identical or very similar incident to this study? 9 10 yes no 11 12 13 14 44. Please make any suggestions about the kind of measures that could be taken to generally 15 reduce the frequencyFor of suchpeer events (in blockreview letters): only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 Please keep a copy of this questionnaire in the patient files. Thank you for filling it out! 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 40 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Appendix B Initial questionnaire 4 5 6 Medication incidents in primary care (MIPC) 7 8 Initial reporting (Version 1.0, January 25th 2015) 9 Internet version: not available in English 10 11 1. Sentinella identification number: . 12 13 14 2. Number of physicians within your practice: . 15 For peer review only 16 17 3. Number of them who report to Sentinella . 18 19 20 21 4. Your weekly workload, h/w: <16 , 16 30 , >30 22 23 24 5. Number of hours within your practice (%): <50% , ≥50% 25 26 6. Approximate proportion of medication prescribed (as compared to directly delivered drugs): 27 <33% , 33 66% , >66% 28 29 7. Do you have an X ray (machine?)? yes no 30 31 32 8. Do you have an ECG? yes no
33 http://bmjopen.bmj.com/ 34 35 9. Do you have an ultrasound? yes no 36 37 38 10. Do you have an electronic system for controlling electronic drug interaction? 39 yes no 40
41 on September 28, 2021 by guest. Protected copyright. 42 11. Do you have electronic patient history documentation? 43 yes no 44 45 46 12. Do you prescribe electronically? 47 yes, with a medication thesaurus yes, but without one (use of a typewriter) no 48 49 50 13. Is your practice certified (e.g. EQUAM)? yes no 51 52 53 14. Do you regularly schedule team sessions? 54 Yes, at least monthly , yes, but less frequently , no 55 56 57 58 41 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 15. Do you attend quality circle sessions (in accordance with "Hausärzte Schweiz")? 5 yes, regularly , yes, now and then , no 6 7 8 16. Did you complete a special education (e.g. manual or psychosomatic medicine), or do you 9 have special interests (e.g. toxic maniac patients)? yes , no 10 if yes, please specify : . 11 12 13 17. Are you contracted by an institution? yes no 14 if yes, please specify (prison, home etc.): 15 For peer review only 16 18. If yes, does this institution have specific problems with medication? 17 yes , no , if yes, please specify: 18 19 19. Are you involved in other special activities (teaching, research, insurance doctor)? 20 yes no 21 If yes, please specify the kind of activity: .. 22 23 Thank you very much! 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 42 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 43 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Appendix C Final questionnaire 4 5 6 Medication Incidents in Primary Care (MIPC) 7 th 8 Final questionnaire (Version 3.0 / December 14 2015) 9 10 Online version: not available in English 11 12 Dear colleagues, 13 14 Thank you for your dedicated support of our study on medication incidents in 2015, regardless of 15 whether you Forreported much peer of them or not.review We would like to ask only some final questions which will 16 help us put the other information you sent us in the right context. 17 18 19 1. Sentinella identification number: 20 21 2. Did you not report medication incidents that you had noticed during the last year (e.g. because 22 of lack of time)? 23 24 never or almost never , yes, but rarely , yes, frequently always or almost always 25 26 If this was frequently the case, please explain why: .. 27 28 3. Did your practice participate in the fortnight morbidity denominator study in March 2015 (calen 29 dar weeks 11 and 12)? 30 yes, fully , yes, but only partly (by omission of certain variables) , no 31 32 If you did not fully participate or not at all, what was the reason?
33 http://bmjopen.bmj.com/ 34 35 If you did not participate in the denominator study, please continue with question 12. 36 37 4. How big was your effort for coding the morbidity variables of the denominator study? 38 manageable , rather big , too much , impossible 39 40 Did you have any difficulties when coding the morbidity variables?
41 on September 28, 2021 by guest. Protected copyright.
42 5. Hospitalisation during the previous 12 months? none , a little , considerable , severe 43 44 6. Care dependency? none , a little , considerable , severe 45 46 7. Number of medications? none , a little , considerable , severe 47 48 8. Number of conditions? none , a little , considerable , severe 49 50 9. Thurgau Morbidity Index? none , a little , considerable , severe 51 52 10. Repeat consultation during the fortnight? none , a little , considerable , severe 53 54 If you named considerable to severe difficulties in questions 5 to 10, please list them in item 13. 55 56 57 58 43 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 44 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 11. How long did it take for you and your practice nurse together for coding all of the variables of 4 one patient (Dr. Gnädinger needed less than 3 minutes)? 5 . minutes 6 7 12. Would you be willing to be interviewed for a focus group on the subject of medication safety? 8 9 yes , no 10 11 13. Other comments: 12 13 14 15 For peer review only 16 17 18 19 20 21 Thank you very much! 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 44 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 45 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Appendix D: electronic tables and figures 4 5 Figure e1: Distribution of the incident notifications over the year 6 Table e1: What went wrong with the incidents 7 Table e2: Organ system involved 8 9 Table e3: Causes of the incident 10 Table e4: Patient sided possible risk factors 11 Table e5: Reactions to the incident 12 Table e6: Proposals to avoid further incidents 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 45 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 46 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 Figure e1. Distribution of the incident notifications over the year. Because only the 38 39 week has been reported, the assignment to the month is approximate. 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 46 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 47 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Type of error N* Percent*, % 5 6 7 Dosing too high 41 (12) 20.9 (6.1) 8 9 10 Dosing too low 21 (6) 10.7 (3.0) 11 12 13 Dosing too short 1 (2) 0.5 (1.0) 14 15 For peer review only 16 Dosing too long 5 (4) 2.6 (2.0) 17 18 19 Wrong way of administration 1 (1) 0.5 (0.5) 20 21 22 Wrong medication applied 56 (12) 28.6 (6.1) 23 24 25 Necessary medication not applied 12 (9) 6.1 (4.6) 26 27 28 Defective or expired medication applied 1 (1) 0.5 (0.5) 29 30 31 Problems with insurance reimbursing 1 (0) 0.5 (0.0) 32
33 http://bmjopen.bmj.com/ 34 Other problem** 31 (6) 15.8 (3.0) 35 36 37 Unknown 1 (0) 0.5 (0.0) 38 39 40 Multiple naming 26 (n.a.) 13.3 (n.a.)
41 on September 28, 2021 by guest. Protected copyright. 42 43 Total 197 (53) 100.0 (26.9) 44 45 46 47 Table e1. What went wrong with the incidents. * Parenthesis denotes the additional naming 48 within the category “multiple naming”. ** Naming within the category “other” were: confusion of sim 49 ilar trade names (2), dosing error (6), erroneous package size (1), confusion of similar looking 50 preparations (1), error when controlling blood levels (2), double dosing (4), missed discontinuation 51 of medication (1), necessary treatment not applied (1), wrong vaccine applied (4), missed re 52 uptake of anticoagulation after operation (1), contra indication overlooked (2), drug drug interaction 53 overlooked (3), known intolerance overlooked (3), necessary monitoring missed (2), transgression 54 of maximal drug allowance (2), delivery of incomplete package (1), confusion by contradictory 55 medication lists (3), incorrect handling of administering device (1), uncontrolled taking of “nature 56 medicines” (1), unreliable compliance with medication plan (2), two medications of the same drug 57 58 47 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 48 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 class (3), forbidden bisection of pills (2), confusion of similar named patients (1), intake of medica 4 tion of the neighbor resident in a home (1), other (9). 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 48 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 49 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Organ system N* Percent*, % 4 5 6 Cardiovascular 14 (10) 15.4 (11.0) 7 8 9 Central nervous system 23 (10) 25.3 (11.0) 10 11 12 Gastro intestinal 5 (3) 5.5 (3.3) 13 14 15 Kidneys For peer review only2 (2) 2.2 (2.2) 16 17 18 Liver 2 (0) 2.2 (2.2) 19 20 21 Lung 2 (1) 2.2 (1.1) 22 23 24 Skin 6 (0) 6.6 (0.0) 25 26 27 Other** 24 (6) 26.4 (6.6) 28 29 30 Multiple naming 13 (n.a.) 14.3 (n.a.) 31 32
33 Total 91 (32) 100 (35.2) http://bmjopen.bmj.com/ 34 35 36 Question does not apply 106 / 37 38 39 40 Table e2. Organ system involved. * Parenthesis denotes the additional naming within the catego
41 on September 28, 2021 by guest. Protected copyright. 42 ry “multiple naming”. ** Naming within the category “other” were: endocrine system (9), musculo 43 skeletal (8), ear nose throat (2), psychic (1), other (9). 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 49 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 50 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Causes of the incident N* Percent*;% 4 5 6 Out of hours 2 (3) 1.0 (1.5) 7 8 Communication problems within staff 10 (11) 5.1 (5.6) 9 10 11 Generic substitution of original trade medication by pharmacist 3 (3) 1.5 (1.5) 12 13 Difficulties when reading hand written prescription 1 (1) 0.5 (0.5) 14 15 For peer review only 16 Multiple conflicting prescriptions 5 (8) 2.5 (4.1) 17 18 19 Lacking alertness of physician or practice staff 50 (17) 25.4 (8.6) 20 21 Insufficient documentation 3 (11) 1.5 (5.6) 22 23 24 Insufficient patient instruction 7 (17) 3.6 (8.6) 25 26 Lacking cooperation of patient / proxies 6 (17) 3.0 (8.6) 27 28 29 Confusion by reading package leaflet 1 (0) 0.5 (0.0) 30 31 Confusion after “Googleing” 1 (0) 0.5 (0.0) 32
33 http://bmjopen.bmj.com/ 34 Administrative problems 2 (3) 1.0 (1.5) 35 36 Defective medication as caused by manufacturer 3 (0) 1.5 (0.0) 37 38 39 Lacking maintenance (e.g. emergency case) 1 (0) 0.5 (0.0) 40
41 on September 28, 2021 by guest. Protected copyright. 42 Lacking use of treatment aids (e.g. Dosett) 0 (6) 0.0 (3.0) 43 44 Other source of trouble** 54 (18) 27.2 (9.1) 45 46 47 Unknown 3 (0) 1.5 (0.0) 48 49 Multiple naming 45 (n.a.) 23.1 (n.a.) 50 51 52 Total 197 (115) 100.0 (58.4) 53 54 Table e3. Causes of the incident. * Parenthesis denotes the additional naming within the category “multiple naming”. ** Naming within 55 56 the category “other” were: Erroneous delivery in pharmacy (5), treatment delayed/hampered by patient (2), transcription error (2), 57 missed discontinuation of an ongoing treatment (2), communication problem within helpers’ network (16), similar trade names (2), inter 58 50 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 51 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 face problems with hospital (4), reading error of patient / proxies (3), erroneous execution of a medical prescription (14), missed follow 4 up control (9), difficult handling of a preparation (2), disregard of possible interactions (2), transgression of competence by care workers 5 (5), incomplete information by patient / proxies (8), stress / lack of time (12), other 25. 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 51 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 52 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Item Relevance All 4 5 (number of missing observations in groups “less” / n (%) 6 “more”) less more 7 8 n (%) n (%) 9 10 Number of patients 124 (100.0) 73 (100.0) 197 (100.0) 11 12 13 Living situation (1/0) 14 15 together withFor mate / family peer review61 (49.6) only31 (42.5) 92 (46.9) 16 alone 28 (22.8) 23 (31.5) 51 (26.0) 17 institution 33 (27.6) 19 (26.0) 53 (27.0) 18 19 20 Social problems (2/4) 21 22 yes 22 (18.1) 18 (26.1) 49 (20.9) 23 24 Dementia or mental illness (2/2) 25 26 yes 31 (25.4) 18 (25.4) 49 (25.4) 27 28 29 30 Psychiatric problems (1/2) 31 yes 28 (22.8) 28 (38.4)* 56 (28.9) 32
33 http://bmjopen.bmj.com/ 34 Treatment with psychotropic drugs (0/0) 35 36 yes 50 (40.3) 39 (53.4) 89 (44.2) 37 38 39 Linguistic problems (0/1) 40 yes 11 (8.9) 5 (6.8) 16 (8.2)
41 on September 28, 2021 by guest. Protected copyright. 42 43 Smoking (7/4) 44 yes 11 (9.4) 8 (11.6) 19 (10.2) 45 46 47 Substance abuse other than nicotine (2/1) 48 yes 3 (2.5) 5 (6.9) 8 (4.1) 49 50 51 Visual blurring (6/6) 52 53 yes 4 (3.4) 4 (6.0) 8 (4.3) 54 55 Hearing problems (3/5) 56 57 58 52 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 53 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 yes 8 (6.6) 4 (5.9) 12 (6.3) 4 5 6 Gait problems (2/2) 7 yes 40 (32.8) 18 (25.4) 58 (30.1) 8 9 2 10 Renal insufficiency (GFR<60 ml/min*1.73m ) (6/2) 11 yes 24 (20.3) 21 (29.6) 45 (23.8) 12 13 14 15 Liver cirrhosisFor of other hepatic peer function problem review (3/1) only 16 17 yes 3 (2.5) 3 (4.2) 6 (3.1) 18 19 20 Table e4. Patient-sided possible risk factors. * p=0.021 “more” vs. “less” by chi square testing. 21 We calculated the variable “incident relevance” from the variables “disturbance” and “endangering” 22 23 of the patients; if any of them was graded with “medium” or higher, the variable relevance was set 24 to “more”, otherwise to “less”. 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 53 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 54 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Reactions to the incident N* Percent*; % 4 5 6 Changing standard operating procedures of the practice 14 (3) 7.2 (1.5) 7 8 9 Better instruction of patients 23 (4) 11.7 (2.0) 10 11 12 Communication with other institutions 31 (3) 15.7 (1.5) 13 14 15 Notifying manufacturerFor peer review only1 (2) 0.5 (1.0) 16 17 Reporting the incident to the critical incident reporting system 10 (6) 5.1 (3.0) 18 19 20 No reaction at all 55 (n.a.) 27.9 (n.a) 21 22 23 Other type of reactions to the incident** 49 (9) 24.9 (4.6) 24 25 26 Missing information 1 (0) 0.5 (0.0) 27 28 29 Multiple naming 13 (n.a.) 6.6 (n.a.) 30 31 32 Total 197 (27) 100.0 (13.7)
33 http://bmjopen.bmj.com/ 34 35 Table e5. Reactions to the incident. * Parenthesis denotes the additional naming within the category “mul 36 tiple naming”. ** Naming within the category “other” were: Hire more workforce (1), arrangements with other 37 physicians (2), with pharmacist (4), with community nurse (4), with institution (12), with practice nurse (6), 38 39 with specialist (1), with patient (3), with supplier (1), sending a new medication plan (1), having regular staff 40 meetings (1), remove Digoxin 0.25 from assortment because of safety reasons (1), to not perform vaccina
41 on September 28, 2021 by guest. Protected copyright. 42 tions in absence of vaccination card (2), to clarify intolerances (2), to clarify interactions (1), to hand out an 43 allergy card (1), to actualize patient records (3), to apply for insurance cost credit (1), to organize follow up 44 controls (20), other (5). 45 46 47 48 49 50 51 52 53 54 55 56 57 58 54 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 55 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Proposal N* Percent*;% 4 5 6 Cross check of medication lists 6 (14) 4.8 (11.2) 7 Patient instructions 2 (3) 1.6 (2.4) 8 9 Reduction of time pressure / stress 3 (6) 2.4 (4.8) 10 11 To observe adverse drug reactions also with “nature 1 (0) 0.8 (0.0) 12 products” 13 14 Four eyes check when dispensing medication 2 (4) 1.6 (3.2) 15 For peer review only 16 No medication without prescription 3 (6) 2.4 (4.8) 17 18 To critically audit polymedication 3 (2) 2.4 (1.6) 19 20 To avoid similarly looking or named medication in drug 2 (3) 1.6 (2.4) 21 master 22 23 To demand medication plans to be brought to the consul 0 (4) 0.0 (3.2) 24 tation 25 26 To provide medication plans routinely 2 (8) 1.6 (6.4) 27 28 In deep checking before delivering “new” medication to 4 (1) 3.2 (0.8) 29 the patient 30 31 To provide patients with allergy / intolerance cards 2 (1) 1.6 (0.8) 32 To let patients themselves write their medication card 0 (3) 0.0 (2.4) 33 http://bmjopen.bmj.com/ 34 (and controlling afterwards) 35 36 To instruct patients to come into the practice immediately 3 (0) 2.4 (0.0) 37 after hospitalization 38 39 To broach regularly “medication safety” on staff meetings 5 (6) 4.0 (4.8) 40 To provide information in patients’ native language 0 (2) 0.0 (1.6)
41 on September 28, 2021 by guest. Protected copyright. 42 43 To share important information about patients with prac 0 (2) 0.0 (1.6) 44 tice nurse 45 46 To improve information flow 11 (16) 8.8 (10.4) 47 48 To organize follow up controls 10 (28) 8.0 (22.4) 49 50 To wait for laboratory results before prescribing 1 (0) 0.8 (0.0) 51 52 Other** 12 (9) 9.6 (7.2) 53 54 Multiple answering 53 (n.a.) 42.4 (n.a.) 55 56 Total 125 (118) 100.0 (94.4) 57 58 55 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 56 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 None 72 (n.a.) (n.a.) 4 5 6 7 Table e6. Proposals to avoid further incidents. * Parenthesis denotes the additional nam 8 ing within the category “multiple naming”. ** Among naming within the category “other” were: timely 9 10 involving of community nurse, stopping of self medication in multimorbid or poly medicated pa 11 tients, to clearly labelling desensitization suspensions, to separating adult and pediatric vaccines, 12 13 to implementing timely new guidelines, to refusing to be in care of a patient with bad compliance, to 14 clearly separating and labelling of medication prepared for different patients, to advice patients for 15 For peer review only 16 separating short and long acting insulins at their home, to providing short time prescriptions with a 17 clear expiration date, to allowing only adequately educated people to align and deliver medication 18 19 in homes for the elderly and that delivering may be done by the same person who did the prepar 20 ing of medication boxes, to avoiding similar looking medication boxes being muddled up by pa 21 22 tients at the refectory room. 23 24
25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 56 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 57 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Appendix E 4 5 cross-sectional studies 6 STROBE Statement—Checklist of items that should be included in reports of 7 Item 8 No Recommendation 9 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or the 10 abstract [yes] 11 12 (b) Provide in the abstract an informative and balanced summary of what 13 was done and what was found [yes] 14 15 Introduction For peer review only 16 Background/rationale 2 Explain the scientific background and rationale for the investigation being 17 reported [yes] 18 19 Objectives 3 State specific objectives, [yes] including any prespecified hypotheses [no] 20 21 Methods 22 Study design 4 Present key elements of study design early in the paper [yes] 23 24 Setting 5 Describe the setting, locations, and relevant dates, including periods of 25 recruitment, exposure, follow up, and data collection [yes] 26 27 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of 28 participants [yes] 29 30 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, 31 and effect modifiers. [yes] Give diagnostic criteria, if applicable [no] 32
33 Data sources/ 8* For each variable of interest, give sources of data and details of methods http://bmjopen.bmj.com/ 34 measurement of assessment (measurement). Describe comparability of assessment 35 methods if there is more than one group [yes] 36 37 Bias 9 Describe any efforts to address potential sources of bias [yes] 38 39 Study size 10 Explain how the study size was arrived at [yes] 40 Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If appli
41 on September 28, 2021 by guest. Protected copyright. cable, describe which groupings were chosen and why [yes] 42 43 Statistical methods 12 (a) Describe all statistical methods, including those used to control for con 44 founding [yes] 45 46 (b) Describe any methods used to examine subgroups and interactions 47 [yes] 48 49 (c) Explain how missing data were addressed [yes] 50 51 (d) If applicable, describe analytical methods taking account of sampling 52 strategy [no] 53 54 (e) Describe any sensitivity analyses [no] 55 56 Results 57 58 57 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 58 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers 4 potentially eligible, examined for eligibility, confirmed eligible, included in 5 the study, completing follow up, and analysed [no] 6 7 (b) Give reasons for non participation at each stage [yes] 8 (c) Consider use of a flow diagram [yes] 9 10 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, 11 social) and information on exposures and potential confounders [yes] 12 13 (b) Indicate number of participants with missing data for each variable of 14 interest [yes] 15 For peer review only 16 Outcome data 15* Report numbers of outcome events or summary measures [yes] 17 18 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder adjusted es 19 timates and their precision (eg, 95% confidence interval). Make clear which 20 confounders were adjusted for and why they were included [yes] 21 22 (b) Report category boundaries when continuous variables were catego 23 rized [yes] 24 25 (c) If relevant, consider translating estimates of relative risk into absolute 26 risk for a meaningful time period [no] 27 28 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and sensitivity analyses [yes] 29 30 31 Discussion 32 Key results 18 Summarise key results with reference to study objectives [yes]
33 http://bmjopen.bmj.com/ 34 Limitations 19 Discuss limitations of the study, taking into account sources of potential 35 bias or imprecision. Discuss both direction and magnitude of any potential 36 bias [yes] 37 38 Interpretation 20 Give a cautious overall interpretation of results considering objectives, 39 limitations, multiplicity of analyses, results from similar studies, and other relevant evidence [yes] 40
41 on September 28, 2021 by guest. Protected copyright. Generalisability 21 Discuss the generalisability (external validity) of the study results [yes] 42 43 44 Other information 45 Funding 22 Give the source of funding and the role of the funders for the present study 46 and, if applicable, for the original study on which the present article is 47 based [yes] 48 49 50 *Give information separately for exposed and unexposed groups. 51 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 52 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 53 54 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 55 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is avail- 56 able at www.strobe-statement.org 57 58 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 59 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Appendix F 4 5 6 Swiss Pharmaceutical Market: Number of packaged sold by ATC-groups, 2015 7 8 9 10 2015 11 number of 12 packaged sold 13 Total 210,992,389 14 A Alimentäres System und Stoffwechsel 31,455,252 15 For peer review only 16 A01 Stomatologika, Medizinische Präparate zur Mund und Zahnpflege 1,473,520 17 A02 Antacida Ulcustherapheutika, Antiflatulentia 6,051,251 18 A02A Antacida,Antiflatulentia 1,500,822 19 A02B Ulcustherapeutika 4,426,496 20 A02C Sonstige Magentherapeutika 123,933 21 A03 Produkte gegen funktionelle Magen Darm Störungen 2,087,299 22 A03A Antispasmodika+Anticholinergika rein 595,861 23 A03C Antispasmodika/Ataraktika Kombinationen 25,411 24 25 A03E Antispasmodika, sonstige Kombinationen 222,102 26 A03F Gastroprokinetika 1,233,461 27 A03G Modulatoren der gastrointestinalen Sensomotorik 10,464 28 A04 Antiemetica und Antinausea 920,841 29 A05 Cholagoga und Leberschutz Mittel 337,970 30 A05A Gallentherapeutica und Cholagoga 254,744 31 A05B Leberschutzpräparate 83,226 32 A06 Mittel gegen Verstopfung und Darmreinigung 4,583,118 33 http://bmjopen.bmj.com/ 34 A06A Mittel Gegen Verstopfung 4,329,037 35 A06B Darmreinigungsmittel 254,081 36 A07 Antidiarrhoica, Elektrolytzufuhr und intestinale Antiphlogistica 3,005,035 37 A07B Intestinale Adsorbierende Antidiarrhoica 228,076 38 A07E Intestinale Antiphlogistica 222,680 39 A07F Antidiarrhoica Micro Organismen 1,340,432 40 A07G Orale Elektrolyt Zufuhr 107,030
41 on September 28, 2021 by guest. Protected copyright. 42 A07H Motilitätshemmer 1,095,216 43 A07X Übrige Antidiarrhoica inkl. A07A Antidiarrhoica intestinale Antiinfectiva 11,601 44 A08 Antiadiposita excl. Diätetica 51,654 45 A09 Digestiva und Enzyme 804,667 46 A10 Antidiabetica 3,376,580 47 A10C Humaninsulin und Analoga 1,078,495 48 A10H Sulfonylharnstoff Antidiabetika 321,219 49 A10J Biguanid Antidiabetika 1,217,578 50 51 A10K Glitazone Antidiabetika 25,305 52 A10M Glinide Antidiabetika 38,782 53 A10N DPP IV Inhibitor Antidiabetika 483,559 54 A10P SGLT2 Hemmer Antidiabetika 45,927 55 A10S GLP 1 Agonisten Antidiabetika 156,656 56 A10X Übrige Antidiabetika inkl. Insulin tierischen Ursprungs und 8,616 57 Alphaglukosidaseinhibitor Antidiabetika 58 59 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 60 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 A11 Vitamine 5,001,408 4 A11A Multivitamine mit Mineralstoffen 755,243 5 A11B Multivitamine ohne Mineralstoffe 17,507 6 A11C Vitamin A+D inkl. einfache Kombinationen 2,618,699 7 A11D Vitamin B1 und Kombinationen 170,884 8 A11E Vitamin B Komplex 739,860 9 10 A11F Vitamin B 12 rein 123,445 11 A11G Vitamin C, inkl. Kombinationen mit Mineralstoffen 353,644 12 A11X Sonstige Vitamine 222,126 13 A12 Mineralverbindungen 3,490,891 14 A12A Mineralverbindung Calcium 1,254,334 15 A12B MineralverbindungFor peer Kalium review only 200,438 16 A12C Sonstige Mineralverbindungen 2,036,119 17 18 A13 Tonica und Roborantia 229,735 19 A16 Übrige Präparate des alimentären und Stoffwechsel Systems 34,295 20 B Blut und Blutbildende Organe 5,507,624 21 B01 Antithrombotika 4,009,465 22 B01A Vitamin K Antagonisten 399,963 23 B01B Heparine 714,872 24 B01C Thrombozytenaggregationshemmer 2,265,838 25 B01D Fibrinolytika 14,293 26 27 B01X Direkte Thrombin Inhibit und direkte Faktor XA Hemmber 598,766 28 B02 Blutgerinnungssystem – Sonstige Produkte 237,495 29 B02B Antagonisten (Antidot.Anticoag) 146,201 30 B02D Blutcoagulation 1,805 31 B02G Glitazone Antidiabetika 48,969 32 B02X Übrige Blutgerinnungsprodukte (Tissues sealing preparations, 40,520
33 Thrombopoietin Agonisten, Antifibrinolytica synthetisch etc.) http://bmjopen.bmj.com/ 34 B03 Antianämica 1,259,896 35 B03A Antianämica mit Eisen und Kombinationen 862,744 36 B03C Erythropoietin Präparate 67,258 37 B03X Sonstige Antianämica inkl. Folsäure/Folinsäure 329,894 38 C Herz Kreislauf Therapie 16,027,143 39 40 C01 Cardica 954,962 C01B Anti Arrhythmie Präparate 213,952
41 on September 28, 2021 by guest. Protected copyright. 42 C01C Cardiale Stimulantien excl. Herzglykoside 266,028 43 C01D Coronartherapeutica excl. Calcium Antagonisten und Nitropräparate 121,815 44 C01E Nitropräparate und Analog 228,438 45 C01X Positive Inotrope, Herzglykoside und übrige Herzpräparate 124,729 46 C02 Antihypertonica 129,957 47 C03 Diuretica 1,815,738 48 49 C04 Cerebrale und periphere Vasotherapeutika 390,255 50 C05 Antivaricosa/Antihaemorrhoidalia 2,271,175 51 C05A Antihaemorrhoidalia topisch 773,800 52 C05B Antivaricosa topisch 1,029,010 53 C05C Antivaricosa systemisch 468,365 54 C06 Übrige Herz+Kreislaufmittel 339,930 55 C07 Betablocker 2,398,309 56 C07A Betablocker rein 2,272,988 57 58 60 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 61 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 C07B Betablocker Kombination 125,321 4 C08 Calciumantagonisten 1,157,267 5 C09 Stoffe mit Wirkung auf das Renin Angiotensin System 4,116,677 6 C09A Ace Hemmer, rein 1,142,465 7 C09B Ace Hemmer, Kombinationen 584,697 8 C09C Angiotensin II Antagonisten, rein 1,084,218 9 10 C09D Angiotensin II Antagonisten, Kombinationen 1,261,616 11 C09X Sonstige Stoffe mit Wirkung auf das Renin Angiotensin System 43,681 12 C10 Lipidregu/Antiarteriosklerotika 2,446,927 13 C10A Lipidregulatoren 2,341,963 14 C10A1 Statine (Cholesterolsynthesehemmer) 2,190,227 15 C10A2 ForLipidsenker, Fibrate,peer Ionenaustauscher review und sonstige Lipidsenker only 151,736 16 C10C Lipidregulatoren kombiniert mit anderen Lipidregulatoren 104,964 17 18 C11 Übrige Präparate für die Herz Kreislauf Therapie 5,946 19 D Dermatologica 17,314,810 20 D01 Dermatologische Antimykotica 1,360,570 21 D02 Emollientia und Hautschutzmittel 2,703,273 22 D03 Wundheilmittel 4,389,791 23 D04 Antipruriginosa inkl. Antihistaminica und Anästhetica usw. 1,458,276 24 D05 Nichtsteroidale Präparate gegen entzündliche Hauterkrankungen 327,245 25 D06 Topische antibakterielle Produkte und antivirale Mittel 864,829 26 27 D06A Topische antibakterielle Produkte 379,013 28 D06D Topische Produkte gegen Virusinfektionen 485,816 29 D07 Corticosteroide topisch 1,982,173 30 D07A Reine Corticosteroide topisch 960,182 31 D07B Corticosteroid Kombinationen topisch 1,021,991 32 D08 Antiseptica + Desinfizientia 1,836,788
33 D10 Aknemittel 1,227,850 http://bmjopen.bmj.com/ 34 D10A Topische Aknemittel 829,114 35 36 D10B Orale Aknemittel 398,736 37 D11 Übrige Dermatologica 1,164,015 38 G Urogenital System+Sexualhormon 7,936,641 39 G01 Gynaekologische Antiinfektiva 812,985 40 G01A Trichomonadenmittel 185,395
41 G01B Gynaekologische Antimykotica 475,911 on September 28, 2021 by guest. Protected copyright. 42 G01X Gynaekologische antibakterielle Produkte und Antiseptika 151,679 43 G02 Sonstige Gynaekologica 1,060,376 44 45 G02A Oxytocica 91,111 46 G02F Topische Sexualhormone 496,735 G02X Lokale Antikonzipientia, Prolactininhibitoren, wehenhemmende Mittel 47 472,530 48 und sonstige Gynaekologika 49 G03 Sexualhormone und Stimulantien des Genitalsystems 3,917,793 50 G03A Hormonale Kontrazeptiva systemisch 2,477,919 51 G03B Androgene, excl.G3E,G3F 70,852 52 G03C Oestrogene, excl.G3A,G3E,G3F 397,050 53 G03D Progestogene, excl. G3A,G3F 245,149 54 G03F Oestrogen Progestogen Kombinationen 487,291 55 G03G Gonadotropin inl. sonstige Ovulationsstimulantien 100,879 56 G03X Androgen Kombinationen mit weiblichen Sexualhormonen, SERMS und 138,653 57 58 61 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 62 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 sonstige Sexualhormone und ähnliche Produkte 4 G04 Urologica 2,145,487 5 G04A Urologische Antiinfektiva und Harnantiseptika 410,545 6 G04C BPH Produkte (Benigne Prostata Hyperplasie) 715,845 7 G04D Harninkontinenzprodukte 390,832 8 G04E Produkte gegen Erektionsstörungen 311,659 9 10 G04X Alle sonstigen Urologika 316,606 11 H Hormonales System, exkl. Sexualhormone 2,875,760 12 H01 Hypophysen /Hypothalamus Hormone 36,944 13 H02 Corticosteroide Systemisch 1,575,880 14 H03 Schilddruesentherapie 1,109,934 15 H04 SonstigeFor Hormone (wachstumshormone, peer reviewantidiuretische Hormone, Anti only 153,002 16 Parathyroid Hormone etc.) 17 J Antiinfektiva systemisch 8,446,623 18 J01 System.Antibakter.Prod. 5,721,173 19 J01A Tetracycline und Kombinationen 155,659 20 J01C Breitspektrum Penicilline 2,353,136 21 J01D Cephalosporine 897,839 22 23 J01F Macrolide und vergleichbare Substanzen 778,785 24 J01G Fluorchinolone 817,056 25 J01H Mittel und Schmalspektrum Penicilline 111,484 26 J01K Aminoglykoside 46,447 27 J01P Sonstige Beta Lactam. Antibiot. Penicilline, Cephalosporine (Monobac 69,074 28 tame, Peneme und Carbapeneme) J01X Rifamycin und vergleichbare Substanzen, Trimethoprim, Kombinationen 29 491,693 30 und vergleichbare Substanzen und sonstige antibakterielle Produkte 31 J02 Antimykotika, Systemisch 591,346 32 J04 Tuberkulosemittel 37,320
33 J05 Antivirale Mittel, systemisch 570,063 http://bmjopen.bmj.com/ 34 J05B Antivirale Mittel, exkl. Produkte gegen HIV 236,676 35 J05C Virustatika gegen HIV 333,387 36 J05C1 Nucleoside und Nucleotide Reverse Transkriptase Hemmer 108,107 37 J05C2 Protease Hemmer 98,581 38 J05CX Nicht Nucleoside Reverse Transkriptase Hemmer, HIV 126,699 39 Fusionshemmer, HIV Antivirale Integrase Hemmer und sonst.HIV Antivirale Mittel 40 J06 Sera und Gammaglobulin 23,164
41 J07 Vakzine 1,406,674 on September 28, 2021 by guest. Protected copyright. 42 J07B Vakzine Kombinationen 463,631 43 44 J07B1 Kombinationen mit Tetanusimpfstoff 387,139 J07BX Mehrfach Impfstoffe mit Masern und/oder Mumps und alle anderen 76,492 45 Kombinationen 46 J07D Bakterielle Impfstoffe 164,542 47 J07E Virale Impfstoffe 687,366 48 49 J07E1 Influenza Impfstoff 205,391 50 J07E4 Hepatitis Impfstoff 212,503 51 J07E9 Alle uebrigen viralen Impfstoffe, inkl. HPV, Rotaviren, Varizellen etc. 269,472 52 J07X Alle uebrigen Impfstoffe und aehnliche Produkte 91,135 53 J08B Übrige Antiinfektiva 96,883 54 K Infusionslösungen 24,158,749 55 K01 Intravenoese Lösungen ab 100 ML 12,064,154 56 K01A Elektrolytlösungen 3,372,213 57 58 62 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 63 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 K01B Standardlösungen 8,461,628 4 K01D Fettemulsionen, inkl. Produkte zur totalen parenteralen Ernährung 156,484 5 K01E Aminosäuren Lösungen 51,140 6 K01X Kalorische Lös. über 10% und Lösungen zur Osmotherapie 22,689 7 K02 Plasma Expander 70,547 8 K03 Blut und Blutersatzinfusionslösungen 1,817 9 10 K04 Injektionslösungen/Infusionszusätze 12,005,621 11 K04A Elektrolytlösungen 752,499 12 K04B Standardlösungen 9,981,186 13 K04C Kalorische Lösungen < 100 ML 34,050 14 K04D Sonstige Injektionslösungen/Infusionszusätze < 100 ML 1,237,886 15 K05 SpüllösungenFor peer review only 16,610 16 L Antineoplastika und Immunmodulatoren 1,934,950 17 18 L01 Antineoplastika 984,072 19 L01A Alkylisierende Substanzen 46,698 20 L01B Antimetaboliten 323,741 21 L01C Vinca Alkaloide & sonstige pflanzliche Produkte 157,793 22 L01D Antineoplastisch wirkende Antibiotika 46,619 23 L01F Platinhalt.Antineoplast. 95,347 24 L01G Monoklonale Antikörper zur antineoplastischen Therapie 159,208 25 L01H Proteinkinasehemmer zur antineoplastischen Therapie 33,895 26 27 L01X Sonstige Antineoplastika 120,771 28 L02 Cytostatische Hormonetherapie 183,379 29 L02A Cytostatische Hormone 43,050 30 L02B Cytostatische Hormonantagonisten 140,329 31 L03 Immunostimulantien 79,294 32 L03A Immunstimulantien 40,990
33 L03B Interferone 38,304 http://bmjopen.bmj.com/ 34 L04 Immunsuppressiva 688,205 35 36 L04B Anti TNF Produkte 340,857 37 L04X Sonstige Immunsuppressiva inkl. Interleukin Inhibitoren 347,348 38 M Muskel und Skelettsystem 14,787,413 39 M01 Antirheumatica 6,373,707 40 M01A Corticoidfreie Antirheumatica 5,700,326
41 M01C Antirheumatica spezifisch 673,381 on September 28, 2021 by guest. Protected copyright. 42 M02 Rheumaeinreibungen Rubefatientia 6,372,249 43 M03 Muskelrelaxantien 665,391 44 45 M03A Muskelrelaxantien peripher wirkend 85,757 46 M03B Muskelrelaxantien zentral wirkend 579,634 47 M04 Gichtmittel 368,020 48 M05 Übrigige Muskel Skelettsystem Präparate 1,008,046 49 N Nervensystem 42,690,195 50 N01 Anaesthetica 1,554,391 51 N01A Narkosemittel 1,078,794 52 N01A1 Narkosemittel Inhalationspräparate 20,582 53 54 N01A2 Narkosemittel Injektionspräparate 1,058,212 55 N01B Lokalanaesthetika 475,597 56 N01B1 Medizinische injizierbare Lokalanästhetika 198,964 57 58 63 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 64 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 N01B3 Topische Lokalanästhetika 270,962 4 N01B9 Sonstige Lokalanästhetika und dentale injizierbare Lokalanästhetika 5,671 5 N02 Analgetica 24,080,469 6 N02A Betäubungsmittel 1,661,607 7 N02B Analgetica und Antipyretica 21,918,098 8 N02B1 Analgetica rezeptpflichtig 8,592,792 9 10 N02B2 Analgetica rezeptfrei 13,325,306 11 N02C Migränemittel 500,764 12 N02C1 Triptane 432,741 13 N02C9 Sonstige Migränemittel 68,023 14 N03 Antiepileptica 1,957,745 15 N04 ParkinsonmittelFor peer review only 658,710 16 N05 Psycholeptica 9,075,955 17 18 N05A Antipsychotika 2,002,404 19 N05A1 Atypische Antipsychotika 1,521,618 20 N05A9 Konventionelle Antipsychotika 480,786 21 N05B Hypnotica und Sedativa 4,462,062 22 N05B1 Hypnotica und Sedativa, barbituratfreie Reinsubstanzen 2,830,382 23 N05B2 Hypnotica und Sedativa, barbituratfreie Kombinationen 129,461 24 N05B5 Pflanzliche Hypnotica und Sedativa 1,502,219 25 N05C Tranquilizer 2,611,489 26 27 N06 Psychoanaleptica Ex.Antiadipos 4,081,575 28 N06A Antidepressiva und stimmungsstabilisierende Produkte 3,503,509 29 N06A2 Pflanzliche Antidepressiva 245,372 30 N06A3 Stimmungsstabilisierende Produkte 50,396 31 N06A4 SSRI Antidepressiva 1,319,897 32 N06A5 SNRI Antidepressiva 664,070
33 N06A9 Sonstige Antidepressiva 1,223,774 http://bmjopen.bmj.com/ 34 N06B Psychostimulantia 341,172 35 N06X Psycholeptica Psychoanaleptica Kombinationen, Nootropica und 236,894 36 Neurotonica sowie verschiedene andere Produkte 37 N07 Sonstige ZNS wirksame Produkte 1,281,350 38 N07B Raucherentwöhnungsmittel 652,146 39 40 N07C Produkte gegen Schwindel 286,538 N07D Anti Alzheimer Produkte 121,928
41 on September 28, 2021 by guest. Protected copyright. 42 N07E Alkohol und Opiatentwöhnungsmittel 126,973 43 N07X Alle anderen ZNS wirksamen Präparate 93,765 44 P Parasitologie 462,559 45 P01 Antiprotozoenmittel und Anthelmintika 437,402 46 P01B Anthelmintica, excl. Schistosomiasis Mittel 296,327 47 P01D Malariamittel 139,242 48 49 P02 Sonstige Antiparasitäre Präparate inkl. Insektizide und Repellentien 26,990 50 R Respirationssystem 28,837,468 51 R01 Rhinologica 7,246,454 52 R01A Rhinologica topisch 6,868,431 53 R01AX Corticosteroide Rhinologika 1,067,991 54 R01A6 Antiallergische Rhinologica 230,180 55 R01A7 Abschwellende Rhinologica 4,290,641 56 R01A9 Sonstige Rhinologica topisch inkl. Antiinfektive Rhinologika ohne 1,279,619 57 58 64 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 65 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Corticosteroide 4 R01B Rhinologica systemisch 378,023 5 R02 Halsschmerzmittel und Antiinfectiva 3,977,806 6 R03 Antiasthmatika und Produkte gegen chronisch obstruktive 2,931,292 7 Lungenerkrankungen 8 R03A B2 Stimulatoren 923,570 9 R03B Xanthine 16,188 10 R03C Degranulationshemmer 9,088 11 R03D Corticoide 258,075 12 13 R03F Komb. B2 Stimulatoren mit Corticoiden 1,172,896 14 R03G Anticholinergica rein und in Komb. mit B2 Stimulatoren 422,479 15 R03G3 For Anticholinergica peer rein, Inhalate review only 269,418 16 R03G4 Anticholinergica – Kombinationen mit Beta 2 Stimulantien, Inhalate 153,061 17 R03J Antileukotrien Antiasthmatika 90,062 18 R03X Sonstige Antiasthmatika und Produkte gegen chronisch obstruktive 19 Lungenerkrankungen inkl. PDE4 Hemmer gegen Asthma / Chronisch Obstruktive 38,934 20 Lungenerkrankungen 21 R04 Percutane Einreibemittel und Inhalationspräparate 1,093,346 22 R05 Husten und Erkältungspräparate 10,899,026 23 R05A Erkältungspräparate 2,498,594 24 R05C Expectorantia ohne Antiinfectiva 4,499,364 25 R05D Hustensedativa 3,297,144 26 R05D1 Hustenmittel plain (Einzelsubstanzen) 1,971,766 27 28 R05D2 Sonstige Hustensedativa, inkl. Kombinationen 1,325,378 29 R05F Sonstige Husten und Erkältungspräparate 603,924 30 R06 Antihistamine systemisch 2,687,483 31 R07 Sonstige Präparate des Respirationssystems 2,061 32 S Sinnesorgane 7,658,311
33 S01 Ophthalmologica 7,214,097 http://bmjopen.bmj.com/ 34 S01A Ophtalmologische Antiinfectiva 451,604 35 S01B Ophtalmologische Corticoide rein 271,663 36 37 S01C Ophthalm. Antiphlogistika und Antiinfektiva Kombinationen 613,682 38 S01E Myotika und Mittel zur Glaukombehandlung 1,038,592 39 S01F Mydriatika und Cykloplegika 46,547 40 S01G Ophthalmologische Antiallergika, Abschwellende Mittel, Antiseptika 1,473,188
41 S01H Ophtalmolog. Lokalanesthetika 40,383 on September 28, 2021 by guest. Protected copyright. 42 S01K Künstliche Tränen und Netzmittel für Augen 2,503,750 43 S01P Ophthalmologische Antineovaskularisationsprodukte 127,501 44 S01R Nonsteroidale entzündungshemmende Ophthalmologica (NSAID'S) 236,207 45 S01X Übrige Ophtalmologica inkl. Präparate für den Gebrauch von 46 Kontaktlinsen, Augentonika und Augenvitamine, Präparate zur Verhütung von 411,009 47 Katarakt und Antikataraktogenika, Ophthalmologische Operationshilfsmittel und 48 Ophthalmologische Diagnostika 49 S02 Otologica 444,214 50 T Diagnostika 43,184 51 T01 Diagnostika für bildgebende Verfahren 18,885 52 T02 Testdiagnostika 24,299 53 V Verschiedenes 855,707 54 55 V01 Allergene 6,429 56 V03 Übrige Therapeutische Präparate 820,500 57 V03D Entgiftungspräparate für die Cytostatikatherapie 45,668 58 65 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 66 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 V03E Antidote 4,532 4 V03G Hyperkaliämie/Hyperphosphatämie Produkte 42,259 5 V03X Sonstige therapeutische Präparate inkl. Entzündungshemmende 728,041 6 Enzyme, Eisen Chelatbildner und Radiopharmazeutika 7 V05 Chirurgische Antiseptica 134 8 V06 Allgemeine Nährmittel 5,612 9 V07 Alle übrigen nicht therapeutischen Präparate 23,032 10 11 12 Source: Interpharma with data from IMS Health Schweiz, 2016. 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 66 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 67 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 4 Item Manu- 5 No script 6 Recommendation page 7 Title and abstract 1 (a) Indicate the study’s design with a commonly used 1 8 term in the title or the abstract [yes] 9 10 (b) Provide in the abstract an informative and 1-2 11 balanced summary of what was done and what was 12 found [yes] 13 14 Introduction 15 Background/rationaleFor 2peer Explain the reviewscientific background and only rationale for the 3 16 investigation being reported [yes] 17 18 Objectives 3 State specific objectives, [yes] including any 3 19 prespecified hypotheses [no] 20 21 22 Methods 23 Study design 4 Present key elements of study design early in the 3 paper [yes] 24 25 Setting 5 Describe the setting, locations, and relevant dates, 3 26 including periods of recruitment, exposure, follow-up, 27 and data collection [yes] 28 29 Participants 6 (a) Give the eligibility criteria, and the sources and 3 30 methods of selection of participants [yes] 31 32 Variables 7 Clearly define all outcomes, exposures, predictors, 3
33 potential confounders, and effect modifiers. [yes] http://bmjopen.bmj.com/ 34 Give diagnostic criteria, if applicable [no] 35 36 Data sources/ 8* For each variable of interest, give sources of data 4 37 measurement and details of methods of assessment 38 (measurement). Describe comparability of 39 assessment methods if there is more than one group 40 [yes]
41 on September 28, 2021 by guest. Protected copyright. Bias 9 Describe any efforts to address potential sources of 4 42 bias [yes] 43 44 Study size 10 Explain how the study size was arrived at [yes] 4-5 45 46 Quantitative variables 11 Explain how quantitative variables were handled in 4-5 47 the analyses. If applicable, describe which groupings 48 were chosen and why [yes] 49 50 Statistical methods 12 (a) Describe all statistical methods, including those 5 51 used to control for confounding [yes] 52 53 (b) Describe any methods used to examine 5 54 subgroups and interactions [yes] 55 56 (c) Explain how missing data were addressed [yes] 5 57 58 (d) If applicable, describe analytical methods taking n.a. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 68 of 69 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 account of sampling strategy [no] 4 5 (e) Describe any sensitivity analyses [no] n.a. 6 7 Results 8 Participants 13* (a) Report numbers of individuals at each stage of 5 9 study—eg numbers potentially eligible, examined for 10 eligibility, confirmed eligible, included in the study, 11 completing follow-up, and analysed [no] 12 13 (b) Give reasons for non-participation at each stage 5 14 [yes] 15 For peer review only 16 (c) Consider use of a flow diagram [yes] 18 17 18 Descriptive data 14* (a) Give characteristics of study participants (eg 22-24 demographic, clinical, social) and information on 19 exposures and potential confounders [yes] 20 21 (b) Indicate number of participants with missing data 5 22 for each variable of interest [yes] 23 24 Outcome data 15* Report numbers of outcome events or summary 6 25 measures [yes] 26 27 Main results 16 (a) Give unadjusted estimates and, if applicable, 28 28 confounder-adjusted estimates and their precision 29 (eg, 95% confidence interval). Make clear which 30 confounders were adjusted for and why they were 31 included [yes] 32 b 33 ( ) Report category boundaries when continuous n.a. http://bmjopen.bmj.com/ 34 variables were categorized n.a. 35 c 36 ( ) If relevant, consider translating estimates of n.a. 37 relative risk into absolute risk for a meaningful time period [no] 38 39 Other analyses 17 Report other analyses done—eg analyses of 24 40 subgroups and interactions, and sensitivity analyses
41 [yes] on September 28, 2021 by guest. Protected copyright. 42 43 44 Discussion 45 Key results 18 Summarise key results with reference to study 8 46 objectives [yes] 47 Limitations 19 Discuss limitations of the study, taking into account 2 48 sources of potential bias or imprecision. Discuss both 49 direction and magnitude of any potential bias [yes] 50 51 Interpretation 20 Give a cautious overall interpretation of results 8-11 52 considering objectives, limitations, multiplicity of 53 analyses, results from similar studies, and other 54 relevant evidence [yes] 55 56 Generalisability 21 Discuss the generalisability (external validity) of the 6 57 study results [yes] 58 59 Other information 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 69 of 69 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Funding 22 Give the source of funding and the role of the funders 13 4 for the present study and, if applicable, for the original 5 study on which the present article is based [yes] 6 7 8 *Give information separately for exposed and unexposed groups. 9 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological 10 background and published examples of transparent reporting. The STROBE checklist is best used in conjunction 11 12 with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of 13 Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the 14 STROBE Initiative is available at www.strobe-statement.org 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from
Medication Incidents in Primary Care Medicine: a Prospective Study in the Swiss Sentinel Surveillance Network (Sentinella)
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2016-013658.R1
Article Type: Research
Date Submitted by the Author: 21-Nov-2016
Complete List of Authors: Gnädinger, Markus; University of Zurich, Institute for General Medicine Conen, Dieter; Swiss Patient Safety, Herzig, Lilli; University of Lausanne, Institute of General Medicine Puhan, Milo; University of Zurich, Institute of Epidemiology, Biostatistics & Prevention Staehelin, Alfred; Sentinel Surveillance Network, Swiss Federal Office of Public Health, Zoller, Marco; Zurich University Hospital, Instiute for General Medicine Ceschi, Alessandro; National Poisons Centre, Tox Info Suisse, Associated Institute of the University of Zurich, Division of Science, Head of ; University Hospital Zurich , Dept. Clinical Pharmacology & Toxicology
Primary Subject http://bmjopen.bmj.com/ General practice / Family practice Heading:
Secondary Subject Heading: Pharmacology and therapeutics, Paediatrics, Medical management, Nursing
Health & safety < HEALTH SERVICES ADMINISTRATION & MANAGEMENT, Keywords: Adverse events < THERAPEUTICS, Patient safety, Pharmaceutical preparations, Medication errors
on September 28, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Medication Incidents in Primary Care Medicine 5 6 7 8 A Prospective Study in the Swiss Sentinel Surveil 9 10 lance Network (Sentinella) 11 12 1Markus Gnädinger (corresponding author, [email protected]), 2Dieter Conen, 3,4Lilli 13 5 1,4 1 6 14 Herzig, Milo A. Puhan, Alfred Staehelin, Marco Zoller, Alessandro Ceschi 15 For peer review only 16 1Institute of Primary Care, University of Zurich, 2Patientensicherheit Schweiz, Zurich 3Policlinique 17 4 18 Médicale, University of Lausanne, Sentinel Surveillance Network, Swiss Federal Office of Public 19 Health, Bern (Sentinella), 5Epidemiology, Biostatistics, and Prevention Institute, University of Zur 20 6 21 ich, Division of Clinical Pharmacology and Toxicology, Department of Internal Medicine, Ente 22 Ospedaliero Cantonale, Lugano, Switzerland and Department of Clinical Pharmacology and Toxi 23 24 cology, University Hospital Zurich, Zurich, Switzerland and National Poisons Centre, Tox Info 25 Suisse, Associated Institute of the University of Zurich, Zurich, Switzerland 26 27 28 (Correspondence: Markus Gnädinger, Dr. med. Facharzt für Innere Medizin, Birkenweg 8, 9323 Steinach, 29 0041 71 446 04 64) 30 31 32
33 http://bmjopen.bmj.com/ 34 Abstract 35 36 37 Objectives: To describe the type, frequency, seasonal and regional distribution of medication inci 38 39 dents in primary care in Switzerland and to elucidate possible risk factors for medication incidents. 40
41 Design: Prospective surveillance study. on September 28, 2021 by guest. Protected copyright. 42 43 44 Setting: Swiss primary health care, Swiss Sentinel Surveillance Network. 45 46 Participants: Patients with drug treatment who experienced any erroneous event related to the 47 48 medication process and interfering with normal treatment course, as judged by their physician. The 49 180 physicians in the study were general practitioners or pediatricians participating in the Swiss 50 51 Federal Sentinel reporting system in 2015. 52 53 Outcomes: Primary: medication incidents; secondary: potential risk factors like age, gender, 54 55 poly medication, morbidity, care dependency, previous hospitalization. 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Results: The mean rates of detected medication incidents were 2.07 per general practitioner per 4 year (46.5 per 100,000 contacts) and 0.15 per pediatrician per year (2.8 per 100,000 contacts), 5 6 respectively. The following factors were associated with medication incidents (OR, 95% CI): higher 7 age 1.004 per year (1.001;1.006), care by community nurse 1.458 (1.025;2.073) , and care by an 8 9 institution 1.802 (1.399;2.323) , chronic conditions 1.052 (1.029;1.075) per condition, medications 10 1.052 (1.030;1.074) per medication, as well as Thurgau Morbidity Index for stage 4: 1.292 11 12 (1.004;1.662), 5: 1.420 (1.078;1.868), and 6: 1.680 (1.178;2.396), respectively. Most cases were 13 14 linked to an incorrect dosage for a given patient, while prescription of an erroneous medication was 15 the second mostFor common peer error. review only 16 17 18 Conclusions: Medication incidents are common in adult primary care whereas they rarely occur in 19 pediatrics. Older and multimorbid patients are at a particularly high risk for medication incidents. 20 21 Reasons for medication incidents are diverse but often seem to be linked to communication prob 22 lems. 23 24 25 Trial registration: www.clinicaltrials.gov, NCT0229537 26 27 Keywords: Patient safety, pharmaceutical preparations, medication errors. 28 29 30 Strength and limitations 31 This is the first prospective and systematic collection of incident data in primary care in Switzerland 32
33 that is characterized by three linguistic regions and two drug distribution systems. http://bmjopen.bmj.com/ 34 It was conducted by experienced physicians and with high response rates. 35 36 37 There is likely – as expected – bias from selective and underreporting or non detection of medica 38 tion incidents. 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 Introduction 45 46 47 Patient safety is a major concern in healthcare systems worldwide. Although most safety research 48 has been conducted in the in patient setting [1], evidence indicates that medical errors and ad 49 50 verse events pose a serious threat for patients in the primary care setting as well, since most pa 51 tients receive ambulatory care [2 4]. The rationale of this project has been published in our study 52 53 protocol [5]. The aim of the project was to describe the type, incidence, seasonal and regional dis 54 tribution of medication incidents in primary care in Switzerland and to elucidate risk factors for 55 56 medication incident. 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 Method 7 8 9 Study design 10 11 We conducted a prospective surveillance study among primary care patients during 2015 to identi 12 13 fy cases of medication incidents. 14 15 Study populationFor peer review only 16 17 18 The study population was any person undergoing drug treatment in general internal or pediatric 19 practices participating the Sentinella network. The latter covers a representative sample of patients 20 21 in primary care for Switzerland [6, included to this manuscript]. Founded in 1986, it was mainly 22 designed to survey transmissible diseases. Later, it also assessed other health problems of public 23 24 interest. It generates daily to weekly current data and covers the entire geographic and linguistic 25 regions of our country. Children, the mentally handicapped or the elderly were also included, all of 26 27 whom might be at increased risk for medication errors. 28 29 Medication incidents 30 31 We defined medication incidents as any erroneous event (as defined by the physician) related to 32
33 the medication process and interfering with normal treatment course (e.g. administration of an er http://bmjopen.bmj.com/ 34 35 roneous medication). We did not include lack of treatment effect, adverse drug reactions or drug 36 drug or drug disease interactions without detectable treatment error. Nor did we consider medica 37 38 tion incidents if patients refused to have them reported to the Sentinella system. 39 40 Data sources
41 on September 28, 2021 by guest. Protected copyright. 42 43 The study physicians recorded the patient’s year of birth and gender on their weekly reporting 44 45 form. After a maximum of four weeks they had to fill in a detailed incident questionnaire (Appendix 46 A). It was comprised of their Sentinella number and the calendar week of notification. Concerning 47 48 the patients, they reported the living situation, several supposed risk factors for an incident, as well 49 as the following variables: hospitalization during the previous year, care dependency, number of 50 51 drugs used chronically, number of chronic conditions, and the Thurgau Morbidity Index (TMI), to be 52 compared with a denominator analysis (below) [6]. We further received a detailed description of the 53 54 incident and proposals to avoid future incidents. 55 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 We got the annual number of patient to physician contacts (PPC) per practice from the Sentinella 4 administration, as well as morbidity data from a fortnight cross sectional denominator analysis of all 5 6 patients consulting a Sentinella practice during weeks 11 or 12 [6]. 7 8 We received the anonymized list of participating physicians, their specialty, as well as the commu 9 10 nity size (Swiss Federal Statistical Office) and the linguistic region from the Sentinella administra 11 tion. Information on Swiss medication sales in 2015 by ATC groups was derived from Interpharma 12 13 Switzerland (Appendix F). 14 15 The questionnairesFor were peercompleted either review electronically or on aonly paper/pencil version, the former as 16 17 online SurveyMonkey™ questionnaires, the latter as sealed envelopes sent from the Sentinella 18 19 administration. We had three study questionnaires: a detailed incident questionnaire, an initial one, 20 and a final one (Appendices A, B, and C). The initial questionnaire served to describe the physi 21 22 cian’s practices in terms of e.g. number of physicians, availability of electronic patient history, elec 23 tronic drug drug interaction control system, and drug distribution system (Appendix B). The final 24 25 questionnaire investigated non reporting and difficulties with coding the morbidity variables (Ap 26 pendix C). 27 28 29 We estimated “incident relevance” from the items “disturbance” and “endangering” of the patients; 30 if any of them was graded with “medium” or higher, the variable relevance was set to “more”, oth 31 32 erwise it was set to “less”. Because of question ambiguity, we set coding of the item care
33 dependency to “missing” for patients younger than 20 years of age. Evans’ Index – a prognostic http://bmjopen.bmj.com/ 34 35 index – was calculated by simple addition of the number of chronic drug treatments with the num 36 ber of chronic conditions [7]. 37 38 39 The following free text variables were manually coded: relationship of the incident to the suspected 40 medication, preventability of the incident, reactions to the incident, and proposals to avoid further
41 on September 28, 2021 by guest. Protected copyright. 42 incidents. 43 44 Statistical methods 45 46 Values are given as frequencies, mean ± SD or median [interquartile range (IQR)] (as denoted with 47 (first quartile; forth quartile), depending on non normal distribution or non interval scaled data level. 48 49 To assess the association of medication incidents with potential risk factors we used the SPSS 50 GENLINMIXED procedure. Clustering of patients was addressed by using a mixed binary logistic 51 52 regression with the fixed factors of gender, year of birth, care dependency, number of chronic drug 53 54 treatments, number of chronic conditions, and TMI as well as the physicians’ practice number as a 55 random factor; if one item was missing, the whole record was excluded from the analysis. Numera 56 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 tor was affiliation to the case group; denominator was all PPC during weeks 11 and 12. We used 4 IBM SPSS 23. 5 6 7 8 Results 9 10 The Sentinella system 11 12 During the year 2015, 149 practices were enrolled to the Sentinella system. Of them, 144 practices 13 were known to report regularly; their properties are listed in Table 1. The Sentinella physicians are 14 15 representativeFor for the overall peer Swiss physician review population [5,6]. onlyDrugs were auto distributed by 42% 16 of the study practices. Approx. half of the physicians had electronic and the other paper based 17 18 medical records. Systematic drug drug interaction control systems were installed only by a minority 19 (36.8%) of the physicians. During the year 2015 (which included unusually for calendar adaptation, 20 21 53 instead of 52 reporting weeks), the general practitioners (GPs) had 4,456±2,137 PPC; for the 22 23 pediatricians (PEDs), these were 5,297±2,715 (mean±SD). In their initial study questionnaire, 24 11.1% of the reporting physicians admitted to be working part time (i.e. less than 30 h per week). 25 26 27 Study flow 28 29 During the year 2015, we received 216 incident notifications (Figure 1). In 11 cases, we did not 30 receive the detailed notification form. Eight cases had to be removed from the database because 31 32 they fulfilled the exclusion criterion (i.e. “adverse drug reactions without detectable error”), and one
33 case had to be removed because of double data entry. This led to 197 cases which could be ana http://bmjopen.bmj.com/ 34 35 lyzed. The distribution of the monthly incident notifications throughout 2015 is depicted in Figure e1 36 (Appendix D), and the distribution of the numbers of cases reported by each practice in Figure 2. 37 38 39 Description of patients 40 Table 2 lists age, gender and geographic distribution as well as the physician to patient relation 41 on September 28, 2021 by guest. Protected copyright. 42 ship and the observer of the incident. Only three cases evolved in PED practices. No statistically 43 44 significant differences were found between the two relevance classes of the incidents. 45
46 47 Number of incidents, non-reporting 48 49 In the GPs 194 incidents, 148 physicians, 4,456 yearly PPC led to 1.31 incidents per physician per 50 year or 29.4 per 100,000 PPC; in PEDs 3 incidents, 32 physicians, 5,297 yearly PPC led to 0.1 51 52 incident per physician per year or 1.8 per 100,000 PPC. 53 54 55 To evaluate the non-reporting of incidents, we asked the physicians in the final study question 56 naire. Out of 180 actively reporting physicians we received 145 questionnaires (80.6% response 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 rate). To our question: “Did you not report medication incidents that you had noticed during the last 4 year?”, they answered: “Never or almost” 110 (75.9%), “yes, but seldom” 22 (15.2%), “yes, fre 5 6 quently” 9 (6.2%), “always, or almost” 4 (2.8%). Reasons for not reporting incidents were “lack of 7 time” or “forgetfulness”. If we speculate that these answers represent reporting rates of 76 100, 51 8 9 75, 26 50, or 0 25% respectively, we could divide the observed rates by the middle of the reporting 10 classes: 0.875, 0.625, 0.375 and 0.125. By doing so, we calculated a rate of 50% of underreport 11 12 ing; if we furthermore consider the 5% of the incidents where the questionnaires were not sent, this 13 14 rate increases to 58%. We therefore have to multiply the observed rates with a factor of 1.58 re 15 sulting in theFor following rates peer of detected incidents:review GP 2.07 per onlyphysician per year, 46.5 per 16 17 100,000 PPC and PED 0.15 per physician per year, 2.8 per 100,000 PPC. 18 19 20 Types and causes of incidents, organ systems involved, preventability 21 The types of error are listed in Table e1 (Appendix D) and Figure 3; in 26 cases more than one 22 23 was mentioned. Most cases were linked to an incorrect dosage for a given patient, while prescrip 24 tion of an erroneous medication was the second most common error. 25 26 27 Most errors concerned orally ingested medication. Parenteral drugs led to fewer error reporting; in 28 our study, they comprised insulin and vaccinations. There were cases of an incorrect (influenza vs. 29 30 anti tetanus) or incomplete (Boostrix® vs. Boostrix Polio®) vaccination and of undue vaccination (a 31 third anti HPV vaccine). The first case was due to a communication problem within the practice 32
33 staff, the other to a vaccination in absence of the vaccination card. http://bmjopen.bmj.com/ 34 35 Three cases concerned pediatric patients, all were linked to inadequate dosing, two times of an 36 37 antibiotic prescription (Cotrimoxazole, Co Amoxicillin) and one time of an antiemetic drug 38 (Thiethylperazine). 39 40 In 89 of the 195 cases, organ system damage was reported, in 13 cases more than one organ sys 41 on September 28, 2021 by guest. Protected copyright. 42 tem was involved, most frequently the central nervous system (Figure 4) (Table e2, Appendix D). 43 44 Possible triggers of the incident were reported in 194 of the 197 cases; in 45 cases there was more 45 than one single reason, most frequently lacking alertness of the reporting physicians or their staff 46 47 (Table e3). When asked who might be the “responsible” for the incident, 191 replied. The most 48 49 common person or institution possibly responsible for the medication incident was: the reporting 50 physician 41 (21.5%), followed by the practice nurse 26 (13.6%), the institution where the patient 51 52 lives 33 (17.3%), the pharmacy 7 (3.7%), the hospital 12 (6.3%), the community nurse 4 (2.1%), 53 the patients or their proxies 15 (7.9%), and the manufacturer 2 (1.0%); in 9 cases (4.6%) this was 54 55 unclear, in 37 cases (19.4%), there was more than source one to blame. Preventability of the inci 56 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 dents was classified (by our study board): unlikely in 6 cases (3.0%), possible in 58 (29.4%), prob 4 able in 114 (57.9%), and definite in 19 (9.6%). 5 6 7 Endangering and disturbances 8 9 In 192 of the 197 cases we received information about patients’ endangering as estimated by the 10 physicians. In 39 cases (20.3%) there was no endangering, in 83 (43.2%) light, in 51 (26.6%) 11 12 moderate, in 19 (9.9%) severe. The disturbances caused by the incident are listed in Table 3. Out 13 14 of the 197 incidents, 74 (37.5%) were classified as “more” relevant (Table 2). 15 For peer review only 16 17 Interface problems, orientation, predictability, repeat incidents 18 The presence of interface problems was reported in 64 of 197 cases (32.4%); these were: with a 19 20 hospital in 28 cases (43.8%), with an institution in 14 (21.9%), with a community nurse in 6 (9.4%), 21 with a pharmacist in 9 (14.0%), with a specialist in 3 (4.7%), and with others in 4 (6.3%). In 184 22 23 cases, we received information about orienting patients about the incident; in 98 cases, the patient 24 was oriented by the reporting physician or his staff (50.3%), non orientation was stated: because 25 26 the patient was not able to understand (children, demented) in 26 cases (14.1%), because the 27 problem had already been solved and the notification would have unduly disturbed the confidence 28 29 in 18 cases (9.8%), because the patient or the proxies had observed themselves in 23 cases 30 (12.5%), because the patient had already been oriented by others in 7 cases (3.8%) or because of 31 32 another reason in 12 cases (6.5%). As for predictability of the incident, we received 183 valid an
33 http://bmjopen.bmj.com/ swers; of them 82 (44.8%) stated “yes, in the given constellation, the incident was to be expected”. 34 35 When asked whether they had already reported a similar incident in the study, 29 out of 196 36 (14.8%) answered in the affirmative. 37 38 39 40 Risk factors to undergo a medication incident
41 To detect patient risk factors to undergo an incident, we compared the incident data with those of a on September 28, 2021 by guest. Protected copyright. 42 43 fortnight denominator analysis [6]; the following univariate factors accumulated preferentially in 44 incident patients: higher age, care dependency, higher numbers of chronic conditions or medica 45 46 tions (or higher Evans’ Index), as well as higher TMI (table 4). In the multivariate analysis only in 47 patient care by an institution remained a significant factor. Other suspected risk factors are listed in 48 49 Table e4 (Appendix D); these items were not included in the denominator study and therefore lack 50 a comparator. Within the patient group with a more as compared to a less relevant incident, only 51 52 psychiatric illness reached a significantly increased proportion. We did not detect major differences 53 between the two drug distribution systems. 54 55 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Relationship between incidents and drug class, ATC-group 4 We assessed semi quantitatively whether the class of the suspected drug was causally linked to 5 6 the emergence of the incident; the frequencies were: none 49 (25.1%), unlikely 31 (15.9%), possi 7 ble 77 (39.5%), probable 34 (17.4%), definite 4 (2.1%), did not apply 2 (n.a). ATC codes of the 8 9 suspected medications are listed in Table 5. Most naming concerned the groups C “cardiovascular” 10 (23.2%) and N “nervous system” (22.1%). Among the medication classes judged to be related to 11 12 the incident (n=37) the most frequently named group were oral anticoagulants: rivaroxaban 9, 13 14 phenprocoumon 7, and acenocoumarol 2 cases; this explains the 7 fold increased relative risk of 15 ATC group BFor as compared peer to sales. As anreview example for errors withoutonly relation to the drug class we 16 17 must mention the 17 cases of institutionalized patients ingesting medications scheduled for other 18 residents, in most cases this was person eating at the same table, but in other cases the person in 19 20 care mixed up patient names. 21 22 23 Reactions to the incident and suggestions to prevent further ones 24 In 141 of the 197 cases, the physicians reported to have changed something after the incident, in 25 26 13 cases this was more than one reaction, mostly often named were communication with other 27 caregivers or better instruction of patients (Table e5, Appendix D). The respondents were asked for 28 29 proposals how to prevent future incidents of the reported type; 125 of them made a total of 243 30 suggestions (Table e6, Appendix D). Of these, 37 (15.2%) were related to accurate medication 31 32 lists, 10 (4.1%) to better patient instructions, 38 (15.6%) to organization of regular follow up con
33 http://bmjopen.bmj.com/ trols, and 55 (22.6%) involved organizational changes within the practice and its staff. 34 35 36 37 Discussion 38 39 In a representative group of primary care physicians [8], we found approximately one case of a 40 medication incident per GP and year.
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 Incident rates 45 We calculated the rates of detected medication incidents as follows: GP 2.07 per physician and 46 47 year, 46.5 per 100,000 PPC and PED 0.15 per physician and year, 2.8 per 100,000 PPC. Medica 48 tion incidents may make up a proportion of approximately one third of all incidents [9]; the rates for 49 50 all safety incidents may amount 6.20 per physician and year or 139.4 per 100,000 PPC in GPs and 51 0.45 per physician and year or 8.4 per 100,000 PPC in PEDs. 52 53 In Australian primary care patients, an incident rate of 4.98 per GP and year was reported; this is 54 close to our estimated rate of 6.20; however, there was no sub typing of the incidents [10]. In a 55 56 three year study, O’Beirne and colleagues reported a rate of 1.8 safety incidents per year and phy 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 sician [9]. In a literature review of western countries’ publications, Sandars and Esmail described a 4 rate of 5 to 80 errors per 100,000 consultations [2], a rate somewhat lower than the 139.4 cases 5 6 per 100,000 consultations estimated in our study. Kuo and colleagues reported a proportion of 7 15% of all incidents to be related to medication [11]; since we estimated a proportion of 33% this 8 9 would give rise to even higher rates of all safety incidents when calculated from our study data 10 base. A 11% group of the physicians worked part time, the rate per full time physician per year 11 12 may therefore be somewhat higher than calculated. The 13 fold higher incident rate in GPs as 13 14 compared to PEDs is not surprising given the lower medication rates in children. A recent British 15 study confirmedFor this much peer lower rate of incidentsreview in children; outonly of 46,902 family practice safety 16 17 reports, 1,788 concerned children (26 times less than adults) [12]. 18 19 Definition of incidents and reliability of reporting 20 21 On the other hand, incident rates may be influenced by their definition. Gandhi and colleagues 22 coined the term “avoidable adverse drug reaction” [13]. In our study, we explicitly excluded adverse 23 24 drug reactions without detectable error. Runciman and colleagues defined a patient safety incident 25 as follows: “An event or a circumstance that could have resulted or did result in unnecessary harm 26 27 to a patient” [14]. An intuitive definition of a medical error was given by Makeham et al: “That was a 28 threat to patient wellbeing and should not happen. I don’t want it to happen again.” [15]. As shown 29 30 in Figure e1 (Appendix D), reporting frequency was higher at the beginning of the study as com 31 pared to the later course of it. This could reflect some loss of interest or forgetfulness by the report 32
33 ing physicians. As calculated from our final questionnaire after the study, the reporting physicians http://bmjopen.bmj.com/ 34 failed to report about one in three cases of the detected medication errors. Non detection of inci 35 36 dents may even be more frequent than non reporting of observed incidents. A missed possible 37 drug drug interaction may be detected by chart review, a documentation error would probably have 38 39 been found only in 1:1 supervision, which is very time consuming, costly and may additionally in 40 fluence performance of the observed physician. It is therefore virtually impossible to decipher the
41 on September 28, 2021 by guest. Protected copyright. 42 real rate of non detected incidents. The problems in detection of incidents were the reason to pos 43 44 tulate a “mix of methods” as needed to identify adverse events in general practice [16]. 45 46 47 48 Other approaches 49 In a retrospective, semi quantitative analysis, Gehring and colleagues investigated safety incidents 50 51 in Swiss primary care [17]; among 23 predefined classes of safety incidents, the respondents ad 52 mitted 15 to have occurred at least yearly – four of them being linked to drug treatment. Another 53 54 approach to incidents is to ask patients as performed by Mira and colleagues [18]; they interviewed 55 patients (> 65 years with five or more chronic drug treatments) and found that 75% of the patients 56 57 reported to have been affected by at least one medication error during the previous twelve months. 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 We did not collect data on appropriateness of treatment; admittedly medication inappropriate for 4 some groups of patients (elderly or patients with impaired renal function) may provoke incidents 5 6 [19]. A recent Scottish study demonstrated an impressive reduction in high risk prescriptions (non 7 steroidal anti inflammatory drugs, antiplatelet, and anticoagulation) as well as hospital admissions 8 9 for gastrointestinal ulcers or heart failure by a combination of educational measures, informatics 10 and financial incentives [20]. 11 12 13 14 Type, consequences, causes and preventability of error 15 For peer review only 16 Most cases involved application of an erroneous dosing or of a wrong medication; non application 17 18 of necessary medication was also frequent. When a wrong medication was dispensed, the error 19 was often caused by confounding of prepared medication in home residents. The classic case of 20 21 non application of a necessary medication was the missed re uptake of anticoagulation after an 22 operation. Non application of necessary drugs seems to be a relevant source of unnecessary harm 23 24 to patients [21]. The causing of the incidents was similar as reported by others [11,17,18,22]. Most 25 studies found – as in our patients – lacking alertness of and communication problems within prac 26 27 tice staff, but there was a large variety of other topics. All three pediatric cases were linked to pre 28 scription of an inadequate dosing for age and weight. More than half of the patients did not have 29 30 any disturbances after the incident; otherwise in most cases the nervous system was affected. No 31 fatalities were reported, but seven patients (3.5%) needed stationary care. In 2004 Pirohamed and 32
33 colleagues published a study on adverse drug reactions as a cause for admission to hospital; they http://bmjopen.bmj.com/ 34 found that 6.5% of all hospitalizations and 4.0% of all hospital days were caused by adverse drug 35 36 reactions, 72% of them being preventable and 2% leading to death [23], however, this study in 37 cluded all cases, and therefore a majority of cases of adverse drug reactions without a detectable 38 39 error. An older Swiss study was published by Livio and colleagues; out of 3,195 hospitalizations, 40 she identified 229 cases (7.2%) as probably caused by adverse drug reactions [24]. In that study,
41 on September 28, 2021 by guest. Protected copyright. 42 32% of the events were classified as being preventable (which does not strictly mean that an error 43 44 had caused the incident), and in 6% of the cases, fatalities were reported. The results on outpa 45 tients contrast to inpatients where one in ten drug administrations was described to be erroneous 46 47 [25]. Medical errors seem to be the “third leading cause of death in the US” [26]. 48 49 50 51 Risk factors 52 When looking for factors accumulating in the incident group, the univariate analysis comparing a 53 54 fortnight denominator with our data revealed all factors investigated (except gender) as significantly 55 accumulated in the incident group; these were higher age, care dependency, higher numbers of 56 57 chronic conditions or medications (and higher Evans’ Index), as well as higher TMI. When perform 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 ing a multivariate analysis, only those living in a home for the elderly or handicapped remained a 4 significant risk factor, but the small number of observations precluded possible less important risk 5 6 factors to be detected. The only significant risk factor for undergoing an incident of higher rele- 7 vance was psychiatric disease. In the literature there are several factors described to correlate with 8 9 proneness to undergo a medication incident, which quite well reflects the results of our study. 10 Mainly, this is evidently the number of drugs ingested [13,27]. Higher or young age [28], or morbidi 11 12 ty [29,30] were also described. Most incidents concerned ATC groups N “nervous system” or C 13 14 “cardiovascular”, which were also of the mostly sold drugs; an exception was group B with antico 15 agulants andFor a sevenfold peer increased relative review risk as compared toonly Swiss sales in 2015. It seems 16 17 wise to be alert to avoid errors when prescribing medication of these groups. The prevailing posi 18 tion of anticoagulants (18.5% of cases) was described also by Field and colleagues [30]. Otherwise 19 20 the repartition of our suspected drugs was similar to other primary care studies [11,30], a literature 21 review [31] or a theoretical paper [22]. 22 23 24 Limitations 25 26 There is likely – as expected – bias from selective and underreporting or non detection of medica 27 28 tion incidents. Therefore, the true rate of incidents could only be estimated, and the proportions of 29 incident characteristics would probably substantially have been altered by a more complete record 30 31 ing of incidents. 32
33 http://bmjopen.bmj.com/ 34 Conclusion 35 36 37 Medication incidents are common in general medicine whereas they rarely occur in pediatrics. 38 Reasons for medication incidents are diverse but often seem to be linked to communication prob 39 40 lems. Older and multimorbid patients are at a particularly high risk for medication incidents.
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 Abbreviations 46 47 GP general practitioner 48 49 50 PED pediatrician 51 52 PPC patient to physician contacts 53 54 55 TMI Thurgau Morbidity Index 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Author affiliations 4 5 MG Institute of Primary Care, University of Zurich, Switzerland. 6 7 AC Division of Clinical Pharmacology and Toxicology, Department of Internal Medicine, Ente 8 9 Ospedaliero Cantonale, Lugano, and Department of Clinical Pharmacology and Toxicology, 10 University Hospital Zurich, Zurich, and National Poisons Centre, Tox Info Suisse, Associated 11 12 Institute of the University of Zurich, Zurich, Switzerland 13 DC Patientensicherheit Schweiz, Zurich, Switzerland. 14 15 LH PolicliniqueFor Médicale, peer University of Lausanne,review Switzerland andonly Sentinel Surveillance Network, 16 Swiss Federal Office of Public Health, Bern, Switzerland. 17 18 MP Epidemiology, Biostatistics, and Prevention Institute, University of Zurich, Switzerland. 19 AS Sentinel Surveillance Network, Swiss Federal Office of Public Health, Bern, Switzerland. 20 21 MZ Institute of General Medicine, University of Zurich, Switzerland. 22 23 24 Acknowledgements 25 26 27 We are grateful to Lee Wennerberg for the English language corrections, Dr Sven Staender, 28 Männedorf for the helpful comments, and Simon Gnädinger, Zurich for the manual coding of free 29 30 text items. We thank the Sentinella program commission for their support, the reporting physicians 31 of Sentinella for their unflagging enthusiastic collection of data, and the Federal Office of Public 32
33 Health for providing data and translating the questionnaires into French. http://bmjopen.bmj.com/ 34 35 36 37 38 Contributorship statement 39 40 MG lead the study, did the pilot study (questionnaire development, data entering and processing)
41 on September 28, 2021 by guest. Protected copyright. 42 wrote all documents, did all the contacts with the Sentinella administration, ethics committee, and 43 others, programmed the electronic questionnaires, entered hand written questionnaires into the 44 45 database, did the data processing and wrote the publication after data collection. 46 AC is an expert on clinical pharmacology and drug safety. 47 48 DC is an expert on patient safety. 49 LH is French speaking and helped to interpret the French questionnaires. She is an expert on mul 50 51 timorbidity. She is a member of the Sentinella program commission. 52 MP is head of Epidemiology, Biostatistics & Prevention Institute. He is responsible for the sound 53 54 methodology. 55 AS had the idea for the study. He is vice president of the Sentinella program commission. 56 57 MZ is expert on electronic data exchange in primary care. 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 All of them have seen all the study documents and have contributed intellectually to their elabora 4 tion. All have contributed to revise the draft of this publication and approve the submitted version of 5 6 this publication. 7 8 Funding 9 10 11 The study was funded by Bangerter Rhyner Foundation, Basel. 12 13 14 Competing interests 15 For peer review only 16 17 The authors declare that they have no financial interest conflicts with this study. 18 19 Patient consent 20 21 22 Not necessary for anonymous data. 23 24 Ethics approval 25 26 27 The ethical committee of Canton Zurich decided that our study did not need formal approval be 28 cause the data are completely anonymous (KEK ZH 2014 0400). The study was recorded in 29 30 www.ClinicalTrials.gov: NCT02295371, as well as in our national study registry (www.kofam.ch; 31 32 SNCTP000001207).
33 http://bmjopen.bmj.com/ 34 Data sharing statement 35 36 37 No additional data are available. 38 39 Strobe statement 40
41 on September 28, 2021 by guest. Protected copyright. 42 Whenever possible, the guidelines of the Strobe statement were followed (Appendix E). 43 44 45 46 References 47 48 49 1. Manias E: Detection of medication related problems in hospital practice: a review. Br J Clin 50 Pharmacol 2012; 76 (1): 7 20. 51 52 2. Sandars J, Esmail A: The frequency and nature of medical error in primary care: under 53 standing the diversity across studies. Fam Pract 20 (3): 231 6, 2003. 54 55 3. Miller GC, Britth HC, Valenti L. Adverse drug events in general practice patients in Aus 56 tralia. Med J Aust. 2006;184(7):321 324. 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4. Thomsen LA, Winterstein AG, Søndergaard B, Haugbølle LS, Melander A. Systematic 4 review of the incidence and characteristics of preventable adverse drug events in ambulato 5 6 ry care. Ann Pharmacother. 2007; 41(9):1411 1426. 7 8 5. Gnädinger M, Ceschi A, Conen D, et al: Medication incidents in primary care medicine: 9 10 protocol of a study by the Swiss Federal Sentinel Reporting System. BMJ Open 04/2015; 11 5(4): e007773. DOI:10.1136/bmjopen 2015 007773. 12 13 6. Gnädinger M, Herzig L, Ceschi A et al: The Burden Related to Chronic Conditions and 14 15 MultimorbidityFor in thepeer Swiss Primary review Care Population: A onlyProspective Study in the Swiss Sen 16 17 tinel Surveillance Network (Sentinella), 2016, to be submitted. That manuscript is includ- 18 ed in the present submission. 19 20 7. Evans DC, Cook CH, Christy JM et al: Comorbidtiy polypharmacy scoring facilitates out 21 come prediction in older trauma patients. J Am Geriatr Soc 60: 1465 70, 2012. 22 23 24 8. Cohidon C, Cornuz C, Senn N: Primary care in Switzerland: evolution of physicians‘ pro 25 file and activities in twenty years (1993 2012). BMC Fam Pract 16: 107, 2015. 26 27 9. Makeham M, Dovey S, Runciman W, et al: Methods and Measures used in Primary Care 28 Patient Safety Research. Results of a literature review. WHO report on patient safety Qual 29 30 Safety Health Care 2008. 31 32 10. Makeham MAB, Kidd MR, Saltman DC et al: The threats to Australian patient safety
33 (TAPS) study: incidents of reported errors in general practice. Med J Aus 185: 95 98, 2006. http://bmjopen.bmj.com/ 34 35 11. Kuo GM, Phillips RL Graham D et al: Medication errors reported by US family physicians 36 and their office staff. Qual Saf Health Care 17: 286 90, 2008. 37 38 12. Rees P, Edwards A, Panesar S et al: Safety incidents in the primary care office setting. 39 Pediatrics 135 (6): 1027 35, 2015. 40
41 on September 28, 2021 by guest. Protected copyright. 42 13. Gandhi TK, Weingart SN, Borus BA, et al: Adverse drug events in ambulatory care. N 43 Engl J Med 348: 1556 64, 2003. 44 45 14. Runciman W, Hibbert P, Thomoson R et al: Towards an international classification ofr 46 47 patient safety: key concepts and terms. Int J Qual Health Care 21 (1): 18 26, 2009. 48 15. Makeham MAB, Dovey SM, County M et al: An international taxonomy for errors in gen 49 50 eral practice: a pilot study. Med J Aus 177:68 72; 2002. 51 52 16. Wetzels R, Wolters R, van Weel C et al: Mix of methods is needed to identify adverse 53 events in general practice: a prospective observational study. BMC Fam Pract 2008; 9:35. 54 55 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 17. Gehring K, Schwappach DLB, Battaglia M et al: Frequency of and harm associated with 4 primary care safety incidents. Am J Managed Care 18 (9): e323 7, 2012. 5 6 18. Mira JJ, Orozco-Beltran D, Perez-Jover V et al: Physician patient communication failure 7 8 facilitates medication errors in older polymedicated patients with multiple comorbidities. 9 Fam Pract 30: 56 63, 2013. 10 11 12 19. Cooper JA, Moriarty F, Ryan C et al: Potentially inappropriate prescribing in two popula 13 tions with differing socio economic profiles: a cross sectional database study using the 14 15 PROMPTFor criteria. peer Eur J Clin Pharmacol review 72: 583 91, 2016. only 16 17 20. Dreischulte T, Donnan P, Grant A et al: Safer prescribing – a trial of education, informat 18 ics, and financial incentives. New Engl J Med 374: 1053 64, 2016. 19 20 21. O’Grady I, Gerrett D: Minimising harm from missed drug doses. Nursing Times 111 (44): 21 22 12 15, 2015. 23 22. Huges RG, Ortiz E: Medication errors. Why they happen, and how they can be prevented. 24 25 Am J Nursing 205 (3): 14 24, 2005. 26 23. Pirmohamed M, James S, Meakin S, et al: Adverse drug reactions as cause of admission 27 28 to hospital: prospective analysis of 18 820 patients. BMJ 2004; 329:15–19. 29 24. Livio F, Buclin T, Yersin B et al: Hospitalisations pour effet indésirable médicamenteux. 30 31 Recensement prospectif dans un service d’urgences médicales [French]. Raisons de Santé 32 23, 1998.
33 http://bmjopen.bmj.com/ 34 35 25. Berdot S, Gillaizeau F, Caruba T et al: Drug administration errors in hospital inpatients: a 36 systematic review. PLoS ONE 8 (6): e68856, 2013; doi: 10.1371/journal.pone.0068856. 37 38 26. Makary M, Daniel M : Medical error – the third leading cause of death in the US. BMJ 39 40 2016 ; 353: i2139.
41 27. Nobili A, Pasina L, Tettamanti M et al: Potentially severe drug interactions in elderly out on September 28, 2021 by guest. Protected copyright. 42 43 patients: results of an observational study of an administrative prescription database. J Clin 44 Pharm Ther 34: 377 86, 2009. 45 46 47 28. Avery AJ, Ghaleb M, Barber N, et al: The prevalence and nature of prescribing and moni 48 toring errors in English general practice: a retrospective case note review. Br. J Gen Pract 49 50 e543, 2013. 51 52 29. Field TS, Gurwitz JH, Avorn J et al: Risk factors for adverse drug events among nursing 53 home residents. Arch Intern Med 161: 1629 34, 2001. 54 55 30. Field TS, Mazor KM, Briesacher B, et al: Adverse drug events resulting from patient er 56 rors in older adults. J Am Geriatr Soc 55: 271 6, 2007. 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 31. Saedder EA, Brock B, Nielsen LP et al: Identifying high risk medication: a systematic lit 4 erature review. Eur J Clin Pharmacol 70: 637 45, 2014. 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 List of figures and tables 4 5
6 7 8 Figure 1: Study flow chart 9 10 Figure 2: Distribution of the number of cases reported by practice 11 12 13 Figure 3: Type of error 14 15 Figure 4: OrganFor system involvedpeer review only 16 17 18 Table 1: Characteristics of the reporting physicians in 2015 19 20 Table 2: General description of the cases 21 22 23 Table 3: Disturbances after the incident 24 25 Table 4: Possible risk factors for incident as compared to a denominator analysis during calendar 26 27 weeks 11 and 12 [Gnädinger M 2016] 28 29 Table 5: ACT Groups of suspected medications 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 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 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Physicians’ gender 5 male 128 (71.1%) 6 female 52 (28.9%) 7 Physicians’ age class 8 < 40 12 (6.7%) 40 49 44 (24.4%) 9 50 59 66 (36.7%) 10 60 and over 58 (32.2%) 11 Specialty 12 GP 148 (82.2%) 13 PED 32 (17.8%) 14 Number of physicians in practice (n=144) 15 1 For peer review only69 (47.8%) 16 2 39 (27.1%) 17 3 17 (11.8%) 18 4 to 5 8 (5.6%) 19 6 to 9 7 (4.9%) 20 10 and over 4 (2.8%) Number of physicians per practice reporting to Sentinella (n=144) 21 1 119 (82.6%) 22 2 19 (13.2%) 23 3 5 (3.5%) 24 8 1 (0.7%) 25 Linguistic region 26 German 122 (67.8%) 27 French 44 (24.4%) 28 Italian 14 (7.8%) 29 Urbanity of the practice 30 urban 93 (51.7%) 31 agglomeration 60 (33.3%) 32 rural 27 (15.0%) Workload per week 33 http://bmjopen.bmj.com/ < 15 h 9 (5.0%) 34 15 30 h 36 (20.0%) 35 > 30 h 135 (75.0%) 36 Drug distribution system 37 dispensing by physician 73 (42.2%) 38 mixed system 19 (10.6%) 39 dispensing by pharmacy 85 (47.2%) 40 Electronic documentation yes 89 (49.4%)
41 on September 28, 2021 by guest. Protected copyright. 42 no 91 (50.6%) 43 Electronic interaction control 44 yes 65 (36.1%) 45 no 115 (63.9%) 46 Electronic prescription yes, with thesaurus 62 (34.4%) 47 yes, but without thesaurus 24 (13.3%) 48 none 94 (52.2%) 49 Certification of the practice 50 yes 46 (25.6%) 51 none 134 (74.4%) 52 Staff meetings 53 yes, at least monthly 69 (38.3%) 54 yes, but less frequently 70 (38.9%) 55 none 41 (22.8%) 56 Quality circle participation 57 yes, at least monthly 134 (74.4%) 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 yes, but less frequently 23 (12.8%) 4 none 23 (12.8%) 5 6 Table 1. Characteristics of the reporting physicians in 2015 (144 practices, 180 physicians) 7 . 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Relevance All 5 6 7 less more 8 9 10 Number of cases 124 73 197 11 12 13 Patients age 69.2 ± 20.6 69.4 ± 21.2 69.3 ± 20.8 14 15 For peer review only 16 Patients gender, 17 % males 40.3 32.9 37.6 18 19 20 Physicians specialty 21 % pediatricians 1.6 1.4 1.5 22 23 24 Linguistic region, % 25 26 German 75.0 68.5 72.6 27 French 18.5 28.8 22.3 28 Italian 6.5 2.7 5.1 29 30 31 Physician to patient relationship, % 32 own family physician 86.3 78.1 83.2 33 http://bmjopen.bmj.com/ 34 urgency / holiday replacing 0.8 4.1 2.0 35 institution physician 11.3 17.8 13.7 36 37 other 1.6 0.0 1.0 38 39 40 Observer of the incident; % physician / practice staff 50.0 50.7 50.3 41 on September 28, 2021 by guest. Protected copyright. 42 patient / proxies 21.8 23.3 22.3 43 44 community nurse 1.6 4.1 2.5 45 institution (where patient lives) 15.3 16.4 15.7 46 hospital 0.8 1.4 1.0 47 48 other physicians 2.4 0.0 1.5 49 pharmacist 7.3 4.1 6.1 50 51 other 0.8 0.0 0.5 52 53 54 Table 2. General description of the cases. Differences between “less” and “more” categories are 55 not statistically significant. We calculated the variable “incident relevance” from the variables “dis 56 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 turbance” and “endangering” of the patients; if any of them was graded with “medium” or higher, 4 the variable relevance was set to “more”, otherwise to “less”. 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Item N percent 4 5 6 Severity of disturbance 7 8 no symptoms but pathological laboratory tests 15 8.0 9 light 44 23.4 10 11 moderate 22 11.7 12 severe 10 5.3 13 14 fatality 0 0.0 15 Subtotal (thisFor is the base peerof the next rows) review only 91 48.4 16 17 no symptoms, normal (or no) laboratory tests 97 51.6 18 19 Total 188 100.0 20 Missing data 9 n.a. 21 22 All patients 197 n.a. 23 24 25 Time until recovery 26 hours 26 28.5 27 28 days 41 45.1 29 weeks 15 16.5 30 31 not yet known or missing information 9 9.9 32 All patients with disturbances 91 100.0
33 http://bmjopen.bmj.com/ 34 35 Recovering 36 37 without sequels 78 85.8 38 with light to moderate sequels 2 2.2 39 40 with severe sequels or fatality* 5 5.4
41 not yet known or missing information 6 6.6 on September 28, 2021 by guest. Protected copyright. 42 43 All patients with disturbances 91 100.0 44 45 46 Treatment / surveillance 47 not needed 48 52.7 48 49 ambulatory care 33 36.3 50 hospital care** 7 7.7 51 52 missing information 3 3.3 53 All patients with disturbances 91 100.0 54 55 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Table 3. Disturbances after the incident. * In one case, there was a reduced kidney 4 5 function, the other cases remained unclear, and no fatalities were reported. ** Two cases 6 had to be surveilled in the emergency room, the hospital stays were: intoxications with 7 8 thiethylperazine, with fenoterol plus ipratropium bromide, with zolpidem, further a derailed 9 10 diabetes type 2 (after missed treatment with metformin), and a gastro intestinal hemor 11 rhage in a patient where antithrombotic treatment with rivaroxaban was not communicated 12 13 to the physician. 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Item Patient group OR (95% CI) OR (95% CI) 5 (crude) (adjusted) 6 7 Denominator Incidents 8 9 10 Number of observations 26,852 197 / / 11 12 Age (mean±SD), years 46.7 ± 27.5 69.3 ± 20.8 1.004 (1.001;1.006)**a 1.001 (0.996;1.005) 13 14 15 Gender, For peer review only 16 ● male 47.0 37.6 1 1 17 ● female 53.0 62.4 1.048 (0.916;1.197) 1.055 (0.878;1.196) 18 19 20 Care dependency, 21 d 22 number (%)* 23 ● none 16,335 (85.5%) 96 (51.6%) 1 1 24 ● yes, by proxies 954 (5.0%) 12 (6.5%) 1.121 (0.789;1.594) 0.979 (0.674;1.423) 25 a ● yes, by community nurse 723 (3.8%) 22 (11.8%) 1.458 (1.025;2.073) 1.201 (0.821;1.758) 26 1,099 (5.8%) 56 (30.1%) 1.802 (1.399;2.323)c 1.528 (1.141;2.046) a 27 ● yes, by institution 28 29 30 c Number of conditions 2 (0;4) 5 (3;7) 1.052 (1.029;1.075)** 1.030 (0.994;1.067)** 31 32 (median, IQR)
33 http://bmjopen.bmj.com/ 34 Number of chronic active 1 (0;4) 6 (3;9) 1.052 (1.030;1.074)**c 1.030 (0.995;1.067)** 35 36 treatments (median, IQR) 37 38 Evans’ Index 3 (0;8) 11 (6;17) 1.009 (1.005;1.013)** c n.a.*** 39 40 (median, IQR)
41 on September 28, 2021 by guest. Protected copyright. 42 Thurgau Morbidity Index d 43 d 44 value (%) 45 ● 0 8,463 (31.5%) 24 (12.2%) 1 1 46 ● 1 3,611 (13.4%) 8 (4.1%) 0.989 (0.787;1.242) 0.908 (0.694;1.190) 47 ● 2 4,102 (15.3%) 23 (11.7%) 1.049 (0.847;1.300) 0.898 (0.685;1.169) 48 3,877 (14.4%) 39 (19.8%) 1.131 (0.914;1.399) 0.830 (0.611;1.127) 49 ● 3 a 50 ● 4 2,119 (7.9%) 39 (19.8%) 1.292 (1.004;1.662) 0.901 (0.643;1.265) b 51 ● 5 1,539 (5.7%) 38 (19.3%) 1.420 (1.078;1.868) 0.823 (0.547;1.239) 52 709 (2.6%) 26 (13.2%) 1.680 (1.178;2.396)c 0.866 (0.523;1.436) 53 ● 6 18 0 54 missing values 55 56 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Table 4. Possible risk factors for incident as compared to a denominator analysis during 4 calendar weeks 11 and 12 [Gnädinger M 2016]. * Because of question ambiguity this analysis 5 6 was restricted to adult patients (age > 19 years); this led to 19,812 valid observations in the de 7 nominator and 183 in the incident groups. ** Per one conditions, medication, year or index point. 8 9 *** Because Evans’ Index is a composite of condition and medication numbers; it was not included 10 in the multiple regression analysis. Significance levels: a p<0.05, b p<0.01, c p<0.001. d Because 11 12 GENLIN procedure was not able to process ordinal scaled variables, correlations between study 13 14 group and TMI or care dependency were tested with Spearman’s rho: the correlation coefficients 15 were +0.075For or +0.094, respectively,peer p<0.001. review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 ATC Class All incidents Incidents with prob- Relative Risk Swiss 2015 sales* 4 able or definite (percentages left (number of packages) 5 relationship with by right column) 6 7 medication 8 9 A Alimentary tract and metabolism 28 (14.6%) 6 (15.8%) 1.06 31,455,252 (14.9%) 10 11 12 B Blood and blood forming organs 23 (12.0%) 7 (18.4%) 7.08 5,507,624 (2.6%) 13 14 C Cardiovascular system 44 (22.9%) 1 (2.6%) 0.35 16,027,143 (7.5%) 15 For peer review only 16 17 D Dermatologics 1 (0.5%) 0 (0.0%) 0.00 17,314,810 (8.2%) 18 19 G Genito urinary system and sex 1 (0.5%) 0 (0.0%) 0.00 7,936,641 (3.7%) 20 hormones 21 22 23 H Systemic hormonal preparations 7 (3.6%) 1 (2.6%) 2.00 2,875,760 (1.3%) 24 (excluding sex hormones and 25 insulins) 26 27 28 J Anti infectives for systemic use 23 (12.0%) 6 (15.8%) 3.95 8,444,623 (4.0%) 29
30 K Infusion liquids 0 (0.0%) 0 (0.0%) 0.00 24,158,749 (11.5%) 31 32
33 L Anti neoplastic and immunomodu 5 (2.6%) 1 (2.6%) 2.89 1,934,950 (0.9%) http://bmjopen.bmj.com/ 34 lating agents 35 36 37 M Musculo skeletal system 4 (2.1%) 2 (5.3%) 0.76 14,787,413 (7.0%) 38 39 N Nervous system 43 (22.4%) 10 (26.3%) 1.30 42,690,195 (20.2%) 40
41 on September 28, 2021 by guest. Protected copyright. 42 P Anti parasitic products, insecti 1 (0.5%) 0 (0.0%) 0.00 462,559 (0.2%) 43 cides and repellents 44 45 R Respiratory system 7 (3.6%) 3 (7.9%) 0.57 28,837,468 (13.7%) 46 47 48 S Sensory organs 2 (1.0%) 0 (0.0%) 0.00 7,658,311 (3.6%) 49 50 51 T Diagnostic use 0 (0.0%) 0 (0.0%) 0.00 43,184 (0.0%) 52 53 V Various 3 (1.6%) 1 (2.6%) 6.50 855,707 (0.4%) 54 55 56 Total 192 (100.0%) 38 (100%) 1.0 210,990,389 (100.0%) 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 does not apply 5 0 n.a. n.a. 4 5 6 Table 5. ACT-Groups of suspected medications. Relative risk of drugs with probable or definite 7 relationship with the incident as compared to sales proportions. * Information by Interpharma Swit 8 9 zerland (Appendix F). 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Study flow chart
33 http://bmjopen.bmj.com/ 34 254x190mm (96 x 96 DPI) 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 Distribution of the number of cases reported by practice 36 37 340x283mm (96 x 96 DPI) 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 30 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 Type of error 36 348x288mm (96 x 96 DPI) 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 31 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 Organ system involved 35 36 356x290mm (96 x 96 DPI) 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 32 of 68
1 2 3
4 Appendix A Incident questionnaire BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 Medication Incidents in Primary Care (MIPC) 6 (Version 1.0, January 25th 2015) (Internet version not available in English) 7 8 9 10 11 Incident reporting form 12 13 14 15 16 Administrative Information 17 For peer review only 18 19 20 21 1. Sentinella identification number: 2. Week of reporting: 22 23 24 25 26 27 28 The Patient 29 30 31 32 33 3. Year of birth: 4. Gender: m f 34 35 36 37 5. What was your relationship to the patient when the incident happened? Were you the http://bmjopen.bmj.com/ 38 39 40 family physician emergency / substitute physician institution physician other 41 if other what kind………………………. 42 43 44 45 46 6. What is the patient’s living situation? on September 28, 2021 by guest. Protected copyright. 47 48 with partner / family alone institution unknown 49 50 51 52 53 7. Are there social problems? 54 55 yes no unknown 56 57 58 59 8. Is the patient demented or otherwise mentally handicapped? 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 68 BMJ Open
1 2 3 yes no unknown 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 8 9. Does he suffer from psychological problems? 9 10 11 yes no unknown 12 13 14 15 10. Does he take psychotropic drugs (esp. antidepressants, neuroleptics, benzodiazepines, 16 opiates)? 17 For peer review only 18 19 yes no unknown 20 21 22 23 11. Are there linguistic problems? 24 25 26 yes no unknown 27 28 29 30 12. Does he or she smoke? 31 32 33 yes no unknown 34 35 36 37 13. Is there substance abuse (other than nicotine)? http://bmjopen.bmj.com/ 38 39 yes no unknown , if yes, what substance? …………………….. 40 41 42 43 44 14. Does the patient have uncorrected / uncorrectable visual impairment? 45 on September 28, 2021 by guest. Protected copyright. 46 yes no unknown 47 48 49 50 51 15. Does the patient have uncorrected / uncorrectable hearing impairment? 52 53 54 yes no unknown 55 56 57 58 16. Does the patient have uncorrected / uncorrectable mobility impairment? 59 yes no unknown 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 68
1 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 17. Is there renal insufficiency (GFR: <60 ml/min/1.73 m2)? 7 8 yes no unknown 9 10 11 12 13 18. Is there hepatic insufficiency or liver cirrhosis? 14 15 16 yes no unknown 17 For peer review only 18 19 20 21 19. Was the patient hospitalized in the past 12 months? 22 yes no unknown 23 24 25 26 20. Is the patient taken care of by others? (only one answer permitted)? 27 28 29 yes, family / proxies yes, community nurse yes, institution no unknown 30 31 32 33 21. Number of regularly applied active substances (including non-daily applied ones, see 34 guidelines)? 35 36 37 unknown http://bmjopen.bmj.com/ 38 39 40 41 42 22. Number of chronic conditions (see guidelines)? 43 44 unknown 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 23. Scale value of „Thurgau Morbidity Index” (chronic part, see guidelines)? 50 51 unknown 52 53 54 55 56 57 58 59 60
3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 68 BMJ Open
1 2 3 Details of the incident
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 24. Please, give a short description of the incident (in block letters): 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 25. Who observed the incident? (multiple answers possible)? 23 24 Physician / staff patient / relatives community nurse home / institution hospi- 25 tal 26 27 other physicians pharmacist other unknown 28 for „other” please specify: …………………………… 29 30 31 32 33 26. What happened (multiple answers possible)? 34 35 dosage too high 36 dosage too low 37 http://bmjopen.bmj.com/ 38 application too short 39 application too long 40 41 wrong administration route 42 43 wrong medication 44 indicated medication not received 45 46 expired / defective medication on September 28, 2021 by guest. Protected copyright. 47 problems with insurance reimbursement 48 49 50 unclear / undefined 51 52 other (please specify): ……………………… 53 54 55 56 57 58 59 60
4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 68
1 2 3 27. Please state the trade name of the medication used in the incident:
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 ……………………………………………….. 6
7 8 9 28. Please note other medication names, presuming they are relevant to the case. 10 11 12 13 14 15 none unknown 16 17 For peer review only 18 19 29. How would you judge the degree of hazard to the patient during the incident? 20 21 mild medium severe none does not apply unknown 22 23 24 25 30. How intense was the impairment (as judged by the patient)? 26 27 28 mild medium severe fatal no symptoms, but pathological lab values 29 30 no impairment does not apply unknown 31 If there wasn’t any impairment, please skip to question 34. 32 33 34 35 31. How long did the impairment last? 36 37 hours days weeks longer unknown http://bmjopen.bmj.com/ 38 39 40 41 42 32. How was the recovery? 43 without residues with mild residues with severe residues / fatal unknown 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 33. Which organ system was affected (multiple answers possible)? 49 50 cardiovascular 51 central nervous 52 53 gastro-enteral 54 kidneys 55 56 liver 57 lung 58 59 skin 60 other, please specify …………………. 5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 68 BMJ Open
1 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 34. Did you have to apply a specific surveillance or treatment for the incident? 7 8 yes, ambulatory yes, hospital no unknown 9 If yes, please specify which? …………………………………………….. 10 11 12 13 14 35. What factors contributed causally to the emergence of the incident (multiple answers pos- 15 16 sible)? 17 off duty hours For peer review only 18 19 20 communication failure within practice 21 generic substitution by pharmacist 22 23 24 hand-written prescription incorrectly interpreted 25 conflicting multiple prescriptions 26 27 lack of alertness within practice 28 29 lack of documentation 30 insufficient patient instruction 31 32 lack of aids (e.g. Dosette®) 33 lack of cooperation by patient / relatives 34 35 misleading package leaflet information 36 patient’s internet search 37 http://bmjopen.bmj.com/ 38 administrative problems 39 40 manufacturer related (defective medication) 41 distributer related (out of stock) 42 43 lack of maintenance (e.g. first aid kit) 44 other, please specify……………………….. 45 46 unknown on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 36. Was there an interface problem? If yes, which (multiple answers possible)? 52 53 yes, with hospital 54 55 yes, with institution 56 yes, with community nurse 57 58 yes, with pharmacist 59 60
6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 68
1 2 3 yes, with specialist physician 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 yes, with other, please specify? …………… 6 7 8 no 9 10 If no, please skip to question 38. 11 12 13 37. Was there an explicit comparison of prescription lists with the institution / person? 14 15 16 yes, verbal / by phone yes, written / by fax no unknown 17 For peer review only 18 19 38/39. Was the patient informed about the incident? 20 21 22 yes, by myself / practice staff 23 no, because he was not able to understand the message (children, demented) 24 25 no, because the problem was solved and communication would have impaired confidence 26 no, because the patient had moved or was deceased 27 28 no, this was not needed because patient / relatives themselves had observed the incident 29 no, because others had already informed him 30 31 32 no, because: ………………………………………………………………………. 33 34 unknown 35 36 37 If yes, what was the patient’s reaction? …………………………………………………… http://bmjopen.bmj.com/ 38 39 40 41 40. What did you do as a result to prevent similar incidents in the future (multiple answers 42 possible)? 43 44 change standard operations procedures 45
better instruction of patients on September 28, 2021 by guest. Protected copyright. 46 47 communication with institution(s) 48 notification of manufacturer 49 50 notification of liability insurer 51 52 notification of drug authority („yellow leaflet“) 53 notification of the “critical incident reporting system” 54 55 other, please specify? ……………………….. 56 nothing 57 58 59 60
7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 68 BMJ Open
1 2 3 41. Who or what was ultimately responsible for the occurrence of the incident? 4 …………………………………………. BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 8 9 42. In the given situation, do you think one could have anticipated the event? 10 11 yes no 12 13 14 15 43. Did you already report an identical or very similar incident to this study? 16 yes no 17 For peer review only 18 19 20 21 44. Please make any suggestions about the kind of measures that could be taken to generally re- 22 duce the frequency of such events (in block letters): 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 Please keep a copy of this questionnaire in the patient files. Thank you for filling it out! 52 53
54 55 56 57 58 59 60
8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 68
1 2 3
4 Appendix B Initial questionnaire BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 Medication incidents in primary care (MIPC) 7 8 Initial reporting (Version 1.0, January 25th 2015) 9 Internet version: not available in English 10 11 1. Sentinella identification number: …………. 12 13 14 15 2. Number of physicians within your practice: …………. 16 17 For peer review only 18 3. Number of them who report to Sentinella …………. 19 20 21 22 4. Your weekly workload, h/w: <16 , 16-30 , >30 23 24 25 5. Number of hours within your practice (%): <50% , ≥50% 26 27 6. Approximate proportion of medication prescribed (as compared to directly delivered drugs): 28 <33% , 33-66% , >66% 29 30 31 7. Do you have an X-ray (machine?)? yes no 32 33 34 8. Do you have an ECG? yes no 35 36 37 9. Do you have an ultrasound? yes no http://bmjopen.bmj.com/ 38 39 40 10. Do you have an electronic system for controlling electronic drug interaction? 41 yes no 42 43 44 45 11. Do you have electronic patient history documentation? 46 yes no on September 28, 2021 by guest. Protected copyright. 47 48 49 12. Do you prescribe electronically? 50 yes, with a medication thesaurus yes, but without one (use of a typewriter) no 51 52 53 13. Is your practice certified (e.g. EQUAM)? yes no 54 55 56 14. Do you regularly schedule team sessions? 57 Yes, at least monthly , yes, but less frequently , no 58 59 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 68 BMJ Open
1 2 3 4 15. Do you attend quality circle sessions (in accordance with "Hausärzte Schweiz")? BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 yes, regularly , yes, now and then , no 6 7 8 9 16. Did you complete a special education (e.g. manual or psychosomatic medicine), or do you 10 have special interests (e.g. toxic maniac patients)? yes , no 11 if yes, please specify : ………………………………………………………………………. 12 13 14 17. Are you contracted by an institution? yes no 15 if yes, please specify (prison, home etc.): ……………………………………………………………… 16 17 18. If yes, doesFor this institution peer have specificreview problems with only medication? 18 yes , no , if yes, please specify: ………………………………………………………… 19 20 19. Are you involved in other special activities (teaching, research, insurance doctor)? 21 22 yes no 23 If yes, please specify the kind of activity: …………………………………………………….. 24 25 Thank you very much! 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 68
1 2 3
4 Appendix C Final questionnaire BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 Medication Incidents in Primary Care (MIPC) 7 8 th 9 Final questionnaire (Version 3.0 / December 14 2015) 10 11 Online version: not available in English 12 13 Dear colleagues, 14 15 Thank you for your dedicated support of our study on medication incidents in 2015, regardless of 16 whether you reported much of them or not. We would like to ask some final questions which will 17 help us put the otherFor information peer you sent review us in the right context only. 18 19 20 1. Sentinella identification number: …………… 21 22 2. Did you not report medication incidents that you had noticed during the last year (e.g. because 23 of lack of time)? 24 25 26 never or almost never , yes, but rarely , yes, frequently always or almost always 27 28 If this was frequently the case, please explain why:……………………………………….. 29 30 3. Did your practice participate in the fortnight morbidity denominator study in March 2015 (calen- 31 dar weeks 11 and 12)? 32 yes, fully , yes, but only partly (by omission of certain variables) , no 33 34 If you did not fully participate or not at all, what was the reason? 35 ……………………………………………………… 36 37 If you did not participate in the denominator study, please continue with question 12. http://bmjopen.bmj.com/ 38 39 4. How big was your effort for coding the morbidity variables of the denominator study? 40 manageable , rather big , too much , impossible 41 42 43 Did you have any difficulties when coding the morbidity variables? 44
45 5. Hospitalisation during the previous 12 months? none , a little , considerable , severe 46 on September 28, 2021 by guest. Protected copyright. 47 6. Care-dependency? none , a little , considerable , severe 48 49 7. Number of medications? none , a little , considerable , severe 50 51 8. Number of conditions? none , a little , considerable , severe 52 53 9. Thurgau Morbidity Index? none , a little , considerable , severe 54 55 10. Repeat consultation during the fortnight? none , a little , considerable , severe 56 57 If you named considerable to severe difficulties in questions 5 to 10, please list them in item 13. 58 59 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 43 of 68 BMJ Open
1 2 3 11. How long did it take for you and your practice nurse together for coding all of the variables of 4 one patient (Dr. Gnädinger needed less than 3 minutes)? BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 …………………………………………………. minutes 6 7 8 12. Would you be willing to be interviewed for a focus group on the subject of medication safety? 9 10 yes , no 11 12 13. Other comments: 13 14 15 16 17 For peer review only 18 19 20 21 22 Thank you very much! 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 44 of 68
1 2 3
4 Appendix D: electronic tables and figures BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 Figure e1: Distribution of the incident notifications over the year 6 Table e1: What went wrong with the incidents 7 Table e2: Organ system involved 8 Table e3: Causes of the incident 9 Table e4: Patient-sided possible risk factors 10 Table e5: Reactions to the incident 11 12 Table e6: Proposals to avoid further incidents 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 45 of 68 BMJ Open
1 2 3
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 Figure e1. Distribution of the incident notifications over the year. Because only the 41 week has been reported, the assignment to the month is approximate. 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 46 of 68
1 2 3
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 Type of error N* Percent*, % 6 7 8 Dosing too high 41 (12) 20.9 (6.1) 9 10 11 Dosing too low 21 (6) 10.7 (3.0) 12 13 14 Dosing too short 1 (2) 0.5 (1.0) 15 16 17 For peer review only 18 Dosing too long 5 (4) 2.6 (2.0) 19 20 21 Wrong way of administration 1 (1) 0.5 (0.5) 22 23 24 Wrong medication applied 56 (12) 28.6 (6.1) 25 26 27 Necessary medication not applied 12 (9) 6.1 (4.6) 28 29 30 Defective or expired medication applied 1 (1) 0.5 (0.5) 31 32 33 Problems with insurance reimbursing 1 (0) 0.5 (0.0) 34 35 36 37 Other problem** 31 (6) 15.8 (3.0) http://bmjopen.bmj.com/ 38 39 40 Unknown 1 (0) 0.5 (0.0) 41 42 43 Multiple naming 26 (n.a.) 13.3 (n.a.) 44 45 46 Total 197 (53) 100.0 (26.9) on September 28, 2021 by guest. Protected copyright. 47 48 49 50 Table e1. What went wrong with the incidents. * Parenthesis denotes the additional naming 51 within the category “multiple naming”. ** Naming within the category “other” were: confusion of sim- 52 ilar trade names (2), dosing error (6), erroneous package size (1), confusion of similar looking 53 preparations (1), error when controlling blood levels (2), double dosing (4), missed discontinuation 54 of medication (1), necessary treatment not applied (1), wrong vaccine applied (4), missed re-up- 55 take of anticoagulation after operation (1), contra-indication overlooked (2), drug-drug interaction 56 overlooked (3), known intolerance overlooked (3), necessary monitoring missed (2), transgression 57 of maximal drug allowance (2), delivery of incomplete package (1), confusion by contradictory 58 59 medication lists (3), incorrect handling of administering device (1), uncontrolled taking of “nature 60 medicines” (1), unreliable compliance with medication plan (2), two medications of the same drug
3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 47 of 68 BMJ Open
1 2 3 class (3), forbidden bisection of pills (2), confusion of similar named patients (1), intake of medica- 4 tion of the neighbor resident in a home (1), other (9). BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 48 of 68
1 2 3
4 Organ system N* Percent*, % BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 Cardiovascular 14 (10) 15.4 (11.0) 8 9 10 Central nervous system 23 (10) 25.3 (11.0) 11 12 13 Gastro-intestinal 5 (3) 5.5 (3.3) 14 15 16 Kidneys 2 (2) 2.2 (2.2) 17 For peer review only 18 19 20 Liver 2 (0) 2.2 (2.2) 21 22 23 Lung 2 (1) 2.2 (1.1) 24 25 26 Skin 6 (0) 6.6 (0.0) 27 28 29 Other** 24 (6) 26.4 (6.6) 30 31 32 Multiple naming 13 (n.a.) 14.3 (n.a.) 33 34 35 36 Total 91 (32) 100 (35.2) 37 http://bmjopen.bmj.com/ 38 39 Question does not apply 106 / 40 41 42 43 Table e2. Organ system involved. * Parenthesis denotes the additional naming within the category 44 45 “multiple naming”. ** Naming within the category “other” were: endocrine system (9), musculo-skel- 46 on September 28, 2021 by guest. Protected copyright. 47 etal (8), ear-nose-throat (2), psychic (1), other (9). 48 49 50 51 52 53 54 55 56 57 58 59 60
5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 49 of 68 BMJ Open
1 2 3 Causes of the incident N* Percent*;%
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 Out of hours 2 (3) 1.0 (1.5) 7 8 9 Communication problems within staff 10 (11) 5.1 (5.6) 10 11 12 Generic substitution of original trade medication by pharmacist 3 (3) 1.5 (1.5) 13 14 15 Difficulties when reading hand-written prescription 1 (1) 0.5 (0.5) 16 For peer review only 17 Multiple conflicting prescriptions 5 (8) 2.5 (4.1) 18 19 20 Lacking alertness of physician or practice staff 50 (17) 25.4 (8.6) 21 22 23 Insufficient documentation 3 (11) 1.5 (5.6) 24 25 Insufficient patient instruction 7 (17) 3.6 (8.6) 26 27 28 Lacking cooperation of patient / proxies 6 (17) 3.0 (8.6) 29 30 31 Confusion by reading package leaflet 1 (0) 0.5 (0.0) 32 33 34 Confusion after “Googleing” 1 (0) 0.5 (0.0) 35 36 Administrative problems 2 (3) 1.0 (1.5) 37 http://bmjopen.bmj.com/ 38 39 Defective medication as caused by manufacturer 3 (0) 1.5 (0.0) 40 41 42 Lacking maintenance (e.g. emergency case) 1 (0) 0.5 (0.0) 43 44 Lacking use of treatment aids (e.g. Dosett) 0 (6) 0.0 (3.0) 45 46 on September 28, 2021 by guest. Protected copyright. 47 Other source of trouble** 54 (18) 27.2 (9.1) 48 49 50 Unknown 3 (0) 1.5 (0.0) 51 52 53 Multiple naming 45 (n.a.) 23.1 (n.a.) 54 55 Total 197 (115) 100.0 (58.4) 56 57 58 Table e3. Causes of the incident. * Parenthesis denotes the additional naming within the category “multiple naming”. ** Naming within 59 the category “other” were: Erroneous delivery in pharmacy (5), treatment delayed/hampered by patient (2), transcription error (2), missed 60
6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 50 of 68
1 2 3 discontinuation of an ongoing treatment (2), communication problem within helpers’ network (16), similar trade names (2), interface prob-
4 lems with hospital (4), reading error of patient / proxies (3), erroneous execution of a medical prescription (14), missed follow-up control BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 (9), difficult handling of a preparation (2), disregard of possible interactions (2), transgression of competence by care workers (5), incom- 6 7 plete information by patient / proxies (8), stress / lack of time (12), other 25. 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 51 of 68 BMJ Open
1 2 3 Item Relevance All 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 (number of missing observations in groups “less” / n (%) 6 7 “more”) less more 8 n (%) n (%) 9 10 11 Number of patients 124 (100.0) 73 (100.0) 197 (100.0) 12 13 14 Living situation (1/0) 15 16 - together with mate / family 61 (49.6) 31 (42.5) 92 (46.9) 17 - alone For peer review28 (22.8) only 23 (31.5) 51 (26.0) 18 19 - institution 33 (27.6) 19 (26.0) 53 (27.0) 20 21 22 Social problems (2/4) 23 - yes 22 (18.1) 18 (26.1) 49 (20.9) 24 25 26 Dementia or mental illness (2/2) 27 28 - yes 31 (25.4) 18 (25.4) 49 (25.4) 29 30 31 32 Psychiatric problems (1/2) 33 34 - yes 28 (22.8) 28 (38.4)* 56 (28.9) 35 36 37 Treatment with psychotropic drugs (0/0) http://bmjopen.bmj.com/ 38 - yes 50 (40.3) 39 (53.4) 89 (44.2) 39 40 41 Linguistic problems (0/1) 42 43 - yes 11 (8.9) 5 (6.8) 16 (8.2) 44 45 46 Smoking (7/4) on September 28, 2021 by guest. Protected copyright. 47 - yes 11 (9.4) 8 (11.6) 19 (10.2) 48 49 50 Substance abuse other than nicotine (2/1) 51 52 - yes 3 (2.5) 5 (6.9) 8 (4.1) 53 54 55 Visual blurring (6/6) 56 - yes 4 (3.4) 4 (6.0) 8 (4.3) 57 58 59 60
8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 52 of 68
1 2 3 Hearing problems (3/5) 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 - yes 8 (6.6) 4 (5.9) 12 (6.3) 6 7 8 Gait problems (2/2) 9 10 - yes 40 (32.8) 18 (25.4) 58 (30.1) 11 12 13 Renal insufficiency (GFR<60 ml/min*1.73m2) (6/2) 14 - yes 24 (20.3) 21 (29.6) 45 (23.8) 15 16 17 For peer review only 18 19 Liver cirrhosis of other hepatic function problem 20 (3/1) 3 (2.5) 3 (4.2) 6 (3.1) 21 22 - yes 23 24 25 Table e4. Patient-sided possible risk factors. * p=0.021 “more” vs. “less” by chi-square testing. 26 We calculated the variable “incident relevance” from the variables “disturbance” and “endangering” 27 28 of the patients; if any of them was graded with “medium” or higher, the variable relevance was set to 29 30 “more”, otherwise to “less”. 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
9
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 53 of 68 BMJ Open
1 2 3 Reactions to the incident N* Percent*; % 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 Changing standard operating procedures of the practice 14 (3) 7.2 (1.5) 8 9 10 Better instruction of patients 23 (4) 11.7 (2.0) 11 12 13 Communication with other institutions 31 (3) 15.7 (1.5) 14 15 16 Notifying manufacturer 1 (2) 0.5 (1.0) 17 For peer review only 18 19 Reporting the incident to the critical incident reporting system 10 (6) 5.1 (3.0) 20 21 22 No reaction at all 55 (n.a.) 27.9 (n.a) 23 24 25 Other type of reactions to the incident** 49 (9) 24.9 (4.6) 26 27 28 Missing information 1 (0) 0.5 (0.0) 29 30 31 Multiple naming 13 (n.a.) 6.6 (n.a.) 32 33 34 Total 197 (27) 100.0 (13.7) 35 36 37 Table e5. Reactions to the incident. * Parenthesis denotes the additional naming within the category “multi- http://bmjopen.bmj.com/ 38 ple naming”. ** Naming within the category “other” were: Hire more workforce (1), arrangements with other 39 40 physicians (2), with pharmacist (4), with community nurse (4), with institution (12), with practice nurse (6), with 41 specialist (1), with patient (3), with supplier (1), sending a new medication plan (1), having regular staff meet- 42 43 ings (1), remove Digoxin 0.25 from assortment because of safety reasons (1), to not perform vaccinations in 44 absence of vaccination card (2), to clarify intolerances (2), to clarify interactions (1), to hand out an allergy card 45 46 (1), to actualize patient records (3), to apply for insurance cost credit (1), to organize follow-up controls (20), on September 28, 2021 by guest. Protected copyright. 47 other (5). 48 49 50 51 52 53 54 55 56 57 58 59 60
10
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 54 of 68
1 2 3 Proposal N* Percent*;%
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 Cross-check of medication lists 6 (14) 4.8 (11.2) 8 Patient instructions 2 (3) 1.6 (2.4) 9 10 Reduction of time pressure / stress 3 (6) 2.4 (4.8) 11 To observe adverse drug reactions also with “nature 1 (0) 0.8 (0.0) 12 13 products” 14 15 Four eyes check when dispensing medication 2 (4) 1.6 (3.2) 16 17 No medication Forwithout prescription peer review only3 (6) 2.4 (4.8) 18 19 To critically audit polymedication 3 (2) 2.4 (1.6) 20 21 To avoid similarly looking or named medication in drug 2 (3) 1.6 (2.4) 22 23 master 24 To demand medication plans to be brought to the con- 0 (4) 0.0 (3.2) 25 26 sultation 27 28 To provide medication plans routinely 2 (8) 1.6 (6.4) 29 30 In-deep checking before delivering “new” medication to 4 (1) 3.2 (0.8) 31 the patient 32 33 To provide patients with allergy / intolerance cards 2 (1) 1.6 (0.8) 34 35 To let patients themselves write their medication card 0 (3) 0.0 (2.4) 36 37 (and controlling afterwards) http://bmjopen.bmj.com/ 38 To instruct patients to come into the practice immedi- 3 (0) 2.4 (0.0) 39 40 ately after hospitalization 41 42 To broach regularly “medication safety” on staff meet- 5 (6) 4.0 (4.8) 43 ings 44 45 To provide information in patients’ native language 0 (2) 0.0 (1.6) 46 on September 28, 2021 by guest. Protected copyright. 47 To share important information about patients with prac- 0 (2) 0.0 (1.6) 48 49 tice nurse 50 To improve information flow 11 (16) 8.8 (10.4) 51 52 To organize follow-up controls 10 (28) 8.0 (22.4) 53 54 To wait for laboratory results before prescribing 1 (0) 0.8 (0.0) 55 56 Other** 12 (9) 9.6 (7.2) 57 58 59 Multiple answering 53 (n.a.) 42.4 (n.a.) 60
11
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 55 of 68 BMJ Open
1 2 3 Total 125 (118) 100.0 (94.4)
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 None 72 (n.a.) (n.a.) 6 7 8 9 Table e6. Proposals to avoid further incidents. * Parenthesis denotes the additional naming 10 11 within the category “multiple naming”. ** Among naming within the category “other” were: timely 12 13 involving of community nurse, stopping of self-medication in multimorbid or poly-medicated patients, 14 to clearly labelling desensitization suspensions, to separating adult and pediatric vaccines, to imple- 15 16 menting timely new guidelines, to refusing to be in care of a patient with bad compliance, to clearly 17 For peer review only 18 separating and labelling of medication prepared for different patients, to advice patients for separat- 19 ing short- and long-acting insulins at their home, to providing short-time prescriptions with a clear 20 21 expiration date, to allowing only adequately educated people to align and deliver medication in 22 homes for the elderly and that delivering may be done by the same person who did the preparing of 23 24 medication boxes, to avoiding similar looking medication boxes being muddled-up by patients at the 25 26 refectory room. 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
12
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 56 of 68
1 2 3
4 Appendix E BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 7 8 Item 9 No Recommendation 10 Title and 1 (a) Indicate the study’s design with a commonly used term in the title or the ab- 11 abstract stract [yes] 12 13 (b) Provide in the abstract an informative and balanced summary of what was 14 done and what was found [yes] 15 16 Introduction 17 For peer review only Background/ra- 2 Explain the scientific background and rationale for the investigation being re- 18 tionale ported [yes] 19 20 Objectives 3 State specific objectives, [yes] including any prespecified hypotheses [no] 21 22 23 Methods 24 Study design 4 Present key elements of study design early in the paper [yes] 25 26 Setting 5 Describe the setting, locations, and relevant dates, including periods of recruit- 27 ment, exposure, follow-up, and data collection [yes] 28 29 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of par- 30 ticipants [yes] 31 32 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and 33 effect modifiers. [yes] Give diagnostic criteria, if applicable [no] 34 35 Data sources/ 8* For each variable of interest, give sources of data and details of methods of 36 measurement assessment (measurement). Describe comparability of assessment methods if 37 there is more than one group [yes] http://bmjopen.bmj.com/ 38 39 Bias 9 Describe any efforts to address potential sources of bias [yes] 40 41 Study size 10 Explain how the study size was arrived at [yes] 42 43 Quantitative va- 11 Explain how quantitative variables were handled in the analyses. If applicable, 44 riables describe which groupings were chosen and why [yes] 45 46 on September 28, 2021 by guest. Protected copyright. Statistical me- 12 (a) Describe all statistical methods, including those used to control for con- 47 thods founding [yes] 48 49 (b) Describe any methods used to examine subgroups and interactions [yes] 50 51 52 (c) Explain how missing data were addressed [yes] 53 54 (d) If applicable, describe analytical methods taking account of sampling strat- 55 egy [no] 56 57 (e) Describe any sensitivity analyses [no] 58 59 Results 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 57 of 68 BMJ Open
1 2 3 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers poten-
4 tially eligible, examined for eligibility, confirmed eligible, included in the study, BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 completing follow-up, and analysed [no] 6 7 (b) Give reasons for non-participation at each stage [yes] 8 9 (c) Consider use of a flow diagram [yes] 10 11 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) 12 and information on exposures and potential confounders [yes] 13 14 (b) Indicate number of participants with missing data for each variable of inter- 15 est [yes] 16 17 For peer review only Outcome data 15* Report numbers of outcome events or summary measures [yes] 18 19 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted esti- 20 mates and their precision (eg, 95% confidence interval). Make clear which con- 21 founders were adjusted for and why they were included [yes] 22 23 (b) Report category boundaries when continuous variables were categorized 24 [yes] 25 26 27 (c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful time period [no] 28 29 30 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and 31 sensitivity analyses [yes] 32 33 Discussion 34 Key results 18 Summarise key results with reference to study objectives [yes] 35 36 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or 37 imprecision. Discuss both direction and magnitude of any potential bias [yes] http://bmjopen.bmj.com/ 38 39 Interpretation 20 Give a cautious overall interpretation of results considering objectives, limita- 40 tions, multiplicity of analyses, results from similar studies, and other relevant 41 evidence [yes] 42 43 Generalisability 21 Discuss the generalisability (external validity) of the study results [yes] 44 45 46 Other information on September 28, 2021 by guest. Protected copyright. 47 Funding 22 Give the source of funding and the role of the funders for the present study 48 and, if applicable, for the original study on which the present article is based 49 [yes] 50 51 52 *Give information separately for exposed and unexposed groups. 53 54 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 55 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 56 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 57 58 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is avail- 59 able at www.strobe-statement.org 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 58 of 68
1 2 3 Appendix F 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 Swiss Pharmaceutical Market: Number of packaged sold by ATC-groups, 2015 7 8 9 10 11 12 2015 13 number of packa-
14 ged sold 15 Total 210,992,389 16 A Alimentäres System und Stoffwechsel 31,455,252 17 A01 Stomatologika,For Medizinische peer Präparate zurreview Mund- und Zahnpflege only 1,473,520 18 A02 Antacida-Ulcustherapheutika, Antiflatulentia 6,051,251 19 A02A Antacida,Antiflatulentia 1,500,822 20 21 A02B Ulcustherapeutika 4,426,496 22 A02C Sonstige Magentherapeutika 123,933 23 A03 Produkte gegen funktionelle Magen-Darm-Störungen 2,087,299 24 A03A Antispasmodika+Anticholinergika rein 595,861 25 A03C Antispasmodika/Ataraktika-Kombinationen 25,411 26 27 A03E Antispasmodika, sonstige Kombinationen 222,102 28 A03F Gastroprokinetika 1,233,461 29 A03G Modulatoren der gastrointestinalen Sensomotorik 10,464 30 A04 Antiemetica und Antinausea 920,841 31 A05 Cholagoga und Leberschutz-Mittel 337,970 32 33 A05A Gallentherapeutica und Cholagoga 254,744 34 A05B Leberschutzpräparate 83,226 35 A06 Mittel gegen Verstopfung und Darmreinigung 4,583,118 36 A06A Mittel Gegen Verstopfung 4,329,037 37 A06B Darmreinigungsmittel 254,081 http://bmjopen.bmj.com/ 38 39 A07 Antidiarrhoica, Elektrolytzufuhr und intestinale Antiphlogistica 3,005,035 40 A07B Intestinale Adsorbierende Antidiarrhoica 228,076 41 A07E Intestinale Antiphlogistica 222,680 42 A07F Antidiarrhoica Micro-Organismen 1,340,432 43 A07G Orale Elektrolyt-Zufuhr 107,030 44 45 A07H Motilitätshemmer 1,095,216 46 A07X Übrige Antidiarrhoica inkl. A07A Antidiarrhoica intestinale Antiinfectiva 11,601 on September 28, 2021 by guest. Protected copyright. 47 A08 Antiadiposita excl. Diätetica 51,654 48 A09 Digestiva und Enzyme 804,667 49 A10 Antidiabetica 3,376,580 50 51 A10C Humaninsulin und Analoga 1,078,495 52 A10H Sulfonylharnstoff Antidiabetika 321,219 53 A10J Biguanid Antidiabetika 1,217,578 54 A10K Glitazone Antidiabetika 25,305 55 A10M Glinide Antidiabetika 38,782 56 57 A10N DPP-IV Inhibitor Antidiabetika 483,559 58 A10P SGLT2-Hemmer Antidiabetika 45,927 59 A10S GLP-1 Agonisten Antidiabetika 156,656 60 A10X Übrige Antidiabetika inkl. Insulin tierischen Ursprungs und Alphagluko- 8,616 sidaseinhibitor Antidiabetika 1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 59 of 68 BMJ Open
1 2 3 A11 Vitamine 5,001,408
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from A11A Multivitamine mit Mineralstoffen 755,243 5 6 A11B Multivitamine ohne Mineralstoffe 17,507 7 A11C Vitamin A+D inkl. einfache Kombinationen 2,618,699 8 A11D Vitamin B1 und Kombinationen 170,884 9 A11E Vitamin B-Komplex 739,860 10 A11F Vitamin B 12 rein 123,445 11 12 A11G Vitamin C, inkl. Kombinationen mit Mineralstoffen 353,644 13 A11X Sonstige Vitamine 222,126 14 A12 Mineralverbindungen 3,490,891 15 A12A Mineralverbindung Calcium 1,254,334 16 A12B Mineralverbindung Kalium 200,438 17 For peer review only 18 A12C Sonstige Mineralverbindungen 2,036,119 19 A13 Tonica und Roborantia 229,735 20 A16 Übrige Präparate des alimentären und Stoffwechsel-Systems 34,295 21 B Blut und Blutbildende Organe 5,507,624 22 B01 Antithrombotika 4,009,465 23 24 B01A Vitamin-K-Antagonisten 399,963 25 B01B Heparine 714,872 26 B01C Thrombozytenaggregationshemmer 2,265,838 27 B01D Fibrinolytika 14,293 28 B01X Direkte Thrombin-Inhibit und direkte Faktor XA-Hemmber 598,766 29 30 B02 Blutgerinnungssystem – Sonstige Produkte 237,495 31 B02B Antagonisten (Antidot.Anticoag) 146,201 32 B02D Blutcoagulation 1,805 33 B02G Glitazone-Antidiabetika 48,969 34 B02X Übrige Blutgerinnungsprodukte (Tissues sealing preparations, Throm- 40,520 35 bopoietin Agonisten, Antifibrinolytica synthetisch etc.) 36 B03 Antianämica 1,259,896 37 B03A Antianämica mit Eisen und Kombinationen 862,744 http://bmjopen.bmj.com/ 38 39 B03C Erythropoietin-Präparate 67,258 40 B03X Sonstige Antianämica inkl. Folsäure/Folinsäure 329,894 41 C Herz-Kreislauf Therapie 16,027,143 42 C01 Cardica 954,962 43 C01B Anti-Arrhythmie-Präparate 213,952 44 45 C01C Cardiale Stimulantien excl. Herzglykoside 266,028 46 C01D Coronartherapeutica excl. Calcium-Antagonisten und Nitropräparate 121,815 on September 28, 2021 by guest. Protected copyright. 47 C01E Nitropräparate und Analog 228,438 48 C01X Positive Inotrope, Herzglykoside und übrige Herzpräparate 124,729 49 C02 Antihypertonica 129,957 50 51 C03 Diuretica 1,815,738 52 C04 Cerebrale und periphere Vasotherapeutika 390,255 53 C05 Antivaricosa/Antihaemorrhoidalia 2,271,175 54 C05A Antihaemorrhoidalia topisch 773,800 55 C05B Antivaricosa topisch 1,029,010 56 57 C05C Antivaricosa systemisch 468,365 58 C06 Übrige Herz+Kreislaufmittel 339,930 59 C07 Betablocker 2,398,309 60 C07A Betablocker rein 2,272,988 2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 60 of 68
1 2 3 C07B Betablocker Kombination 125,321
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from C08 Calciumantagonisten 1,157,267 5 6 C09 Stoffe mit Wirkung auf das Renin-Angiotensin-System 4,116,677 7 C09A Ace-Hemmer, rein 1,142,465 8 C09B Ace-Hemmer, Kombinationen 584,697 9 C09C Angiotensin-II-Antagonisten, rein 1,084,218 10 C09D Angiotensin-II-Antagonisten, Kombinationen 1,261,616 11 12 C09X Sonstige Stoffe mit Wirkung auf das Renin - Angiotensin - System 43,681 13 C10 Lipidregu/Antiarteriosklerotika 2,446,927 14 C10A Lipidregulatoren 2,341,963 15 C10A1 Statine (Cholesterolsynthesehemmer) 2,190,227 16 C10A2 Lipidsenker, Fibrate, Ionenaustauscher und sonstige Lipidsenker 151,736 17 For peer review only 18 C10C Lipidregulatoren kombiniert mit anderen Lipidregulatoren 104,964 19 C11 Übrige Präparate für die Herz-Kreislauf-Therapie 5,946 20 D Dermatologica 17,314,810 21 D01 Dermatologische Antimykotica 1,360,570 22 D02 Emollientia und Hautschutzmittel 2,703,273 23 24 D03 Wundheilmittel 4,389,791 25 D04 Antipruriginosa inkl. Antihistaminica und Anästhetica usw. 1,458,276 26 D05 Nichtsteroidale Präparate gegen entzündliche Hauterkrankungen 327,245 27 D06 Topische antibakterielle Produkte und antivirale Mittel 864,829 28 D06A Topische antibakterielle Produkte 379,013 29 30 D06D Topische Produkte gegen Virusinfektionen 485,816 31 D07 Corticosteroide topisch 1,982,173 32 D07A Reine Corticosteroide topisch 960,182 33 D07B Corticosteroid-Kombinationen topisch 1,021,991 34 D08 Antiseptica + Desinfizientia 1,836,788 35 36 D10 Aknemittel 1,227,850 37 D10A Topische Aknemittel 829,114 http://bmjopen.bmj.com/ 38 D10B Orale Aknemittel 398,736 39 D11 Übrige Dermatologica 1,164,015 40 G Urogenital-System+Sexualhormon 7,936,641 41 42 G01 Gynaekologische Antiinfektiva 812,985 43 G01A Trichomonadenmittel 185,395 44 G01B Gynaekologische Antimykotica 475,911 45
G01X Gynaekologische antibakterielle Produkte und Antiseptika 151,679 on September 28, 2021 by guest. Protected copyright. 46 G02 Sonstige Gynaekologica 1,060,376 47 48 G02A Oxytocica 91,111 49 G02F Topische Sexualhormone 496,735 50 G02X Lokale Antikonzipientia, Prolactininhibitoren, wehenhemmende Mittel 472,530 51 und sonstige Gynaekologika 52 G03 Sexualhormone und Stimulantien des Genitalsystems 3,917,793 53 G03A Hormonale Kontrazeptiva systemisch 2,477,919 54 G03B Androgene, excl.G3E,G3F 70,852 55 G03C Oestrogene, excl.G3A,G3E,G3F 397,050 56 57 G03D Progestogene, excl. G3A,G3F 245,149 58 G03F Oestrogen-Progestogen-Kombinationen 487,291 59 G03G Gonadotropin inl. sonstige Ovulationsstimulantien 100,879 60
3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 61 of 68 BMJ Open
1 2 3 G03X Androgen-Kombinationen mit weiblichen Sexualhormonen, SERMS und 138,653
4 sonstige Sexualhormone und ähnliche Produkte BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 G04 Urologica 2,145,487 6 G04A Urologische Antiinfektiva und Harnantiseptika 410,545 7 G04C BPH Produkte (Benigne Prostata-Hyperplasie) 715,845 8 9 G04D Harninkontinenzprodukte 390,832 10 G04E Produkte gegen Erektionsstörungen 311,659 11 G04X Alle sonstigen Urologika 316,606 12 H Hormonales System, exkl. Sexualhormone 2,875,760 13 H01 Hypophysen-/Hypothalamus-Hormone 36,944 14 15 H02 Corticosteroide Systemisch 1,575,880 16 H03 Schilddruesentherapie 1,109,934 17 H04 Sonstige HormoneFor (wachstumshormone, peer review antidiuretische Hormone, only Anti-Pa- 153,002 18 rathyroid-Hormone etc.) 19 J Antiinfektiva systemisch 8,446,623 20 J01 System.Antibakter.Prod. 5,721,173 21 J01A Tetracycline und Kombinationen 155,659 22 J01C Breitspektrum Penicilline 2,353,136 23 24 J01D Cephalosporine 897,839 25 J01F Macrolide und vergleichbare Substanzen 778,785 26 J01G Fluorchinolone 817,056 27 J01H Mittel- und Schmalspektrum-Penicilline 111,484 28 J01K Aminoglykoside 46,447 29 J01P Sonstige Beta-Lactam. Antibiot. Penicilline, Cephalosporine (Monobac- 30 69,074 tame, Peneme und Carbapeneme) 31 J01X Rifamycin und vergleichbare Substanzen, Trimethoprim, Kombinationen 491,693 32 und vergleichbare Substanzen und sonstige antibakterielle Produkte 33 J02 Antimykotika, Systemisch 591,346 34 35 J04 Tuberkulosemittel 37,320 36 J05 Antivirale Mittel, systemisch 570,063 37 J05B Antivirale Mittel, exkl. Produkte gegen HIV 236,676 http://bmjopen.bmj.com/ 38 J05C Virustatika gegen HIV 333,387 39 J05C1 Nucleoside und Nucleotide Reverse Transkriptase Hemmer 108,107 40 41 J05C2 Protease Hemmer 98,581 J05CX Nicht-Nucleoside Reverse Transkriptase Hemmer, HIV Fusionshem- 42 126,699 mer, HIV-Antivirale Integrase Hemmer und sonst.HIV-Antivirale Mittel 43 44 J06 Sera und Gammaglobulin 23,164 45 J07 Vakzine 1,406,674 46 J07B Vakzine Kombinationen 463,631 on September 28, 2021 by guest. Protected copyright. 47 J07B1 Kombinationen mit Tetanusimpfstoff 387,139 48 J07BX Mehrfach-Impfstoffe mit Masern und/oder Mumps und alle anderen 76,492 49 Kombinationen 50 J07D Bakterielle Impfstoffe 164,542 51 J07E Virale Impfstoffe 687,366 52 53 J07E1 Influenza-Impfstoff 205,391 54 J07E4 Hepatitis-Impfstoff 212,503 55 J07E9 Alle uebrigen viralen Impfstoffe, inkl. HPV, Rotaviren, Varizellen etc. 269,472 56 J07X Alle uebrigen Impfstoffe und aehnliche Produkte 91,135 57 J08B Übrige Antiinfektiva 96,883 58 59 K Infusionslösungen 24,158,749 60 K01 Intravenoese Lösungen ab 100 ML 12,064,154
4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 62 of 68
1 2 3 K01A Elektrolytlösungen 3,372,213
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from K01B Standardlösungen 8,461,628 5 6 K01D Fettemulsionen, inkl. Produkte zur totalen parenteralen Ernährung 156,484 7 K01E Aminosäuren-Lösungen 51,140 8 K01X Kalorische Lös. über 10% und Lösungen zur Osmotherapie 22,689 9 K02 Plasma Expander 70,547 10 K03 Blut- und Blutersatzinfusionslösungen 1,817 11 12 K04 Injektionslösungen/Infusionszusätze 12,005,621 13 K04A Elektrolytlösungen 752,499 14 K04B Standardlösungen 9,981,186 15 K04C Kalorische Lösungen < 100 ML 34,050 16 K04D Sonstige Injektionslösungen/Infusionszusätze < 100 ML 1,237,886 17 For peer review only 18 K05 Spüllösungen 16,610 19 L Antineoplastika und Immunmodulatoren 1,934,950 20 L01 Antineoplastika 984,072 21 L01A Alkylisierende Substanzen 46,698 22 L01B Antimetaboliten 323,741 23 24 L01C Vinca-Alkaloide & sonstige pflanzliche Produkte 157,793 25 L01D Antineoplastisch wirkende Antibiotika 46,619 26 L01F Platinhalt.Antineoplast. 95,347 27 L01G Monoklonale Antikörper zur antineoplastischen Therapie 159,208 28 L01H Proteinkinasehemmer zur antineoplastischen Therapie 33,895 29 30 L01X Sonstige Antineoplastika 120,771 31 L02 Cytostatische Hormonetherapie 183,379 32 L02A Cytostatische Hormone 43,050 33 L02B Cytostatische Hormonantagonisten 140,329 34 L03 Immunostimulantien 79,294 35 36 L03A Immunstimulantien 40,990 37 L03B Interferone 38,304 http://bmjopen.bmj.com/ 38 L04 Immunsuppressiva 688,205 39 L04B Anti-TNF-Produkte 340,857 40 L04X Sonstige Immunsuppressiva inkl. Interleukin Inhibitoren 347,348 41 42 M Muskel-und Skelettsystem 14,787,413 43 M01 Antirheumatica 6,373,707 44 M01A Corticoidfreie Antirheumatica 5,700,326 45
M01C Antirheumatica spezifisch 673,381 on September 28, 2021 by guest. Protected copyright. 46 M02 Rheumaeinreibungen Rubefatientia 6,372,249 47 48 M03 Muskelrelaxantien 665,391 49 M03A Muskelrelaxantien peripher wirkend 85,757 50 M03B Muskelrelaxantien zentral wirkend 579,634 51 M04 Gichtmittel 368,020 52 M05 Übrigige Muskel-Skelettsystem-Präparate 1,008,046 53 54 N Nervensystem 42,690,195 55 N01 Anaesthetica 1,554,391 56 N01A Narkosemittel 1,078,794 57 N01A1 Narkosemittel Inhalationspräparate 20,582 58 N01A2 Narkosemittel Injektionspräparate 1,058,212 59 60 N01B Lokalanaesthetika 475,597
5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 63 of 68 BMJ Open
1 2 3 N01B1 Medizinische injizierbare Lokalanästhetika 198,964
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from N01B3 Topische Lokalanästhetika 270,962 5 6 N01B9 Sonstige Lokalanästhetika und dentale injizierbare Lokalanästhetika 5,671 7 N02 Analgetica 24,080,469 8 N02A Betäubungsmittel 1,661,607 9 N02B Analgetica und Antipyretica 21,918,098 10 N02B1 Analgetica rezeptpflichtig 8,592,792 11 12 N02B2 Analgetica rezeptfrei 13,325,306 13 N02C Migränemittel 500,764 14 N02C1 Triptane 432,741 15 N02C9 Sonstige Migränemittel 68,023 16 N03 Antiepileptica 1,957,745 17 For peer review only 18 N04 Parkinsonmittel 658,710 19 N05 Psycholeptica 9,075,955 20 N05A Antipsychotika 2,002,404 21 N05A1 Atypische Antipsychotika 1,521,618 22 N05A9 Konventionelle Antipsychotika 480,786 23 24 N05B Hypnotica und Sedativa 4,462,062 25 N05B1 Hypnotica und Sedativa, barbituratfreie Reinsubstanzen 2,830,382 26 N05B2 Hypnotica und Sedativa, barbituratfreie Kombinationen 129,461 27 N05B5 Pflanzliche Hypnotica und Sedativa 1,502,219 28 N05C Tranquilizer 2,611,489 29 30 N06 Psychoanaleptica Ex.Antiadipos 4,081,575 31 N06A Antidepressiva und stimmungsstabilisierende Produkte 3,503,509 32 N06A2 Pflanzliche Antidepressiva 245,372 33 N06A3 Stimmungsstabilisierende Produkte 50,396 34 N06A4 SSRI Antidepressiva 1,319,897 35 36 N06A5 SNRI Antidepressiva 664,070 37 N06A9 Sonstige Antidepressiva 1,223,774 http://bmjopen.bmj.com/ 38 N06B Psychostimulantia 341,172 39 N06X Psycholeptica-Psychoanaleptica-Kombinationen, Nootropica und Neuro- 236,894 40 tonica sowie verschiedene andere Produkte 41 N07 Sonstige ZNS-wirksame Produkte 1,281,350 42 N07B Raucherentwöhnungsmittel 652,146 43 N07C Produkte gegen Schwindel 286,538 44 45 N07D Anti-Alzheimer Produkte 121,928 46 N07E Alkohol- und Opiatentwöhnungsmittel 126,973 on September 28, 2021 by guest. Protected copyright. 47 N07X Alle anderen ZNS-wirksamen Präparate 93,765 48 P Parasitologie 462,559 49 P01 Antiprotozoenmittel und Anthelmintika 437,402 50 51 P01B Anthelmintica, excl. Schistosomiasis-Mittel 296,327 52 P01D Malariamittel 139,242 53 P02 Sonstige Antiparasitäre Präparate inkl. Insektizide und Repellentien 26,990 54 R Respirationssystem 28,837,468 55 R01 Rhinologica 7,246,454 56 57 R01A Rhinologica topisch 6,868,431 58 R01AX Corticosteroide Rhinologika 1,067,991 59 R01A6 Antiallergische Rhinologica 230,180 60 R01A7 Abschwellende Rhinologica 4,290,641 6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 64 of 68
1 2 3 R01A9 Sonstige Rhinologica topisch inkl. Antiinfektive Rhinologika ohne 1,279,619
4 Corticosteroide BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 R01B Rhinologica systemisch 378,023 6 R02 Halsschmerzmittel und Antiinfectiva 3,977,806 7 R03 Antiasthmatika und Produkte gegen chronisch-obstruktive Lungenerkran- 2,931,292 8 kungen 9 R03A B2-Stimulatoren 923,570 10 11 R03B Xanthine 16,188 12 R03C Degranulationshemmer 9,088 13 R03D Corticoide 258,075 14 R03F Komb. B2-Stimulatoren mit Corticoiden 1,172,896 15 R03G Anticholinergica rein und in Komb. mit B2-Stimulatoren 422,479 16 17 R03G3 AnticholinergicaFor peerrein, Inhalate review only 269,418 18 R03G4 Anticholinergica – Kombinationen mit Beta-2-Stimulantien, Inhalate 153,061 19 R03J Antileukotrien Antiasthmatika 90,062 20 R03X Sonstige Antiasthmatika und Produkte gegen chronisch-obstruktive Lun- 21 generkrankungen inkl. PDE4-Hemmer gegen Asthma / Chronisch-Obstruktive 38,934 22 Lungenerkrankungen 23 R04 Percutane Einreibemittel und Inhalationspräparate 1,093,346 24 R05 Husten- und Erkältungspräparate 10,899,026 25 R05A Erkältungspräparate 2,498,594 26 27 R05C Expectorantia ohne Antiinfectiva 4,499,364 28 R05D Hustensedativa 3,297,144 29 R05D1 Hustenmittel plain (Einzelsubstanzen) 1,971,766 30 R05D2 Sonstige Hustensedativa, inkl. Kombinationen 1,325,378 31 R05F Sonstige Husten- und Erkältungspräparate 603,924 32 33 R06 Antihistamine systemisch 2,687,483 34 R07 Sonstige Präparate des Respirationssystems 2,061 35 S Sinnesorgane 7,658,311 36 S01 Ophthalmologica 7,214,097 37 S01A Ophtalmologische Antiinfectiva 451,604 http://bmjopen.bmj.com/ 38 39 S01B Ophtalmologische Corticoide rein 271,663 40 S01C Ophthalm. Antiphlogistika und Antiinfektiva Kombinationen 613,682 41 S01E Myotika und Mittel zur Glaukombehandlung 1,038,592 42 S01F Mydriatika und Cykloplegika 46,547 43 S01G Ophthalmologische Antiallergika, Abschwellende Mittel, Antiseptika 1,473,188 44 45 S01H Ophtalmolog. Lokalanesthetika 40,383 46 S01K Künstliche Tränen und Netzmittel für Augen 2,503,750 on September 28, 2021 by guest. Protected copyright. 47 S01P Ophthalmologische Antineovaskularisationsprodukte 127,501 48 S01R Nonsteroidale entzündungshemmende Ophthalmologica (NSAID'S) 236,207 49 S01X Übrige Ophtalmologica inkl. Präparate für den Gebrauch von Kontaktlin- 50 sen, Augentonika und Augenvitamine, Präparate zur Verhütung von Katarakt und 411,009 51 Antikataraktogenika, Ophthalmologische Operationshilfsmittel und Ophthalmologi- 52 sche Diagnostika 53 S02 Otologica 444,214 54 T Diagnostika 43,184 55 T01 Diagnostika für bildgebende Verfahren 18,885 56 57 T02 Testdiagnostika 24,299 58 V Verschiedenes 855,707 59 V01 Allergene 6,429 60 V03 Übrige Therapeutische Präparate 820,500
7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 65 of 68 BMJ Open
1 2 3 V03D Entgiftungspräparate für die Cytostatikatherapie 45,668
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from V03E Antidote 4,532 5 6 V03G Hyperkaliämie/Hyperphosphatämie-Produkte 42,259 V03X Sonstige therapeutische Präparate inkl. Entzündungshemmende En- 7 728,041 8 zyme, Eisen-Chelatbildner und Radiopharmazeutika 9 V05 Chirurgische Antiseptica 134 10 V06 Allgemeine Nährmittel 5,612 11 V07 Alle übrigen nicht therapeutischen Präparate 23,032 12 13 14 15 Source: Interpharma with data from IMS Health Schweiz, 2016. 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 66 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 4 Item Manu- 5 No script 6 Recommendation page 7 Title and abstract 1 (a) Indicate the study’s design with a commonly used 1 8 term in the title or the abstract [yes] 9 10 (b) Provide in the abstract an informative and 1-2 11 balanced summary of what was done and what was 12 found [yes] 13 14 Introduction 15 Background/rationaleFor 2peer Explain the reviewscientific background and only rationale for the 3 16 investigation being reported [yes] 17 18 Objectives 3 State specific objectives, [yes] including any 3 19 prespecified hypotheses [no] 20 21 22 Methods 23 Study design 4 Present key elements of study design early in the 3 paper [yes] 24 25 Setting 5 Describe the setting, locations, and relevant dates, 3 26 including periods of recruitment, exposure, follow-up, 27 and data collection [yes] 28 29 Participants 6 (a) Give the eligibility criteria, and the sources and 3 30 methods of selection of participants [yes] 31 32 Variables 7 Clearly define all outcomes, exposures, predictors, 3
33 potential confounders, and effect modifiers. [yes] http://bmjopen.bmj.com/ 34 Give diagnostic criteria, if applicable [no] 35 36 Data sources/ 8* For each variable of interest, give sources of data 4 37 measurement and details of methods of assessment 38 (measurement). Describe comparability of 39 assessment methods if there is more than one group 40 [yes]
41 on September 28, 2021 by guest. Protected copyright. Bias 9 Describe any efforts to address potential sources of 4 42 bias [yes] 43 44 Study size 10 Explain how the study size was arrived at [yes] 4-5 45 46 Quantitative variables 11 Explain how quantitative variables were handled in 4-5 47 the analyses. If applicable, describe which groupings 48 were chosen and why [yes] 49 50 Statistical methods 12 (a) Describe all statistical methods, including those 5 51 used to control for confounding [yes] 52 53 (b) Describe any methods used to examine 5 54 subgroups and interactions [yes] 55 56 (c) Explain how missing data were addressed [yes] 5 57 58 (d) If applicable, describe analytical methods taking n.a. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 67 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 account of sampling strategy [no] 4 5 (e) Describe any sensitivity analyses [no] n.a. 6 7 Results 8 Participants 13* (a) Report numbers of individuals at each stage of 5 9 study—eg numbers potentially eligible, examined for 10 eligibility, confirmed eligible, included in the study, 11 completing follow-up, and analysed [no] 12 13 (b) Give reasons for non-participation at each stage 5 14 [yes] 15 For peer review only 16 (c) Consider use of a flow diagram [yes] 18 17 18 Descriptive data 14* (a) Give characteristics of study participants (eg 22-24 demographic, clinical, social) and information on 19 exposures and potential confounders [yes] 20 21 (b) Indicate number of participants with missing data 5 22 for each variable of interest [yes] 23 24 Outcome data 15* Report numbers of outcome events or summary 6 25 measures [yes] 26 27 Main results 16 (a) Give unadjusted estimates and, if applicable, 28 28 confounder-adjusted estimates and their precision 29 (eg, 95% confidence interval). Make clear which 30 confounders were adjusted for and why they were 31 included [yes] 32 b 33 ( ) Report category boundaries when continuous n.a. http://bmjopen.bmj.com/ 34 variables were categorized n.a. 35 c 36 ( ) If relevant, consider translating estimates of n.a. 37 relative risk into absolute risk for a meaningful time period [no] 38 39 Other analyses 17 Report other analyses done—eg analyses of 24 40 subgroups and interactions, and sensitivity analyses
41 [yes] on September 28, 2021 by guest. Protected copyright. 42 43 44 Discussion 45 Key results 18 Summarise key results with reference to study 8 46 objectives [yes] 47 Limitations 19 Discuss limitations of the study, taking into account 2 48 sources of potential bias or imprecision. Discuss both 49 direction and magnitude of any potential bias [yes] 50 51 Interpretation 20 Give a cautious overall interpretation of results 8-11 52 considering objectives, limitations, multiplicity of 53 analyses, results from similar studies, and other 54 relevant evidence [yes] 55 56 Generalisability 21 Discuss the generalisability (external validity) of the 6 57 study results [yes] 58 59 Other information 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 68 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Funding 22 Give the source of funding and the role of the funders 13 4 for the present study and, if applicable, for the original 5 study on which the present article is based [yes] 6 7 8 *Give information separately for exposed and unexposed groups. 9 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological 10 background and published examples of transparent reporting. The STROBE checklist is best used in conjunction 11 12 with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of 13 Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the 14 STROBE Initiative is available at www.strobe-statement.org 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from
Medication Incidents in Primary Care Medicine: a Prospective Study in the Swiss Sentinel Surveillance Network (Sentinella)
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2016-013658.R2
Article Type: Research
Date Submitted by the Author: 27-Dec-2016
Complete List of Authors: Gnädinger, Markus; University of Zurich, Institute for General Medicine Conen, Dieter; Swiss Patient Safety, Herzig, Lilli; University of Lausanne, Institute of General Medicine Puhan, Milo; University of Zurich, Institute of Epidemiology, Biostatistics & Prevention Staehelin, Alfred; Sentinel Surveillance Network, Swiss Federal Office of Public Health, Zoller, Marco; Zurich University Hospital, Instiute for General Medicine Ceschi, Alessandro; National Poisons Centre, Tox Info Suisse, Associated Institute of the University of Zurich, Division of Science, Head of ; University Hospital Zurich , Dept. Clinical Pharmacology & Toxicology
Primary Subject http://bmjopen.bmj.com/ General practice / Family practice Heading:
Secondary Subject Heading: Pharmacology and therapeutics, Paediatrics, Medical management, Nursing
Health & safety < HEALTH SERVICES ADMINISTRATION & MANAGEMENT, Keywords: Adverse events < THERAPEUTICS, Patient safety, Pharmaceutical preparations, Medication errors
on September 28, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Medication Incidents in Primary Care Medicine 5 6 7 8 A Prospective Study in the Swiss Sentinel Surveil 9 10 lance Network (Sentinella) 11 12 1Markus Gnädinger (corresponding author, [email protected]), 2Dieter Conen, 3,4Lilli 13 5 1,4 1 6 14 Herzig, Milo A. Puhan, Alfred Staehelin, Marco Zoller, Alessandro Ceschi 15 For peer review only 16 1Institute of Primary Care, University of Zurich, 2Patientensicherheit Schweiz, Zurich 3Policlinique 17 4 18 Médicale, University of Lausanne, Sentinel Surveillance Network, Swiss Federal Office of Public 19 Health, Bern (Sentinella), 5Epidemiology, Biostatistics, and Prevention Institute, University of Zur 20 6 21 ich, Division of Clinical Pharmacology and Toxicology, Institute of Pharmacological Sciences of 22 Southern Switzerland, Ente Ospedaliero Cantonale, Lugano, Switzerland and Department of Clini 23 24 cal Pharmacology and Toxicology, University Hospital Zurich, Zurich, Switzerland and National 25 Poisons Centre, Tox Info Suisse, Associated Institute of the University of Zurich, Zurich, Switzer 26 27 land 28 29 (Correspondence: Markus Gnädinger, Dr. med. Facharzt für Innere Medizin, Birkenweg 8, 9323 Steinach, 30 31 0041 71 446 04 64) 32
33 http://bmjopen.bmj.com/ 34 35 36 Abstract 37 38 39 Objectives: To describe the type, frequency, seasonal and regional distribution of medication inci 40 dents in primary care in Switzerland and to elucidate possible risk factors for medication incidents.
41 on September 28, 2021 by guest. Protected copyright. 42 43 Design: Prospective surveillance study. 44 45 Setting: Swiss primary health care, Swiss Sentinel Surveillance Network. 46 47 48 Participants: Patients with drug treatment who experienced any erroneous event related to the 49 medication process and interfering with normal treatment course, as judged by their physician. The 50 51 180 physicians in the study were general practitioners or pediatricians participating in the Swiss 52 Federal Sentinel reporting system in 2015. 53 54 55 Outcomes: Primary: medication incidents; secondary: potential risk factors like age, gender, 56 poly medication, morbidity, care dependency, previous hospitalization. 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Results: The mean rates of detected medication incidents were 2.07 per general practitioner per 4 year (46.5 per 100,000 contacts) and 0.15 per pediatrician per year (2.8 per 100,000 contacts), 5 6 respectively. The following factors were associated with medication incidents (OR, 95% CI): higher 7 age 1.004 per year (1.001;1.006), care by community nurse 1.458 (1.025;2.073) , and care by an 8 9 institution 1.802 (1.399;2.323) , chronic conditions 1.052 (1.029;1.075) per condition, medications 10 1.052 (1.030;1.074) per medication, as well as Thurgau Morbidity Index for stage 4: 1.292 11 12 (1.004;1.662), 5: 1.420 (1.078;1.868), and 6: 1.680 (1.178;2.396), respectively. Most cases were 13 14 linked to an incorrect dosage for a given patient, while prescription of an erroneous medication was 15 the second mostFor common peer error. review only 16 17 18 Conclusions: Medication incidents are common in adult primary care whereas they rarely occur in 19 pediatrics. Older and multimorbid patients are at a particularly high risk for medication incidents. 20 21 Reasons for medication incidents are diverse but often seem to be linked to communication prob 22 lems. 23 24 25 Trial registration: www.clinicaltrials.gov, NCT0229537 26 27 Keywords: Patient safety, pharmaceutical preparations, medication errors. 28 29 30 Strength and limitations 31 This is the first prospective and systematic collection of incident data in primary care in Switzerland 32
33 that is characterized by three linguistic regions and two drug distribution systems. http://bmjopen.bmj.com/ 34 It was conducted by experienced physicians and with high response rates. 35 36 37 There is likely – as expected – bias from selective and underreporting or non detection of medica 38 tion incidents. 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 Introduction 45 46 47 Patient safety is a major concern in healthcare systems worldwide. Although most safety research 48 has been conducted in the in patient setting [1], evidence indicates that medical errors and ad 49 50 verse events pose a serious threat for patients in the primary care setting as well, since most pa 51 tients receive ambulatory care [2 4]. Information about the frequency and outcomes of safety inci 52 53 dents in primary care is required to identify risks or ‘hot spots’, to prioritize them and to take action 54 as needed [5]. The aim of the project was to describe the type, incidence, seasonal and regional 55 56 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 distribution of medication incidents in primary care in Switzerland and to elucidate risk factors for 4 medication incidents. 5 6 7 8 9 10 Method 11 12 13 Study design 14 15 We conductedFor a prospective peer surveillance review study among primary only care patients during 2015 to identi 16 17 fy cases of medication incidents. 18 19 Study population 20 21 22 The study population was any person undergoing drug treatment in general internal or pediatric 23 practices participating the Sentinella network. The latter covers a representative sample of patients 24 25 in primary care for Switzerland [6, included to this manuscript]. Founded in 1986, it was mainly 26 designed to survey transmissible diseases. Later, it also assessed other health problems of public 27 28 interest. It generates daily to weekly current data and covers the entire geographic and linguistic 29 regions of our country. Children, the mentally handicapped or the elderly were also included, all of 30 31 whom might be at increased risk for medication errors. 32
33 http://bmjopen.bmj.com/ Medication incidents 34 35 36 We defined medication incidents as any erroneous event (as defined by the physician) related to 37 the medication process and interfering with normal treatment course (e.g. administration of an er 38 39 roneous medication). We did not include lack of treatment effect, adverse drug reactions or drug 40 drug or drug disease interactions without detectable treatment error. Nor did we consider medica
41 on September 28, 2021 by guest. Protected copyright. 42 tion incidents if patients refused to have them reported to the Sentinella system. 43 44 Data sources 45 46 47 The study physicians recorded the patient’s year of birth and gender on their weekly reporting 48 49 form. After a maximum of four weeks they had to fill in a detailed incident questionnaire (Appendix 50 A). It was comprised of their Sentinella number and the calendar week of notification. Concerning 51 52 the patients, they reported the living situation, several supposed risk factors for an incident, as well 53 as the following variables: hospitalization during the previous year, care dependency, number of 54 55 drugs used chronically, number of chronic conditions, and the Thurgau Morbidity Index (TMI, a 56 seven step Likert scale to measure morbidity in outpatients), to be compared with a denominator 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 analysis (below) [6]. We further received a detailed description of the incident and proposals to 4 avoid future incidents. 5 6 7 We got the annual number of patient to physician contacts (PPC) per practice from the Sentinella 8 administration, as well as morbidity data from a fortnight cross sectional denominator analysis of all 9 10 patients consulting a Sentinella practice during weeks 11 or 12 [6]. 11 12 We received the anonymized list of participating physicians, their specialty, as well as the commu 13 14 nity size (Swiss Federal Statistical Office) and the linguistic region from the Sentinella administra 15 tion. InformationFor on Swiss peer medication sales review in 2015 by ATC (Anatomical only Therapeutic Chemical 16 17 Classification System, an international classification for pharmaceutical products) groups was de 18 19 rived from Interpharma Switzerland (Appendix F). 20 21 The questionnaires were completed either electronically or on a paper/pencil version, the former as 22 23 online SurveyMonkey™ questionnaires, the latter as sealed envelopes sent from the Sentinella 24 administration. We had three study questionnaires: a detailed incident questionnaire, an initial one, 25 26 and a final one. The detailed incident questionnaire collected information about the circumstances 27 of the incident and about the patient’s properties (Appendix A). The initial questionnaire served to 28 29 describe the physician’s practices in terms of e.g. number of physicians, availability of electronic 30 patient history, electronic drug drug interaction control system, and drug distribution system (Ap 31 32 pendix B). The final questionnaire investigated non reporting and difficulties with coding the mor
33 bidity variables (Appendix C). http://bmjopen.bmj.com/ 34 35 36 We constructed a variable “incident relevance” from the items “disturbance” and “endangering” of 37 the patients; if any of them was graded with “medium” or higher, the variable relevance was set to 38 39 “more”, otherwise it was set to “less”. Because of question ambiguity, we set coding of the item 40 care dependency to “missing” for patients younger than 20 years of age. Evans’ Index – a prog
41 on September 28, 2021 by guest. Protected copyright. 42 nostic index – was calculated by simple addition of the number of chronic drug treatments with the 43 number of chronic conditions [7]. 44 45 46 The following free text variables were manually coded: relationship of the incident to the suspected 47 medication, preventability of the incident, reactions to the incident, and proposals to avoid further 48 49 incidents. 50 51 Statistical methods 52 53 Values are given as frequencies, mean ± SD or median [interquartile range (IQR)] (as denoted with 54 (first quartile; forth quartile), depending on non normal distribution or non interval scaled data level. 55 56 To assess the association of medication incidents with potential risk factors we used the SPSS 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 GENLINMIXED procedure (a procedure that fits generalized linear mixed models). Clustering of 4 patients was addressed by using a mixed binary logistic regression with the fixed factors of gender, 5 6 year of birth, care dependency, number of chronic drug treatments, number of chronic conditions, 7 and TMI as well as the physicians’ practice number as a random factor; if one item was missing, 8 9 the whole record was excluded from the analysis. Numerator was affiliation to the case group; de 10 nominator was all PPC during weeks 11 and 12. We used IBM SPSS 23. 11 12 13 14 Results 15 For peer review only 16 The Sentinella system 17 18 During the year 2015, 149 practices were enrolled to the Sentinella system. Of them, 144 practices 19 were known to report regularly; their properties are listed in Table 1. The Sentinella physicians are 20 21 representative for the overall Swiss physician population [5,6]. Drugs were auto distributed by 42% 22 23 of the study practices. Approx. half of the physicians had electronic and the other paper based 24 medical records. Systematic drug drug interaction control systems were installed only by a minority 25 26 (36.8%) of the physicians. During the year 2015 (which included unusually for calendar adaptation, 27 53 instead of 52 reporting weeks), the general practitioners (GPs) had 4,456±2,137 PPC; for the 28 29 pediatricians (PEDs), these were 5,297±2,715 (mean±SD). In their initial study questionnaire, 30 11.1% of the reporting physicians admitted to be working part time (i.e. less than 30 h per week). 31 32
33 Study flow http://bmjopen.bmj.com/ 34 35 During the year 2015, we received 216 incident notifications (Figure 1). In 11 cases, we did not 36 receive the detailed notification form. Eight cases had to be removed from the database because 37 38 they fulfilled the exclusion criterion (i.e. “adverse drug reactions without detectable error”), and one 39 case had to be removed because of double data entry. This led to 197 cases which could be ana 40 lyzed. The distribution of the monthly incident notifications throughout 2015 is depicted in Figure e1 41 on September 28, 2021 by guest. Protected copyright. 42 (Appendix D), and the distribution of the numbers of cases reported by each practice in Figure 2. 43 44 45 Description of patients 46 47 Table 2 lists age, gender and geographic distribution as well as the physician to patient relation 48 49 ship and the observer of the incident. Only three cases evolved in PED practices. No statistically 50 significant differences were found between the two relevance classes of the incidents. 51 52 53 Number of incidents, non-reporting 54 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 In the GPs 194 incidents, 148 physicians, 4,456 yearly PPC led to 1.31 incidents per physician per 4 year or 29.4 per 100,000 PPC; in PEDs 3 incidents, 32 physicians, 5,297 yearly PPC led to 0.1 5 6 incident per physician per year or 1.8 per 100,000 PPC. 7
8 9 To evaluate the non-reporting of incidents, we asked the physicians in the final study question 10 naire. Out of 180 actively reporting physicians we received 145 questionnaires (80.6% response 11 12 rate). To our question: “Did you not report medication incidents that you had noticed during the last 13 14 year?”, they answered: “Never or almost” 110 (75.9%), “yes, but seldom” 22 (15.2%), “yes, fre 15 quently” 9 (6.2%),For “always, peer or almost” 4 (2.8%).review Reasons for not only reporting incidents were “lack of 16 17 time” or “forgetfulness”. If we speculate that these answers represent reporting rates of 76 100, 51 18 75, 26 50, or 0 25% respectively, we could divide the observed rates by the middle of the reporting 19 20 classes: 0.875, 0.625, 0.375 and 0.125. By doing so, we calculated a rate of 50% of underreport 21 ing; if we furthermore consider the 5% of the incidents where the questionnaires were not sent, this 22 23 rate increases to 58%. We therefore have to multiply the observed rates with a factor of 1.58 re 24 sulting in the following rates of detected incidents: GP 2.07 per physician per year, 46.5 per 25 26 100,000 PPC and PED 0.15 per physician per year, 2.8 per 100,000 PPC. 27 28 29 Types and causes of incidents, organ systems involved, preventability 30 The types of error are listed in Table e1 (Appendix D) and Figure 3; in 26 cases, more than one 31 32 was mentioned. Most cases were linked to an incorrect dosage for a given patient, while prescrip
33 http://bmjopen.bmj.com/ tion of an erroneous medication was the second most common error. 34 35 36 Most errors concerned orally ingested medication. Errors in application of parenteral drugs were 37 reported less frequently; in our study, they comprised insulin (one case) and vaccinations (three 38 39 cases). There were cases of an incorrect (influenza vs. anti tetanus) or incomplete (Boostrix® vs. 40 Boostrix Polio®) vaccination and of undue vaccination (a third anti HPV vaccine). One case was
41 on September 28, 2021 by guest. Protected copyright. 42 due to a communication problem within the practice staff, another to a vaccination in absence of 43 44 the vaccination card. 45 46 Three cases concerned pediatric patients, all were linked to inadequate dosing, two times of an 47 48 antibiotic prescription (Cotrimoxazole, Co Amoxicillin) and one time of an antiemetic drug 49 (Thiethylperazine). 50 51 52 In 89 of the 195 cases, organ system damage was reported, in 13 cases more than one organ sys 53 tem was involved, most frequently the central nervous system (Figure 4) (Table e2, Appendix D). 54 55 Possible triggers of the incident were reported in 194 of the 197 cases; in 45 cases, there was 56 more than one single reason, most frequently lacking alertness of the reporting physicians or their 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 staff (Table e3). When asked, who might be the “responsible” for the incident, 191 replied. The 4 most common person or institution possibly responsible for the medication incident was: the report 5 6 ing physician 41 (21.5%), followed by the practice nurse 26 (13.6%), the institution where the pa 7 tient lives 33 (17.3%), the pharmacy 7 (3.7%), the hospital 12 (6.3%), the community nurse 4 8 9 (2.1%), the patients or their proxies 15 (7.9%), and the manufacturer 2 (1.0%); in 9 cases (4.6%) 10 this was unclear, in 37 cases (19.4%), there was more than source one to blame. Preventability of 11 12 the incidents was classified (by our study board): unlikely in 6 cases (3.0%), possible in 58 13 14 (29.4%), probable in 114 (57.9%), and definite in 19 (9.6%). 15 For peer review only 16 17 Endangering and disturbances 18 In 192 of the 197 cases the item “patients’ endangering” as estimated by the physicians was an 19 20 swered. In 39 cases (20.3%) there was no endangering, in 83 (43.2%) light, in 51 (26.6%) moder 21 ate, in 19 (9.9%) severe. The answers to item “disturbances caused by the incident” are listed in 22 23 Table 3. Out of the 197 incidents, 74 (37.5%) were classified as “more” relevant (see methods sec 24 tion, Table 2). 25 26 27 Interface problems, orientation, predictability, repeat incidents 28 29 The presence of interface problems was reported in 64 of 197 cases (32.4%); these were: with a 30 hospital in 28 cases (43.8%), with an institution in 14 (21.9%), with a community nurse in 6 (9.4%), 31 32 with a pharmacist in 9 (14.0%), with a specialist in 3 (4.7%), and with others in 4 (6.3%). In 184
33 http://bmjopen.bmj.com/ cases, we received information about informing patients about the incident; in 98 cases, the patient 34 35 was oriented by the reporting physician or his staff (50.3%), non orientation was stated: because 36 the patient was not able to understand (children, demented) in 26 cases (14.1%), because the 37 38 problem had already been solved and the notification would have unduly disturbed the confidence 39 40 in 18 cases (9.8%), because the patient or the proxies had observed themselves in 23 cases
41 (12.5%), because the patient had already been oriented by others in 7 cases (3.8%) or because of on September 28, 2021 by guest. Protected copyright. 42 43 another reason in 12 cases (6.5%). As for predictability of the incident, we received 183 valid an 44 swers; of them 82 (44.8%) stated “yes, in the given constellation, the incident was to be expected”. 45 46 When asked whether they had already reported a similar incident in the study, 29 out of 196 47 (14.8%) answered in the affirmative. 48 49 50 Risk factors to undergo a medication incident 51 52 To detect patient risk factors to undergo an incident, we compared the incident data with those of a 53 fortnight denominator analysis [6]; the following univariate factors accumulated preferentially in 54 55 incident patients: higher age, care dependency, higher numbers of chronic conditions or medica 56 tions (or higher Evans’ Index), as well as higher TMI (table 4). In the logistic regression only inpa 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 tient care by an institution remained a significant factor. Other suspected risk factors are listed in 4 Table e4 (Appendix D); these items were not included in the denominator study and therefore lack 5 6 a comparator. Within the patient group with a more as compared to a less relevant incident, only 7 psychiatric illness reached a significantly increased proportion. We did not detect major differences 8 9 between the two drug distribution systems. 10 11 Relationship between incidents and drug class, ATC-group 12 13 We assessed semi quantitatively whether the class of the suspected drug was causally linked to 14 the emergence of the incident; the frequencies were: none 49 (25.1%), unlikely 31 (15.9%), possi 15 For peer review only 16 ble 77 (39.5%), probable 34 (17.4%), definite 4 (2.1%), did not apply 2 (n.a). ATC codes of the 17 18 suspected medications are listed in Table 5. Most naming concerned the groups C “cardiovascular” 19 (23.2%) and N “nervous system” (22.1%). Among the medication classes judged to be related to 20 21 the incident (n=37) the most frequently named group were oral anticoagulants: rivaroxaban 9, 22 phenprocoumon 7, and acenocoumarol 2 cases; this explains the 7 fold increased relative risk of 23 24 ATC group B as compared to sales. As an example for errors without relation to the drug class, we 25 must mention the 17 cases of institutionalized patients ingesting medications scheduled for other 26 27 residents, in most cases this was person eating at the same table, but in other cases the person in 28 care mixed up patient names. 29 30 31 Reactions to the incident and suggestions to prevent further ones 32
33 In 141 of the 197 cases, the physicians reported to have changed something after the incident, in http://bmjopen.bmj.com/ 34 13 cases this was more than one single action, mostly often named were communication with other 35 36 caregivers or better instruction of patients (Table e5, Appendix D). The respondents were asked for 37 proposals how to prevent future incidents of the reported type; 125 of them made a total of 243 38 39 suggestions (Table e6, Appendix D). Of these, 37 (15.2%) were related to accurate medication 40 lists, 10 (4.1%) to better patient instructions, 38 (15.6%) to organization of regular follow up con
41 on September 28, 2021 by guest. Protected copyright. 42 trols, and 55 (22.6%) involved organizational changes within the practice and its staff. 43 44 45 46 Discussion 47 48 In a representative group of primary care physicians [6,8], we found approximately one case of a 49 medication incident per GP and year. 50 51 52 Incident rates 53 54 Incident rates in this study were similar to those found in other studies. We calculated the rates of 55 detected medication incidents as follows: GP 2.07 per physician and year, 46.5 per 100,000 PPC 56 57 and PED 0.15 per physician and year, 2.8 per 100,000 PPC. Medication incidents may make up a 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 proportion of approximately one third of all incidents [9]; the rates for all safety incidents may 4 amount 6.20 per physician and year or 139.4 per 100,000 PPC in GPs and 0.45 per physician and 5 6 year or 8.4 per 100,000 PPC in PEDs. 7 In Australian primary care patients, an incident rate of 4.98 per GP and year was reported; this is 8 9 close to our estimated rate of 6.20; however, there was no sub typing of the incidents [10]. In a 10 three year study, O’Beirne and colleagues reported a rate of 1.8 safety incidents per year and phy 11 12 sician [9]. In a literature review of western countries’ publications, Sandars and Esmail described a 13 14 rate of 5 to 80 errors per 100,000 consultations [2], a rate somewhat lower than the 139.4 cases 15 per 100,000 Forconsultations peer estimated in ourreview study. Kuo and colleagues only reported a proportion of 16 17 15% of all incidents to be related to medication [11]; since we estimated a proportion of 33% this 18 would give rise to even higher rates of all safety incidents when calculated from our study data 19 20 base. An 11% group of the physicians worked part time, the rate per full time physician per year 21 may therefore be somewhat higher than calculated. The 13 fold higher incident rate in GPs as 22 23 compared to PEDs is not surprising given the lower medication rates in children. A recent British 24 study confirmed this much lower rate of incidents in children; out of 46,902 family practice safety 25 26 reports, 1,788 concerned children (26 times less than adults) [12]. 27 28 Definition of incidents and reliability of reporting 29 30 On the other hand, incident rates may be influenced by their definition. Gandhi and colleagues 31 coined the term “avoidable adverse drug reaction” [13]. In our study, we explicitly excluded adverse 32
33 drug reactions without detectable error. Runciman and colleagues defined a patient safety incident http://bmjopen.bmj.com/ 34 as follows: “An event or a circumstance that could have resulted or did result in unnecessary harm 35 36 to a patient” [14]. An intuitive definition of a medical error was given by Makeham et al: “That was a 37 threat to patient wellbeing and should not happen. I don’t want it to happen again.” [15]. As shown 38 39 in Figure e1 (Appendix D), reporting frequency was higher at the beginning of the study as com 40 pared to the later course of it. This could reflect some loss of interest or forgetfulness by the report
41 on September 28, 2021 by guest. Protected copyright. 42 ing physicians. As calculated from our final questionnaire after the study, the reporting physicians 43 44 failed to report about one in three cases of the detected medication errors. Non detection of inci 45 dents may even be more frequent than non reporting of observed incidents. A missed possible 46 47 drug drug interaction may be detected by chart review, a documentation error would probably have 48 been found only in 1:1 supervision, which is very time consuming, costly and may additionally in 49 50 fluence performance of the observed physician. It is therefore virtually impossible to decipher the 51 real rate of non detected incidents. The problems in detection of incidents were the reason to pos 52 53 tulate a “mix of methods” as needed to identify adverse events in general practice [16]. 54 55 56 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Other approaches to investigate safety incidents 4 In contrast to our prospective investigation, there are other methods to approach safety incidents. 5 6 In a retrospective, semi quantitative analysis, Gehring and colleagues investigated safety incidents 7 in Swiss primary care [17]; among 23 predefined classes of safety incidents, the respondents ad 8 9 mitted 15 to have occurred at least yearly – four of them being linked to drug treatment. Another 10 approach to incidents is to ask patients as performed by Mira and colleagues [18]; they interviewed 11 12 patients (> 65 years with five or more chronic drug treatments) and found that 75% of the patients 13 14 reported to have been affected by at least one medication error during the previous twelve months. 15 We did not collectFor data onpeer appropriateness review of treatment; admittedly only medication inappropriate for 16 17 some groups of patients (elderly or patients with impaired renal function) may provoke incidents 18 [19]. A recent Scottish study demonstrated an impressive reduction in high risk prescriptions (non 19 20 steroidal anti inflammatory drugs, antiplatelet, and anticoagulation) as well as hospital admissions 21 for gastrointestinal ulcers or heart failure by a combination of educational measures, informatics 22 23 and financial incentives [20]. 24 25 26 27 Type, consequences, causes and preventability of error 28 29 Most cases in this study involved application of an erroneous dosing or of a wrong medication alt 30 31 hough non application of necessary medication was also frequent. The distribution of the incidents 32 was similar to that reported in other studies [11,17,18,22]. When a wrong medication was dis
33 http://bmjopen.bmj.com/ 34 pensed, the error was often caused by confounding of prepared medication in home residents. The 35 classic case of non application of a necessary medication was the missed re uptake of anticoagu 36 37 lation after an operation. Non application of necessary drugs seems to be a relevant source of un 38 necessary harm to patients [21]. The causing of the incidents was similar as reported by others 39 40 [11,17,18,22]. Most studies found – as in our patients – lacking alertness of and communication
41 on September 28, 2021 by guest. Protected copyright. problems within practice staff, but there was a large variety of other topics. All three pediatric cases 42 43 were linked to prescription of an inadequate dosing for age and weight. More than half of the pa 44 45 tients did not have any disturbances after the incident; otherwise in most cases the nervous system 46 was affected. No fatalities were reported, but seven patients (3.5%) needed stationary care. In 47 48 2004 Pirohamed and colleagues published a study on adverse drug reactions as a cause for ad 49 mission to hospital; they found that 6.5% of all hospitalizations and 4.0% of all hospital days were 50 51 caused by adverse drug reactions, 72% of them being preventable and 2% leading to death [23], 52 however, this study included all cases, and therefore a majority of cases of adverse drug reactions 53 54 without a detectable error. An older Swiss study was published by Livio and colleagues; out of 55 3,195 hospitalizations, she identified 229 cases (7.2%) as probably caused by adverse drug reac 56 57 tions [24]. In that study, 32% of the events were classified as being preventable (which does not 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 strictly mean that an error had caused the incident), and in 6% of the cases, fatalities were report 4 ed. The results on outpatients contrast to inpatients where one in ten drug administrations was 5 6 described to be erroneous [25]. Medical errors seem to be the “third leading cause of death in the 7 US” [26]. 8 9 10 11 Risk factors 12 13 14 We were able to identify some risk factors such as higher age, care dependency, higher numbers 15 of chronic conditionsFor or medicationspeer (and review higher Evans’ Index), only as well as higher TMI. However, 16 17 adjusted in the logistic regression analysis, only care dependency remained a significant risk fac 18 19 tor, but this may be due to the small number of observations which possibly precluded less im 20 portant risk factors to be detected. The only significant risk factor for undergoing an incident of 21 22 higher relevance was psychiatric disease. Several factors which reflect quite well the results of our 23 study have been described in the literature to correlate with proneness to undergo a medication 24 25 incident: mainly the number of drugs ingested [13,27], but also higher or young age [28] and mor 26 bidity [29,30] have been described. Most incidents concerned ATC groups N “nervous system” or 27 28 C “cardiovascular”, which were also of the mostly sold drugs; an exception was group B with anti 29 coagulants and a sevenfold increased relative risk as compared to Swiss sales in 2015. It seems 30 31 wise to be alert to avoid errors when prescribing medication of these groups. The prevailing posi 32 tion of anticoagulants (18.5% of cases) was described also by Field and colleagues [30]. Otherwise
33 http://bmjopen.bmj.com/ 34 the repartition of our suspected drugs was similar to other primary care studies [11,30], a literature 35 review [31] or a theoretical paper [22]. 36 37 38 Limitations 39 40 There is likely – as expected – bias from selective and underreporting or non detection of medica
41 on September 28, 2021 by guest. Protected copyright. 42 tion incidents. Therefore, the true rate of incidents could only be estimated, and the proportions of 43 incident characteristics would probably substantially have been altered by a more complete record 44 45 ing of incidents. 46 47 48 49 Conclusion 50 51 Medication incidents are common in general medicine whereas they rarely occur in pediatrics, in 52 which polypharmacy is less prevalent. Reasons for medication incidents are diverse but often 53 54 seem to be linked to communication problems. Older and multimorbid patients are at a particularly 55 high risk for medication incidents. 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 Abbreviations 6 7 8 GP general practitioner 9 10 PED pediatrician 11 12 13 PPC patient to physician contacts 14 15 TMI ThurgauFor Morbidity peer Index review only 16 17 18 Author affiliations 19 20 21 MG Institute of Primary Care, University of Zurich, Switzerland. 22 AC Division of Clinical Pharmacology and Toxicology, Institute of Pharmacological Sciences of 23 24 Southern Switzerland, Ente Ospedaliero Cantonale, Lugano, and Department of Clinical Pharma 25 cology and Toxicology, University Hospital Zurich, Zurich, and National Poisons Centre, Tox Info 26 27 Suisse, Associated Institute of the University of Zurich, Zurich, Switzerland 28 DC Patientensicherheit Schweiz, Zurich, Switzerland. 29 30 LH Policlinique Médicale, University of Lausanne, Switzerland and Sentinel Surveillance Network, 31 Swiss Federal Office of Public Health, Bern, Switzerland. 32
33 MP Epidemiology, Biostatistics, and Prevention Institute, University of Zurich, Switzerland. http://bmjopen.bmj.com/ 34 AS Sentinel Surveillance Network, Swiss Federal Office of Public Health, Bern, Switzerland. 35 36 MZ Institute of General Medicine, University of Zurich, Switzerland. 37 38 39 Acknowledgements 40
41 on September 28, 2021 by guest. Protected copyright. 42 We are grateful to Lee Wennerberg for the English language corrections, Dr Sven Staender, 43 44 Männedorf for the helpful comments, and Simon Gnädinger, Zurich for the manual coding of free 45 text items. We thank the Sentinella program commission for their support, the reporting physicians 46 47 of Sentinella for their unflagging enthusiastic collection of data, and the Federal Office of Public 48 Health for providing data and translating the questionnaires into French. 49 50 51 52 53 Contributorship statement 54 55 MG lead the study, did the pilot study (questionnaire development, data entering and processing) 56 57 wrote all documents, did all the contacts with the Sentinella administration, ethics committee, and 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 others, programmed the electronic questionnaires, entered hand written questionnaires into the 4 database, did the data processing and wrote the publication after data collection. 5 6 AC is an expert on clinical pharmacology and drug safety. 7 DC is an expert on patient safety. 8 9 LH is French speaking and helped to interpret the French questionnaires. She is an expert on mul 10 timorbidity. She is a member of the Sentinella program commission. 11 12 MP is head of Epidemiology, Biostatistics & Prevention Institute. He is responsible for the sound 13 14 methodology. 15 AS had the ideaFor for the study.peer He is vice reviewpresident of the Sentinella only program commission. 16 17 MZ is expert on electronic data exchange in primary care. 18 All of them have seen all the study documents and have contributed intellectually to their elabora 19 20 tion. All have contributed to revise the draft of this publication and approve the submitted version of 21 this publication. 22 23 24 Funding 25 26 The study was funded by Bangerter Rhyner Foundation, Basel. 27 28 29 Competing interests 30 31 32 The authors declare that they have no financial interest conflicts with this study.
33 http://bmjopen.bmj.com/ 34 35 Patient consent 36 37 Not necessary for anonymous data. 38 39 40 Ethics approval
41 on September 28, 2021 by guest. Protected copyright. 42 The ethical committee of Canton Zurich decided that our study did not need formal approval be 43 44 cause the data are completely anonymous (KEK ZH 2014 0400). The study was recorded in 45 www.ClinicalTrials.gov: NCT02295371, as well as in our national study registry (www.kofam.ch; 46 47 SNCTP000001207). 48 49 Data sharing statement 50 51 52 No additional data are available. 53 54 55 Strobe statement 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Whenever possible, the guidelines of the Strobe statement were followed (Appendix E). 4 5 6 7 References 8 9 10 1. Manias E: Detection of medication related problems in hospital practice: a review. Br J Clin 11 Pharmacol 2012; 76 (1): 7 20. 12 13 2. Sandars J, Esmail A: The frequency and nature of medical error in primary care: under 14 standing the diversity across studies. Fam Pract 20 (3): 231 6, 2003. 15 For peer review only 16 3. Miller GC, Britth HC, Valenti L. Adverse drug events in general practice patients in Aus 17 tralia. Med J Aust. 2006;184(7):321 324. 18 19 4. Thomsen LA, Winterstein AG, Søndergaard B, Haugbølle LS, Melander A. Systematic 20 review of the incidence and characteristics of preventable adverse drug events in ambulato 21 22 ry care. Ann Pharmacother. 2007; 41(9):1411 1426. 23 24 5. Gnädinger M, Ceschi A, Conen D, et al: Medication incidents in primary care medicine: 25 26 protocol of a study by the Swiss Federal Sentinel Reporting System. BMJ Open 04/2015; 27 28 5(4): e007773. DOI:10.1136/bmjopen 2015 007773. 29 30 6. Gnädinger M, Herzig L, Ceschi A et al: The Burden Related to Chronic Conditions and 31 Multimorbidity in the Swiss Primary Care Population: A Prospective Study in the Swiss Sen 32 tinel Surveillance Network (Sentinella), 2016, to be submitted. That manuscript is includ- 33 http://bmjopen.bmj.com/ 34 ed in the present submission. 35 36 7. Evans DC, Cook CH, Christy JM et al: Comorbidtiy polypharmacy scoring facilitates out 37 come prediction in older trauma patients. J Am Geriatr Soc 60: 1465 70, 2012. 38 39 40 8. Cohidon C, Cornuz C, Senn N: Primary care in Switzerland: evolution of physicians‘ pro
41 file and activities in twenty years (1993 2012). BMC Fam Pract 16: 107, 2015. on September 28, 2021 by guest. Protected copyright. 42 43 9. Makeham M, Dovey S, Runciman W, et al: Methods and Measures used in Primary Care 44 Patient Safety Research. Results of a literature review. WHO report on patient safety Qual 45 46 Safety Health Care 2008. 47 48 10. Makeham MAB, Kidd MR, Saltman DC et al: The threats to Australian patient safety 49 (TAPS) study: incidents of reported errors in general practice. Med J Aus 185: 95 98, 2006. 50 51 11. Kuo GM, Phillips RL Graham D et al: Medication errors reported by US family physicians 52 and their office staff. Qual Saf Health Care 17: 286 90, 2008. 53 54 12. Rees P, Edwards A, Panesar S et al: Safety incidents in the primary care office setting. 55 Pediatrics 135 (6): 1027 35, 2015. 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 13. Gandhi TK, Weingart SN, Borus BA, et al: Adverse drug events in ambulatory care. N 4 Engl J Med 348: 1556 64, 2003. 5 6 14. Runciman W, Hibbert P, Thomoson R et al: Towards an international classification ofr 7 8 patient safety: key concepts and terms. Int J Qual Health Care 21 (1): 18 26, 2009. 9 15. Makeham MAB, Dovey SM, County M et al: An international taxonomy for errors in gen 10 11 eral practice: a pilot study. Med J Aus 177:68 72; 2002. 12 16. Wetzels R, Wolters R, van Weel C et al: Mix of methods is needed to identify adverse 13 14 events in general practice: a prospective observational study. BMC Fam Pract 2008; 9:35. 15 For peer review only 16 17 17. Gehring K, Schwappach DLB, Battaglia M et al: Frequency of and harm associated with 18 primary care safety incidents. Am J Managed Care 18 (9): e323 7, 2012. 19 20 18. Mira JJ, Orozco-Beltran D, Perez-Jover V et al: Physician patient communication failure 21 22 facilitates medication errors in older polymedicated patients with multiple comorbidities. 23 Fam Pract 30: 56 63, 2013. 24 25 26 19. Cooper JA, Moriarty F, Ryan C et al: Potentially inappropriate prescribing in two popula 27 tions with differing socio economic profiles: a cross sectional database study using the 28 29 PROMPT criteria. Eur J Clin Pharmacol 72: 583 91, 2016. 30 20. Dreischulte T, Donnan P, Grant A et al: Safer prescribing – a trial of education, informat 31 32 ics, and financial incentives. New Engl J Med 374: 1053 64, 2016.
33 http://bmjopen.bmj.com/ 34 21. O’Grady I, Gerrett D: Minimising harm from missed drug doses. Nursing Times 111 (44): 35 12 15, 2015. 36 37 22. Huges RG, Ortiz E: Medication errors. Why they happen, and how they can be prevented. 38 Am J Nursing 205 (3): 14 24, 2005. 39 40 23. Pirmohamed M, James S, Meakin S, et al: Adverse drug reactions as cause of admission
41 to hospital: prospective analysis of 18 820 patients. BMJ 2004; 329:15–19. on September 28, 2021 by guest. Protected copyright. 42 43 24. Livio F, Buclin T, Yersin B et al: Hospitalisations pour effet indésirable médicamenteux. 44 Recensement prospectif dans un service d’urgences médicales [French]. Raisons de Santé 45 46 23, 1998. 47 48 25. Berdot S, Gillaizeau F, Caruba T et al: Drug administration errors in hospital inpatients: a 49 50 systematic review. PLoS ONE 8 (6): e68856, 2013; doi: 10.1371/journal.pone.0068856. 51 52 26. Makary M, Daniel M : Medical error – the third leading cause of death in the US. BMJ 53 54 2016 ; 353: i2139. 55 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 27. Nobili A, Pasina L, Tettamanti M et al: Potentially severe drug interactions in elderly out 4 patients: results of an observational study of an administrative prescription database. J Clin 5 6 Pharm Ther 34: 377 86, 2009. 7 8 28. Avery AJ, Ghaleb M, Barber N, et al: The prevalence and nature of prescribing and moni 9 10 toring errors in English general practice: a retrospective case note review. Br. J Gen Pract 11 e543, 2013. 12 13 29. Field TS, Gurwitz JH, Avorn J et al: Risk factors for adverse drug events among nursing 14 home residents. Arch Intern Med 161: 1629 34, 2001. 15 For peer review only 16 30. Field TS, Mazor KM, Briesacher B, et al: Adverse drug events resulting from patient er 17 18 rors in older adults. J Am Geriatr Soc 55: 271 6, 2007. 19 20 31. Saedder EA, Brock B, Nielsen LP et al: Identifying high risk medication: a systematic lit 21 erature review. Eur J Clin Pharmacol 70: 637 45, 2014. 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 List of figures and tables 4 5
6 7 8 Figure 1: Study flow chart 9 10 Figure 2: Distribution of the number of cases reported by practice 11 12 13 Figure 3: Type of error 14 15 Figure 4: OrganFor system involvedpeer review only 16 17 18 Table 1: Characteristics of the reporting physicians in 2015 19 20 Table 2: General description of the cases 21 22 23 Table 3: Disturbances after the incident 24 25 Table 4: Possible risk factors for incident as compared to a denominator analysis during calendar 26 27 weeks 11 and 12 [Gnädinger M 2016] 28 29 Table 5: ACT Groups of suspected medications 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 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 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Physicians’ gender 5 male 128 (71.1%) 6 female 52 (28.9%) 7 Physicians’ age class 8 < 40 12 (6.7%) 40 49 44 (24.4%) 9 50 59 66 (36.7%) 10 60 and over 58 (32.2%) 11 Specialty 12 GP 148 (82.2%) 13 PED 32 (17.8%) 14 Number of physicians in practice (n=144) 15 1 For peer review only69 (47.8%) 16 2 39 (27.1%) 17 3 17 (11.8%) 18 4 to 5 8 (5.6%) 19 6 to 9 7 (4.9%) 20 10 and over 4 (2.8%) Number of physicians per practice reporting to Sentinella (n=144) 21 1 119 (82.6%) 22 2 19 (13.2%) 23 3 5 (3.5%) 24 8 1 (0.7%) 25 Linguistic region 26 German 122 (67.8%) 27 French 44 (24.4%) 28 Italian 14 (7.8%) 29 Urbanity of the practice 30 urban 93 (51.7%) 31 agglomeration 60 (33.3%) 32 rural 27 (15.0%) Workload per week 33 http://bmjopen.bmj.com/ < 15 h 9 (5.0%) 34 15 30 h 36 (20.0%) 35 > 30 h 135 (75.0%) 36 Drug distribution system 37 dispensing by physician 73 (42.2%) 38 mixed system 19 (10.6%) 39 dispensing by pharmacy 85 (47.2%) 40 Electronic documentation yes 89 (49.4%)
41 on September 28, 2021 by guest. Protected copyright. 42 no 91 (50.6%) 43 Electronic interaction control 44 yes 65 (36.1%) 45 no 115 (63.9%) 46 Electronic prescription yes, with thesaurus 62 (34.4%) 47 yes, but without thesaurus 24 (13.3%) 48 none 94 (52.2%) 49 Certification of the practice 50 yes 46 (25.6%) 51 none 134 (74.4%) 52 Staff meetings 53 yes, at least monthly 69 (38.3%) 54 yes, but less frequently 70 (38.9%) 55 none 41 (22.8%) 56 Quality circle participation 57 yes, at least monthly 134 (74.4%) 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 yes, but less frequently 23 (12.8%) 4 none 23 (12.8%) 5 6 Table 1. Characteristics of the reporting physicians in 2015 (144 practices, 180 physicians) 7 . 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Relevance All 5 6 7 less more 8 9 10 Number of cases 124 73 197 11 12 13 Patients age 69.2 ± 20.6 69.4 ± 21.2 69.3 ± 20.8 14 15 For peer review only 16 Patients gender, 17 % males 40.3 32.9 37.6 18 19 20 Physicians specialty 21 % pediatricians 1.6 1.4 1.5 22 23 24 Linguistic region, % 25 26 German 75.0 68.5 72.6 27 French 18.5 28.8 22.3 28 Italian 6.5 2.7 5.1 29 30 31 Physician to patient relationship, % 32 own family physician 86.3 78.1 83.2 33 http://bmjopen.bmj.com/ 34 urgency / holiday replacing 0.8 4.1 2.0 35 institution physician 11.3 17.8 13.7 36 37 other 1.6 0.0 1.0 38 39 40 Observer of the incident; % physician / practice staff 50.0 50.7 50.3 41 on September 28, 2021 by guest. Protected copyright. 42 patient / proxies 21.8 23.3 22.3 43 44 community nurse 1.6 4.1 2.5 45 institution (where patient lives) 15.3 16.4 15.7 46 hospital 0.8 1.4 1.0 47 48 other physicians 2.4 0.0 1.5 49 pharmacist 7.3 4.1 6.1 50 51 other 0.8 0.0 0.5 52 53 54 Table 2. General description of the cases. Differences between “less” and “more” categories are 55 not statistically significant. We calculated the variable “incident relevance” from the variables “dis 56 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 turbance” and “endangering” of the patients; if any of them was graded with “medium” or higher, 4 the variable relevance was set to “more”, otherwise to “less”. 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Item N percent 4 5 6 Severity of disturbance 7 8 no symptoms but pathological laboratory tests 15 8.0 9 light 44 23.4 10 11 moderate 22 11.7 12 severe 10 5.3 13 14 fatality 0 0.0 15 Subtotal (thisFor is the base peerof the next rows) review only 91 48.4 16 17 no symptoms, normal (or no) laboratory tests 97 51.6 18 19 Total 188 100.0 20 Missing data 9 n.a. 21 22 All patients 197 n.a. 23 24 25 Time until recovery 26 hours 26 28.5 27 28 days 41 45.1 29 weeks 15 16.5 30 31 not yet known or missing information 9 9.9 32 All patients with disturbances 91 100.0
33 http://bmjopen.bmj.com/ 34 35 Recovering 36 37 without sequels 78 85.8 38 with light to moderate sequels 2 2.2 39 40 with severe sequels or fatality* 5 5.4
41 not yet known or missing information 6 6.6 on September 28, 2021 by guest. Protected copyright. 42 43 All patients with disturbances 91 100.0 44 45 46 Treatment / surveillance 47 not needed 48 52.7 48 49 ambulatory care 33 36.3 50 hospital care** 7 7.7 51 52 missing information 3 3.3 53 All patients with disturbances 91 100.0 54 55 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Table 3. Disturbances after the incident. * In one case, there was a reduced kidney 4 5 function, the other cases remained unclear, and no fatalities were reported. ** Two cases 6 had to be surveilled in the emergency room, the hospital stays were: intoxications with 7 8 thiethylperazine, with fenoterol plus ipratropium bromide, with zolpidem, further a derailed 9 10 diabetes type 2 (after missed treatment with metformin), and a gastro intestinal hemor 11 rhage in a patient where antithrombotic treatment with rivaroxaban was not communicated 12 13 to the physician. 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 Item Patient group OR (95% CI) OR (95% CI) 5 (crude) (adjusted) 6 7 Denominator Incidents 8 9 10 Number of observations 26,852 197 / / 11 12 Age (mean±SD), years 46.7 ± 27.5 69.3 ± 20.8 1.004 (1.001;1.006)**a 1.001 (0.996;1.005) 13 14 15 Gender, For peer review only 16 ● male 47.0 37.6 1 1 17 ● female 53.0 62.4 1.048 (0.916;1.197) 1.055 (0.878;1.196) 18 19 20 Care dependency, 21 d 22 number (%)* 23 ● none 16,335 (85.5%) 96 (51.6%) 1 1 24 ● yes, by proxies 954 (5.0%) 12 (6.5%) 1.121 (0.789;1.594) 0.979 (0.674;1.423) 25 a ● yes, by community nurse 723 (3.8%) 22 (11.8%) 1.458 (1.025;2.073) 1.201 (0.821;1.758) 26 1,099 (5.8%) 56 (30.1%) 1.802 (1.399;2.323)c 1.528 (1.141;2.046) a 27 ● yes, by institution 28 29 30 c Number of conditions 2 (0;4) 5 (3;7) 1.052 (1.029;1.075)** 1.030 (0.994;1.067)** 31 32 (median, IQR)
33 http://bmjopen.bmj.com/ 34 Number of chronic active 1 (0;4) 6 (3;9) 1.052 (1.030;1.074)**c 1.030 (0.995;1.067)** 35 36 treatments (median, IQR) 37 38 Evans’ Index 3 (0;8) 11 (6;17) 1.009 (1.005;1.013)** c n.a.*** 39 40 (median, IQR)
41 on September 28, 2021 by guest. Protected copyright. 42 Thurgau Morbidity Index d 43 d 44 value (%) 45 ● 0 8,463 (31.5%) 24 (12.2%) 1 1 46 ● 1 3,611 (13.4%) 8 (4.1%) 0.989 (0.787;1.242) 0.908 (0.694;1.190) 47 ● 2 4,102 (15.3%) 23 (11.7%) 1.049 (0.847;1.300) 0.898 (0.685;1.169) 48 3,877 (14.4%) 39 (19.8%) 1.131 (0.914;1.399) 0.830 (0.611;1.127) 49 ● 3 a 50 ● 4 2,119 (7.9%) 39 (19.8%) 1.292 (1.004;1.662) 0.901 (0.643;1.265) b 51 ● 5 1,539 (5.7%) 38 (19.3%) 1.420 (1.078;1.868) 0.823 (0.547;1.239) 52 709 (2.6%) 26 (13.2%) 1.680 (1.178;2.396)c 0.866 (0.523;1.436) 53 ● 6 18 0 54 missing values 55 56 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Table 4. Possible risk factors for incident as compared to a denominator analysis during 4 calendar weeks 11 and 12 [Gnädinger M 2016]. * Because of question ambiguity this analysis 5 6 was restricted to adult patients (age > 19 years); this led to 19,812 valid observations in the de 7 nominator and 183 in the incident groups. ** Per one conditions, medication, year or index point. 8 9 *** Because Evans’ Index is a composite of condition and medication numbers; it was not included 10 in the multiple regression analysis. Significance levels: a p<0.05, b p<0.01, c p<0.001. d Because 11 12 GENLIN procedure was not able to process ordinal scaled variables, correlations between study 13 14 group and TMI or care dependency were tested with Spearman’s rho: the correlation coefficients 15 were +0.075For or +0.094, respectively,peer p<0.001. review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 ATC Class All incidents Incidents with prob- Relative Risk Swiss 2015 sales* 4 able or definite (percentages left (number of packages) 5 relationship with by right column) 6 7 medication 8 9 A Alimentary tract and metabolism 28 (14.6%) 6 (15.8%) 1.06 31,455,252 (14.9%) 10 11 12 B Blood and blood forming organs 23 (12.0%) 7 (18.4%) 7.08 5,507,624 (2.6%) 13 14 C Cardiovascular system 44 (22.9%) 1 (2.6%) 0.35 16,027,143 (7.5%) 15 For peer review only 16 17 D Dermatologics 1 (0.5%) 0 (0.0%) 0.00 17,314,810 (8.2%) 18 19 G Genito urinary system and sex 1 (0.5%) 0 (0.0%) 0.00 7,936,641 (3.7%) 20 hormones 21 22 23 H Systemic hormonal preparations 7 (3.6%) 1 (2.6%) 2.00 2,875,760 (1.3%) 24 (excluding sex hormones and 25 insulins) 26 27 28 J Anti infectives for systemic use 23 (12.0%) 6 (15.8%) 3.95 8,444,623 (4.0%) 29
30 K Infusion liquids 0 (0.0%) 0 (0.0%) 0.00 24,158,749 (11.5%) 31 32
33 L Anti neoplastic and immunomodu 5 (2.6%) 1 (2.6%) 2.89 1,934,950 (0.9%) http://bmjopen.bmj.com/ 34 lating agents 35 36 37 M Musculo skeletal system 4 (2.1%) 2 (5.3%) 0.76 14,787,413 (7.0%) 38 39 N Nervous system 43 (22.4%) 10 (26.3%) 1.30 42,690,195 (20.2%) 40
41 on September 28, 2021 by guest. Protected copyright. 42 P Anti parasitic products, insecti 1 (0.5%) 0 (0.0%) 0.00 462,559 (0.2%) 43 cides and repellents 44 45 R Respiratory system 7 (3.6%) 3 (7.9%) 0.57 28,837,468 (13.7%) 46 47 48 S Sensory organs 2 (1.0%) 0 (0.0%) 0.00 7,658,311 (3.6%) 49 50 51 T Diagnostic use 0 (0.0%) 0 (0.0%) 0.00 43,184 (0.0%) 52 53 V Various 3 (1.6%) 1 (2.6%) 6.50 855,707 (0.4%) 54 55 56 Total 192 (100.0%) 38 (100%) 1.0 210,990,389 (100.0%) 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 does not apply 5 0 n.a. n.a. 4 5 6 Table 5. ACT-Groups of suspected medications. Relative risk of drugs with probable or definite 7 relationship with the incident as compared to sales proportions. * Information by Interpharma Swit 8 9 zerland (Appendix F). 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Study flow chart
33 http://bmjopen.bmj.com/ 34 254x190mm (300 x 300 DPI) 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 Distribution of the number of cases reported by practice http://bmjopen.bmj.com/ 34 35 136x106mm (300 x 300 DPI) 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 30 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 Type of error 37 38 148x129mm (300 x 300 DPI) 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 31 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 Organ system involved 35 36 147x120mm (300 x 300 DPI) 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 32 of 68
1 2 3
4 Appendix A Incident questionnaire BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 Medication Incidents in Primary Care (MIPC) 6 (Version 1.0, January 25th 2015) (Internet version not available in English) 7 8 9 10 11 Incident reporting form 12 13 14 15 16 Administrative Information 17 For peer review only 18 19 20 21 1. Sentinella identification number: 2. Week of reporting: 22 23 24 25 26 27 28 The Patient 29 30 31 32 33 3. Year of birth: 4. Gender: m f 34 35 36 37 5. What was your relationship to the patient when the incident happened? Were you the http://bmjopen.bmj.com/ 38 39 40 family physician emergency / substitute physician institution physician other 41 if other what kind………………………. 42 43 44 45 46 6. What is the patient’s living situation? on September 28, 2021 by guest. Protected copyright. 47 48 with partner / family alone institution unknown 49 50 51 52 53 7. Are there social problems? 54 55 yes no unknown 56 57 58 59 8. Is the patient demented or otherwise mentally handicapped? 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 68 BMJ Open
1 2 3 yes no unknown 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 8 9. Does he suffer from psychological problems? 9 10 11 yes no unknown 12 13 14 15 10. Does he take psychotropic drugs (esp. antidepressants, neuroleptics, benzodiazepines, 16 opiates)? 17 For peer review only 18 19 yes no unknown 20 21 22 23 11. Are there linguistic problems? 24 25 26 yes no unknown 27 28 29 30 12. Does he or she smoke? 31 32 33 yes no unknown 34 35 36 37 13. Is there substance abuse (other than nicotine)? http://bmjopen.bmj.com/ 38 39 yes no unknown , if yes, what substance? …………………….. 40 41 42 43 44 14. Does the patient have uncorrected / uncorrectable visual impairment? 45 on September 28, 2021 by guest. Protected copyright. 46 yes no unknown 47 48 49 50 51 15. Does the patient have uncorrected / uncorrectable hearing impairment? 52 53 54 yes no unknown 55 56 57 58 16. Does the patient have uncorrected / uncorrectable mobility impairment? 59 yes no unknown 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 68
1 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 17. Is there renal insufficiency (GFR: <60 ml/min/1.73 m2)? 7 8 yes no unknown 9 10 11 12 13 18. Is there hepatic insufficiency or liver cirrhosis? 14 15 16 yes no unknown 17 For peer review only 18 19 20 21 19. Was the patient hospitalized in the past 12 months? 22 yes no unknown 23 24 25 26 20. Is the patient taken care of by others? (only one answer permitted)? 27 28 29 yes, family / proxies yes, community nurse yes, institution no unknown 30 31 32 33 21. Number of regularly applied active substances (including non-daily applied ones, see 34 guidelines)? 35 36 37 unknown http://bmjopen.bmj.com/ 38 39 40 41 42 22. Number of chronic conditions (see guidelines)? 43 44 unknown 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 23. Scale value of „Thurgau Morbidity Index” (chronic part, see guidelines)? 50 51 unknown 52 53 54 55 56 57 58 59 60
3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 68 BMJ Open
1 2 3 Details of the incident
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 24. Please, give a short description of the incident (in block letters): 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 25. Who observed the incident? (multiple answers possible)? 23 24 Physician / staff patient / relatives community nurse home / institution hospi- 25 tal 26 27 other physicians pharmacist other unknown 28 for „other” please specify: …………………………… 29 30 31 32 33 26. What happened (multiple answers possible)? 34 35 dosage too high 36 dosage too low 37 http://bmjopen.bmj.com/ 38 application too short 39 application too long 40 41 wrong administration route 42 43 wrong medication 44 indicated medication not received 45 46 expired / defective medication on September 28, 2021 by guest. Protected copyright. 47 problems with insurance reimbursement 48 49 50 unclear / undefined 51 52 other (please specify): ……………………… 53 54 55 56 57 58 59 60
4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 68
1 2 3 27. Please state the trade name of the medication used in the incident:
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 ……………………………………………….. 6
7 8 9 28. Please note other medication names, presuming they are relevant to the case. 10 11 12 13 14 15 none unknown 16 17 For peer review only 18 19 29. How would you judge the degree of hazard to the patient during the incident? 20 21 mild medium severe none does not apply unknown 22 23 24 25 30. How intense was the impairment (as judged by the patient)? 26 27 28 mild medium severe fatal no symptoms, but pathological lab values 29 30 no impairment does not apply unknown 31 If there wasn’t any impairment, please skip to question 34. 32 33 34 35 31. How long did the impairment last? 36 37 hours days weeks longer unknown http://bmjopen.bmj.com/ 38 39 40 41 42 32. How was the recovery? 43 without residues with mild residues with severe residues / fatal unknown 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 33. Which organ system was affected (multiple answers possible)? 49 50 cardiovascular 51 central nervous 52 53 gastro-enteral 54 kidneys 55 56 liver 57 lung 58 59 skin 60 other, please specify …………………. 5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 68 BMJ Open
1 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 34. Did you have to apply a specific surveillance or treatment for the incident? 7 8 yes, ambulatory yes, hospital no unknown 9 If yes, please specify which? …………………………………………….. 10 11 12 13 14 35. What factors contributed causally to the emergence of the incident (multiple answers pos- 15 16 sible)? 17 off duty hours For peer review only 18 19 20 communication failure within practice 21 generic substitution by pharmacist 22 23 24 hand-written prescription incorrectly interpreted 25 conflicting multiple prescriptions 26 27 lack of alertness within practice 28 29 lack of documentation 30 insufficient patient instruction 31 32 lack of aids (e.g. Dosette®) 33 lack of cooperation by patient / relatives 34 35 misleading package leaflet information 36 patient’s internet search 37 http://bmjopen.bmj.com/ 38 administrative problems 39 40 manufacturer related (defective medication) 41 distributer related (out of stock) 42 43 lack of maintenance (e.g. first aid kit) 44 other, please specify……………………….. 45 46 unknown on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 36. Was there an interface problem? If yes, which (multiple answers possible)? 52 53 yes, with hospital 54 55 yes, with institution 56 yes, with community nurse 57 58 yes, with pharmacist 59 60
6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 68
1 2 3 yes, with specialist physician 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 yes, with other, please specify? …………… 6 7 8 no 9 10 If no, please skip to question 38. 11 12 13 37. Was there an explicit comparison of prescription lists with the institution / person? 14 15 16 yes, verbal / by phone yes, written / by fax no unknown 17 For peer review only 18 19 38/39. Was the patient informed about the incident? 20 21 22 yes, by myself / practice staff 23 no, because he was not able to understand the message (children, demented) 24 25 no, because the problem was solved and communication would have impaired confidence 26 no, because the patient had moved or was deceased 27 28 no, this was not needed because patient / relatives themselves had observed the incident 29 no, because others had already informed him 30 31 32 no, because: ………………………………………………………………………. 33 34 unknown 35 36 37 If yes, what was the patient’s reaction? …………………………………………………… http://bmjopen.bmj.com/ 38 39 40 41 40. What did you do as a result to prevent similar incidents in the future (multiple answers 42 possible)? 43 44 change standard operations procedures 45
better instruction of patients on September 28, 2021 by guest. Protected copyright. 46 47 communication with institution(s) 48 notification of manufacturer 49 50 notification of liability insurer 51 52 notification of drug authority („yellow leaflet“) 53 notification of the “critical incident reporting system” 54 55 other, please specify? ……………………….. 56 nothing 57 58 59 60
7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 68 BMJ Open
1 2 3 41. Who or what was ultimately responsible for the occurrence of the incident? 4 …………………………………………. BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 8 9 42. In the given situation, do you think one could have anticipated the event? 10 11 yes no 12 13 14 15 43. Did you already report an identical or very similar incident to this study? 16 yes no 17 For peer review only 18 19 20 21 44. Please make any suggestions about the kind of measures that could be taken to generally re- 22 duce the frequency of such events (in block letters): 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 Please keep a copy of this questionnaire in the patient files. Thank you for filling it out! 52 53
54 55 56 57 58 59 60
8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 68
1 2 3
4 Appendix B Initial questionnaire BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 Medication incidents in primary care (MIPC) 7 8 Initial reporting (Version 1.0, January 25th 2015) 9 Internet version: not available in English 10 11 1. Sentinella identification number: …………. 12 13 14 15 2. Number of physicians within your practice: …………. 16 17 For peer review only 18 3. Number of them who report to Sentinella …………. 19 20 21 22 4. Your weekly workload, h/w: <16 , 16-30 , >30 23 24 25 5. Number of hours within your practice (%): <50% , ≥50% 26 27 6. Approximate proportion of medication prescribed (as compared to directly delivered drugs): 28 <33% , 33-66% , >66% 29 30 31 7. Do you have an X-ray (machine?)? yes no 32 33 34 8. Do you have an ECG? yes no 35 36 37 9. Do you have an ultrasound? yes no http://bmjopen.bmj.com/ 38 39 40 10. Do you have an electronic system for controlling electronic drug interaction? 41 yes no 42 43 44 45 11. Do you have electronic patient history documentation? 46 yes no on September 28, 2021 by guest. Protected copyright. 47 48 49 12. Do you prescribe electronically? 50 yes, with a medication thesaurus yes, but without one (use of a typewriter) no 51 52 53 13. Is your practice certified (e.g. EQUAM)? yes no 54 55 56 14. Do you regularly schedule team sessions? 57 Yes, at least monthly , yes, but less frequently , no 58 59 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 68 BMJ Open
1 2 3 4 15. Do you attend quality circle sessions (in accordance with "Hausärzte Schweiz")? BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 yes, regularly , yes, now and then , no 6 7 8 9 16. Did you complete a special education (e.g. manual or psychosomatic medicine), or do you 10 have special interests (e.g. toxic maniac patients)? yes , no 11 if yes, please specify : ………………………………………………………………………. 12 13 14 17. Are you contracted by an institution? yes no 15 if yes, please specify (prison, home etc.): ……………………………………………………………… 16 17 18. If yes, doesFor this institution peer have specificreview problems with only medication? 18 yes , no , if yes, please specify: ………………………………………………………… 19 20 19. Are you involved in other special activities (teaching, research, insurance doctor)? 21 22 yes no 23 If yes, please specify the kind of activity: …………………………………………………….. 24 25 Thank you very much! 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 68
1 2 3
4 Appendix C Final questionnaire BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 Medication Incidents in Primary Care (MIPC) 7 8 th 9 Final questionnaire (Version 3.0 / December 14 2015) 10 11 Online version: not available in English 12 13 Dear colleagues, 14 15 Thank you for your dedicated support of our study on medication incidents in 2015, regardless of 16 whether you reported much of them or not. We would like to ask some final questions which will 17 help us put the otherFor information peer you sent review us in the right context only. 18 19 20 1. Sentinella identification number: …………… 21 22 2. Did you not report medication incidents that you had noticed during the last year (e.g. because 23 of lack of time)? 24 25 26 never or almost never , yes, but rarely , yes, frequently always or almost always 27 28 If this was frequently the case, please explain why:……………………………………….. 29 30 3. Did your practice participate in the fortnight morbidity denominator study in March 2015 (calen- 31 dar weeks 11 and 12)? 32 yes, fully , yes, but only partly (by omission of certain variables) , no 33 34 If you did not fully participate or not at all, what was the reason? 35 ……………………………………………………… 36 37 If you did not participate in the denominator study, please continue with question 12. http://bmjopen.bmj.com/ 38 39 4. How big was your effort for coding the morbidity variables of the denominator study? 40 manageable , rather big , too much , impossible 41 42 43 Did you have any difficulties when coding the morbidity variables? 44
45 5. Hospitalisation during the previous 12 months? none , a little , considerable , severe 46 on September 28, 2021 by guest. Protected copyright. 47 6. Care-dependency? none , a little , considerable , severe 48 49 7. Number of medications? none , a little , considerable , severe 50 51 8. Number of conditions? none , a little , considerable , severe 52 53 9. Thurgau Morbidity Index? none , a little , considerable , severe 54 55 10. Repeat consultation during the fortnight? none , a little , considerable , severe 56 57 If you named considerable to severe difficulties in questions 5 to 10, please list them in item 13. 58 59 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 43 of 68 BMJ Open
1 2 3 11. How long did it take for you and your practice nurse together for coding all of the variables of 4 one patient (Dr. Gnädinger needed less than 3 minutes)? BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 …………………………………………………. minutes 6 7 8 12. Would you be willing to be interviewed for a focus group on the subject of medication safety? 9 10 yes , no 11 12 13. Other comments: 13 14 15 16 17 For peer review only 18 19 20 21 22 Thank you very much! 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 44 of 68
1 2 3
4 Appendix D: electronic tables and figures BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 Figure e1: Distribution of the incident notifications over the year 6 Table e1: What went wrong with the incidents 7 Table e2: Organ system involved 8 Table e3: Causes of the incident 9 Table e4: Patient-sided possible risk factors 10 Table e5: Reactions to the incident 11 12 Table e6: Proposals to avoid further incidents 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 45 of 68 BMJ Open
1 2 3
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 Figure e1. Distribution of the incident notifications over the year. Because only the 41 week has been reported, the assignment to the month is approximate. 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 46 of 68
1 2 3
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 Type of error N* Percent*, % 6 7 8 Dosing too high 41 (12) 20.9 (6.1) 9 10 11 Dosing too low 21 (6) 10.7 (3.0) 12 13 14 Dosing too short 1 (2) 0.5 (1.0) 15 16 17 For peer review only 18 Dosing too long 5 (4) 2.6 (2.0) 19 20 21 Wrong way of administration 1 (1) 0.5 (0.5) 22 23 24 Wrong medication applied 56 (12) 28.6 (6.1) 25 26 27 Necessary medication not applied 12 (9) 6.1 (4.6) 28 29 30 Defective or expired medication applied 1 (1) 0.5 (0.5) 31 32 33 Problems with insurance reimbursing 1 (0) 0.5 (0.0) 34 35 36 37 Other problem** 31 (6) 15.8 (3.0) http://bmjopen.bmj.com/ 38 39 40 Unknown 1 (0) 0.5 (0.0) 41 42 43 Multiple naming 26 (n.a.) 13.3 (n.a.) 44 45 46 Total 197 (53) 100.0 (26.9) on September 28, 2021 by guest. Protected copyright. 47 48 49 50 Table e1. What went wrong with the incidents. * Parenthesis denotes the additional naming 51 within the category “multiple naming”. ** Naming within the category “other” were: confusion of sim- 52 ilar trade names (2), dosing error (6), erroneous package size (1), confusion of similar looking 53 preparations (1), error when controlling blood levels (2), double dosing (4), missed discontinuation 54 of medication (1), necessary treatment not applied (1), wrong vaccine applied (4), missed re-up- 55 take of anticoagulation after operation (1), contra-indication overlooked (2), drug-drug interaction 56 overlooked (3), known intolerance overlooked (3), necessary monitoring missed (2), transgression 57 of maximal drug allowance (2), delivery of incomplete package (1), confusion by contradictory 58 59 medication lists (3), incorrect handling of administering device (1), uncontrolled taking of “nature 60 medicines” (1), unreliable compliance with medication plan (2), two medications of the same drug
3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 47 of 68 BMJ Open
1 2 3 class (3), forbidden bisection of pills (2), confusion of similar named patients (1), intake of medica- 4 tion of the neighbor resident in a home (1), other (9). BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 48 of 68
1 2 3
4 Organ system N* Percent*, % BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 Cardiovascular 14 (10) 15.4 (11.0) 8 9 10 Central nervous system 23 (10) 25.3 (11.0) 11 12 13 Gastro-intestinal 5 (3) 5.5 (3.3) 14 15 16 Kidneys 2 (2) 2.2 (2.2) 17 For peer review only 18 19 20 Liver 2 (0) 2.2 (2.2) 21 22 23 Lung 2 (1) 2.2 (1.1) 24 25 26 Skin 6 (0) 6.6 (0.0) 27 28 29 Other** 24 (6) 26.4 (6.6) 30 31 32 Multiple naming 13 (n.a.) 14.3 (n.a.) 33 34 35 36 Total 91 (32) 100 (35.2) 37 http://bmjopen.bmj.com/ 38 39 Question does not apply 106 / 40 41 42 43 Table e2. Organ system involved. * Parenthesis denotes the additional naming within the category 44 45 “multiple naming”. ** Naming within the category “other” were: endocrine system (9), musculo-skel- 46 on September 28, 2021 by guest. Protected copyright. 47 etal (8), ear-nose-throat (2), psychic (1), other (9). 48 49 50 51 52 53 54 55 56 57 58 59 60
5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 49 of 68 BMJ Open
1 2 3 Causes of the incident N* Percent*;%
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 Out of hours 2 (3) 1.0 (1.5) 7 8 9 Communication problems within staff 10 (11) 5.1 (5.6) 10 11 12 Generic substitution of original trade medication by pharmacist 3 (3) 1.5 (1.5) 13 14 15 Difficulties when reading hand-written prescription 1 (1) 0.5 (0.5) 16 For peer review only 17 Multiple conflicting prescriptions 5 (8) 2.5 (4.1) 18 19 20 Lacking alertness of physician or practice staff 50 (17) 25.4 (8.6) 21 22 23 Insufficient documentation 3 (11) 1.5 (5.6) 24 25 Insufficient patient instruction 7 (17) 3.6 (8.6) 26 27 28 Lacking cooperation of patient / proxies 6 (17) 3.0 (8.6) 29 30 31 Confusion by reading package leaflet 1 (0) 0.5 (0.0) 32 33 34 Confusion after “Googleing” 1 (0) 0.5 (0.0) 35 36 Administrative problems 2 (3) 1.0 (1.5) 37 http://bmjopen.bmj.com/ 38 39 Defective medication as caused by manufacturer 3 (0) 1.5 (0.0) 40 41 42 Lacking maintenance (e.g. emergency case) 1 (0) 0.5 (0.0) 43 44 Lacking use of treatment aids (e.g. Dosett) 0 (6) 0.0 (3.0) 45 46 on September 28, 2021 by guest. Protected copyright. 47 Other source of trouble** 54 (18) 27.2 (9.1) 48 49 50 Unknown 3 (0) 1.5 (0.0) 51 52 53 Multiple naming 45 (n.a.) 23.1 (n.a.) 54 55 Total 197 (115) 100.0 (58.4) 56 57 58 Table e3. Causes of the incident. * Parenthesis denotes the additional naming within the category “multiple naming”. ** Naming within 59 the category “other” were: Erroneous delivery in pharmacy (5), treatment delayed/hampered by patient (2), transcription error (2), missed 60
6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 50 of 68
1 2 3 discontinuation of an ongoing treatment (2), communication problem within helpers’ network (16), similar trade names (2), interface prob-
4 lems with hospital (4), reading error of patient / proxies (3), erroneous execution of a medical prescription (14), missed follow-up control BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 (9), difficult handling of a preparation (2), disregard of possible interactions (2), transgression of competence by care workers (5), incom- 6 7 plete information by patient / proxies (8), stress / lack of time (12), other 25. 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 51 of 68 BMJ Open
1 2 3 Item Relevance All 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 (number of missing observations in groups “less” / n (%) 6 7 “more”) less more 8 n (%) n (%) 9 10 11 Number of patients 124 (100.0) 73 (100.0) 197 (100.0) 12 13 14 Living situation (1/0) 15 16 - together with mate / family 61 (49.6) 31 (42.5) 92 (46.9) 17 - alone For peer review28 (22.8) only 23 (31.5) 51 (26.0) 18 19 - institution 33 (27.6) 19 (26.0) 53 (27.0) 20 21 22 Social problems (2/4) 23 - yes 22 (18.1) 18 (26.1) 49 (20.9) 24 25 26 Dementia or mental illness (2/2) 27 28 - yes 31 (25.4) 18 (25.4) 49 (25.4) 29 30 31 32 Psychiatric problems (1/2) 33 34 - yes 28 (22.8) 28 (38.4)* 56 (28.9) 35 36 37 Treatment with psychotropic drugs (0/0) http://bmjopen.bmj.com/ 38 - yes 50 (40.3) 39 (53.4) 89 (44.2) 39 40 41 Linguistic problems (0/1) 42 43 - yes 11 (8.9) 5 (6.8) 16 (8.2) 44 45 46 Smoking (7/4) on September 28, 2021 by guest. Protected copyright. 47 - yes 11 (9.4) 8 (11.6) 19 (10.2) 48 49 50 Substance abuse other than nicotine (2/1) 51 52 - yes 3 (2.5) 5 (6.9) 8 (4.1) 53 54 55 Visual blurring (6/6) 56 - yes 4 (3.4) 4 (6.0) 8 (4.3) 57 58 59 60
8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 52 of 68
1 2 3 Hearing problems (3/5) 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 - yes 8 (6.6) 4 (5.9) 12 (6.3) 6 7 8 Gait problems (2/2) 9 10 - yes 40 (32.8) 18 (25.4) 58 (30.1) 11 12 13 Renal insufficiency (GFR<60 ml/min*1.73m2) (6/2) 14 - yes 24 (20.3) 21 (29.6) 45 (23.8) 15 16 17 For peer review only 18 19 Liver cirrhosis of other hepatic function problem 20 (3/1) 3 (2.5) 3 (4.2) 6 (3.1) 21 22 - yes 23 24 25 Table e4. Patient-sided possible risk factors. * p=0.021 “more” vs. “less” by chi-square testing. 26 We calculated the variable “incident relevance” from the variables “disturbance” and “endangering” 27 28 of the patients; if any of them was graded with “medium” or higher, the variable relevance was set to 29 30 “more”, otherwise to “less”. 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
9
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 53 of 68 BMJ Open
1 2 3 Reactions to the incident N* Percent*; % 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 Changing standard operating procedures of the practice 14 (3) 7.2 (1.5) 8 9 10 Better instruction of patients 23 (4) 11.7 (2.0) 11 12 13 Communication with other institutions 31 (3) 15.7 (1.5) 14 15 16 Notifying manufacturer 1 (2) 0.5 (1.0) 17 For peer review only 18 19 Reporting the incident to the critical incident reporting system 10 (6) 5.1 (3.0) 20 21 22 No reaction at all 55 (n.a.) 27.9 (n.a) 23 24 25 Other type of reactions to the incident** 49 (9) 24.9 (4.6) 26 27 28 Missing information 1 (0) 0.5 (0.0) 29 30 31 Multiple naming 13 (n.a.) 6.6 (n.a.) 32 33 34 Total 197 (27) 100.0 (13.7) 35 36 37 Table e5. Reactions to the incident. * Parenthesis denotes the additional naming within the category “multi- http://bmjopen.bmj.com/ 38 ple naming”. ** Naming within the category “other” were: Hire more workforce (1), arrangements with other 39 40 physicians (2), with pharmacist (4), with community nurse (4), with institution (12), with practice nurse (6), with 41 specialist (1), with patient (3), with supplier (1), sending a new medication plan (1), having regular staff meet- 42 43 ings (1), remove Digoxin 0.25 from assortment because of safety reasons (1), to not perform vaccinations in 44 absence of vaccination card (2), to clarify intolerances (2), to clarify interactions (1), to hand out an allergy card 45 46 (1), to actualize patient records (3), to apply for insurance cost credit (1), to organize follow-up controls (20), on September 28, 2021 by guest. Protected copyright. 47 other (5). 48 49 50 51 52 53 54 55 56 57 58 59 60
10
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 54 of 68
1 2 3 Proposal N* Percent*;%
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 7 Cross-check of medication lists 6 (14) 4.8 (11.2) 8 Patient instructions 2 (3) 1.6 (2.4) 9 10 Reduction of time pressure / stress 3 (6) 2.4 (4.8) 11 To observe adverse drug reactions also with “nature 1 (0) 0.8 (0.0) 12 13 products” 14 15 Four eyes check when dispensing medication 2 (4) 1.6 (3.2) 16 17 No medication Forwithout prescription peer review only3 (6) 2.4 (4.8) 18 19 To critically audit polymedication 3 (2) 2.4 (1.6) 20 21 To avoid similarly looking or named medication in drug 2 (3) 1.6 (2.4) 22 23 master 24 To demand medication plans to be brought to the con- 0 (4) 0.0 (3.2) 25 26 sultation 27 28 To provide medication plans routinely 2 (8) 1.6 (6.4) 29 30 In-deep checking before delivering “new” medication to 4 (1) 3.2 (0.8) 31 the patient 32 33 To provide patients with allergy / intolerance cards 2 (1) 1.6 (0.8) 34 35 To let patients themselves write their medication card 0 (3) 0.0 (2.4) 36 37 (and controlling afterwards) http://bmjopen.bmj.com/ 38 To instruct patients to come into the practice immedi- 3 (0) 2.4 (0.0) 39 40 ately after hospitalization 41 42 To broach regularly “medication safety” on staff meet- 5 (6) 4.0 (4.8) 43 ings 44 45 To provide information in patients’ native language 0 (2) 0.0 (1.6) 46 on September 28, 2021 by guest. Protected copyright. 47 To share important information about patients with prac- 0 (2) 0.0 (1.6) 48 49 tice nurse 50 To improve information flow 11 (16) 8.8 (10.4) 51 52 To organize follow-up controls 10 (28) 8.0 (22.4) 53 54 To wait for laboratory results before prescribing 1 (0) 0.8 (0.0) 55 56 Other** 12 (9) 9.6 (7.2) 57 58 59 Multiple answering 53 (n.a.) 42.4 (n.a.) 60
11
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 55 of 68 BMJ Open
1 2 3 Total 125 (118) 100.0 (94.4)
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 None 72 (n.a.) (n.a.) 6 7 8 9 Table e6. Proposals to avoid further incidents. * Parenthesis denotes the additional naming 10 11 within the category “multiple naming”. ** Among naming within the category “other” were: timely 12 13 involving of community nurse, stopping of self-medication in multimorbid or poly-medicated patients, 14 to clearly labelling desensitization suspensions, to separating adult and pediatric vaccines, to imple- 15 16 menting timely new guidelines, to refusing to be in care of a patient with bad compliance, to clearly 17 For peer review only 18 separating and labelling of medication prepared for different patients, to advice patients for separat- 19 ing short- and long-acting insulins at their home, to providing short-time prescriptions with a clear 20 21 expiration date, to allowing only adequately educated people to align and deliver medication in 22 homes for the elderly and that delivering may be done by the same person who did the preparing of 23 24 medication boxes, to avoiding similar looking medication boxes being muddled-up by patients at the 25 26 refectory room. 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
12
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 56 of 68
1 2 3
4 Appendix E BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 7 8 Item 9 No Recommendation 10 Title and 1 (a) Indicate the study’s design with a commonly used term in the title or the ab- 11 abstract stract [yes] 12 13 (b) Provide in the abstract an informative and balanced summary of what was 14 done and what was found [yes] 15 16 Introduction 17 For peer review only Background/ra- 2 Explain the scientific background and rationale for the investigation being re- 18 tionale ported [yes] 19 20 Objectives 3 State specific objectives, [yes] including any prespecified hypotheses [no] 21 22 23 Methods 24 Study design 4 Present key elements of study design early in the paper [yes] 25 26 Setting 5 Describe the setting, locations, and relevant dates, including periods of recruit- 27 ment, exposure, follow-up, and data collection [yes] 28 29 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of par- 30 ticipants [yes] 31 32 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and 33 effect modifiers. [yes] Give diagnostic criteria, if applicable [no] 34 35 Data sources/ 8* For each variable of interest, give sources of data and details of methods of 36 measurement assessment (measurement). Describe comparability of assessment methods if 37 there is more than one group [yes] http://bmjopen.bmj.com/ 38 39 Bias 9 Describe any efforts to address potential sources of bias [yes] 40 41 Study size 10 Explain how the study size was arrived at [yes] 42 43 Quantitative va- 11 Explain how quantitative variables were handled in the analyses. If applicable, 44 riables describe which groupings were chosen and why [yes] 45 46 on September 28, 2021 by guest. Protected copyright. Statistical me- 12 (a) Describe all statistical methods, including those used to control for con- 47 thods founding [yes] 48 49 (b) Describe any methods used to examine subgroups and interactions [yes] 50 51 52 (c) Explain how missing data were addressed [yes] 53 54 (d) If applicable, describe analytical methods taking account of sampling strat- 55 egy [no] 56 57 (e) Describe any sensitivity analyses [no] 58 59 Results 60
1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 57 of 68 BMJ Open
1 2 3 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers poten-
4 tially eligible, examined for eligibility, confirmed eligible, included in the study, BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 completing follow-up, and analysed [no] 6 7 (b) Give reasons for non-participation at each stage [yes] 8 9 (c) Consider use of a flow diagram [yes] 10 11 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) 12 and information on exposures and potential confounders [yes] 13 14 (b) Indicate number of participants with missing data for each variable of inter- 15 est [yes] 16 17 For peer review only Outcome data 15* Report numbers of outcome events or summary measures [yes] 18 19 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted esti- 20 mates and their precision (eg, 95% confidence interval). Make clear which con- 21 founders were adjusted for and why they were included [yes] 22 23 (b) Report category boundaries when continuous variables were categorized 24 [yes] 25 26 27 (c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful time period [no] 28 29 30 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and 31 sensitivity analyses [yes] 32 33 Discussion 34 Key results 18 Summarise key results with reference to study objectives [yes] 35 36 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or 37 imprecision. Discuss both direction and magnitude of any potential bias [yes] http://bmjopen.bmj.com/ 38 39 Interpretation 20 Give a cautious overall interpretation of results considering objectives, limita- 40 tions, multiplicity of analyses, results from similar studies, and other relevant 41 evidence [yes] 42 43 Generalisability 21 Discuss the generalisability (external validity) of the study results [yes] 44 45 46 Other information on September 28, 2021 by guest. Protected copyright. 47 Funding 22 Give the source of funding and the role of the funders for the present study 48 and, if applicable, for the original study on which the present article is based 49 [yes] 50 51 52 *Give information separately for exposed and unexposed groups. 53 54 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 55 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 56 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 57 58 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is avail- 59 able at www.strobe-statement.org 60
2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 58 of 68
1 2 3 Appendix F 4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 6 Swiss Pharmaceutical Market: Number of packaged sold by ATC-groups, 2015 7 8 9 10 11 12 2015 13 number of packa-
14 ged sold 15 Total 210,992,389 16 A Alimentäres System und Stoffwechsel 31,455,252 17 A01 Stomatologika,For Medizinische peer Präparate zurreview Mund- und Zahnpflege only 1,473,520 18 A02 Antacida-Ulcustherapheutika, Antiflatulentia 6,051,251 19 A02A Antacida,Antiflatulentia 1,500,822 20 21 A02B Ulcustherapeutika 4,426,496 22 A02C Sonstige Magentherapeutika 123,933 23 A03 Produkte gegen funktionelle Magen-Darm-Störungen 2,087,299 24 A03A Antispasmodika+Anticholinergika rein 595,861 25 A03C Antispasmodika/Ataraktika-Kombinationen 25,411 26 27 A03E Antispasmodika, sonstige Kombinationen 222,102 28 A03F Gastroprokinetika 1,233,461 29 A03G Modulatoren der gastrointestinalen Sensomotorik 10,464 30 A04 Antiemetica und Antinausea 920,841 31 A05 Cholagoga und Leberschutz-Mittel 337,970 32 33 A05A Gallentherapeutica und Cholagoga 254,744 34 A05B Leberschutzpräparate 83,226 35 A06 Mittel gegen Verstopfung und Darmreinigung 4,583,118 36 A06A Mittel Gegen Verstopfung 4,329,037 37 A06B Darmreinigungsmittel 254,081 http://bmjopen.bmj.com/ 38 39 A07 Antidiarrhoica, Elektrolytzufuhr und intestinale Antiphlogistica 3,005,035 40 A07B Intestinale Adsorbierende Antidiarrhoica 228,076 41 A07E Intestinale Antiphlogistica 222,680 42 A07F Antidiarrhoica Micro-Organismen 1,340,432 43 A07G Orale Elektrolyt-Zufuhr 107,030 44 45 A07H Motilitätshemmer 1,095,216 46 A07X Übrige Antidiarrhoica inkl. A07A Antidiarrhoica intestinale Antiinfectiva 11,601 on September 28, 2021 by guest. Protected copyright. 47 A08 Antiadiposita excl. Diätetica 51,654 48 A09 Digestiva und Enzyme 804,667 49 A10 Antidiabetica 3,376,580 50 51 A10C Humaninsulin und Analoga 1,078,495 52 A10H Sulfonylharnstoff Antidiabetika 321,219 53 A10J Biguanid Antidiabetika 1,217,578 54 A10K Glitazone Antidiabetika 25,305 55 A10M Glinide Antidiabetika 38,782 56 57 A10N DPP-IV Inhibitor Antidiabetika 483,559 58 A10P SGLT2-Hemmer Antidiabetika 45,927 59 A10S GLP-1 Agonisten Antidiabetika 156,656 60 A10X Übrige Antidiabetika inkl. Insulin tierischen Ursprungs und Alphagluko- 8,616 sidaseinhibitor Antidiabetika 1
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 59 of 68 BMJ Open
1 2 3 A11 Vitamine 5,001,408
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from A11A Multivitamine mit Mineralstoffen 755,243 5 6 A11B Multivitamine ohne Mineralstoffe 17,507 7 A11C Vitamin A+D inkl. einfache Kombinationen 2,618,699 8 A11D Vitamin B1 und Kombinationen 170,884 9 A11E Vitamin B-Komplex 739,860 10 A11F Vitamin B 12 rein 123,445 11 12 A11G Vitamin C, inkl. Kombinationen mit Mineralstoffen 353,644 13 A11X Sonstige Vitamine 222,126 14 A12 Mineralverbindungen 3,490,891 15 A12A Mineralverbindung Calcium 1,254,334 16 A12B Mineralverbindung Kalium 200,438 17 For peer review only 18 A12C Sonstige Mineralverbindungen 2,036,119 19 A13 Tonica und Roborantia 229,735 20 A16 Übrige Präparate des alimentären und Stoffwechsel-Systems 34,295 21 B Blut und Blutbildende Organe 5,507,624 22 B01 Antithrombotika 4,009,465 23 24 B01A Vitamin-K-Antagonisten 399,963 25 B01B Heparine 714,872 26 B01C Thrombozytenaggregationshemmer 2,265,838 27 B01D Fibrinolytika 14,293 28 B01X Direkte Thrombin-Inhibit und direkte Faktor XA-Hemmber 598,766 29 30 B02 Blutgerinnungssystem – Sonstige Produkte 237,495 31 B02B Antagonisten (Antidot.Anticoag) 146,201 32 B02D Blutcoagulation 1,805 33 B02G Glitazone-Antidiabetika 48,969 34 B02X Übrige Blutgerinnungsprodukte (Tissues sealing preparations, Throm- 40,520 35 bopoietin Agonisten, Antifibrinolytica synthetisch etc.) 36 B03 Antianämica 1,259,896 37 B03A Antianämica mit Eisen und Kombinationen 862,744 http://bmjopen.bmj.com/ 38 39 B03C Erythropoietin-Präparate 67,258 40 B03X Sonstige Antianämica inkl. Folsäure/Folinsäure 329,894 41 C Herz-Kreislauf Therapie 16,027,143 42 C01 Cardica 954,962 43 C01B Anti-Arrhythmie-Präparate 213,952 44 45 C01C Cardiale Stimulantien excl. Herzglykoside 266,028 46 C01D Coronartherapeutica excl. Calcium-Antagonisten und Nitropräparate 121,815 on September 28, 2021 by guest. Protected copyright. 47 C01E Nitropräparate und Analog 228,438 48 C01X Positive Inotrope, Herzglykoside und übrige Herzpräparate 124,729 49 C02 Antihypertonica 129,957 50 51 C03 Diuretica 1,815,738 52 C04 Cerebrale und periphere Vasotherapeutika 390,255 53 C05 Antivaricosa/Antihaemorrhoidalia 2,271,175 54 C05A Antihaemorrhoidalia topisch 773,800 55 C05B Antivaricosa topisch 1,029,010 56 57 C05C Antivaricosa systemisch 468,365 58 C06 Übrige Herz+Kreislaufmittel 339,930 59 C07 Betablocker 2,398,309 60 C07A Betablocker rein 2,272,988 2
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 60 of 68
1 2 3 C07B Betablocker Kombination 125,321
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from C08 Calciumantagonisten 1,157,267 5 6 C09 Stoffe mit Wirkung auf das Renin-Angiotensin-System 4,116,677 7 C09A Ace-Hemmer, rein 1,142,465 8 C09B Ace-Hemmer, Kombinationen 584,697 9 C09C Angiotensin-II-Antagonisten, rein 1,084,218 10 C09D Angiotensin-II-Antagonisten, Kombinationen 1,261,616 11 12 C09X Sonstige Stoffe mit Wirkung auf das Renin - Angiotensin - System 43,681 13 C10 Lipidregu/Antiarteriosklerotika 2,446,927 14 C10A Lipidregulatoren 2,341,963 15 C10A1 Statine (Cholesterolsynthesehemmer) 2,190,227 16 C10A2 Lipidsenker, Fibrate, Ionenaustauscher und sonstige Lipidsenker 151,736 17 For peer review only 18 C10C Lipidregulatoren kombiniert mit anderen Lipidregulatoren 104,964 19 C11 Übrige Präparate für die Herz-Kreislauf-Therapie 5,946 20 D Dermatologica 17,314,810 21 D01 Dermatologische Antimykotica 1,360,570 22 D02 Emollientia und Hautschutzmittel 2,703,273 23 24 D03 Wundheilmittel 4,389,791 25 D04 Antipruriginosa inkl. Antihistaminica und Anästhetica usw. 1,458,276 26 D05 Nichtsteroidale Präparate gegen entzündliche Hauterkrankungen 327,245 27 D06 Topische antibakterielle Produkte und antivirale Mittel 864,829 28 D06A Topische antibakterielle Produkte 379,013 29 30 D06D Topische Produkte gegen Virusinfektionen 485,816 31 D07 Corticosteroide topisch 1,982,173 32 D07A Reine Corticosteroide topisch 960,182 33 D07B Corticosteroid-Kombinationen topisch 1,021,991 34 D08 Antiseptica + Desinfizientia 1,836,788 35 36 D10 Aknemittel 1,227,850 37 D10A Topische Aknemittel 829,114 http://bmjopen.bmj.com/ 38 D10B Orale Aknemittel 398,736 39 D11 Übrige Dermatologica 1,164,015 40 G Urogenital-System+Sexualhormon 7,936,641 41 42 G01 Gynaekologische Antiinfektiva 812,985 43 G01A Trichomonadenmittel 185,395 44 G01B Gynaekologische Antimykotica 475,911 45
G01X Gynaekologische antibakterielle Produkte und Antiseptika 151,679 on September 28, 2021 by guest. Protected copyright. 46 G02 Sonstige Gynaekologica 1,060,376 47 48 G02A Oxytocica 91,111 49 G02F Topische Sexualhormone 496,735 50 G02X Lokale Antikonzipientia, Prolactininhibitoren, wehenhemmende Mittel 472,530 51 und sonstige Gynaekologika 52 G03 Sexualhormone und Stimulantien des Genitalsystems 3,917,793 53 G03A Hormonale Kontrazeptiva systemisch 2,477,919 54 G03B Androgene, excl.G3E,G3F 70,852 55 G03C Oestrogene, excl.G3A,G3E,G3F 397,050 56 57 G03D Progestogene, excl. G3A,G3F 245,149 58 G03F Oestrogen-Progestogen-Kombinationen 487,291 59 G03G Gonadotropin inl. sonstige Ovulationsstimulantien 100,879 60
3
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 61 of 68 BMJ Open
1 2 3 G03X Androgen-Kombinationen mit weiblichen Sexualhormonen, SERMS und 138,653
4 sonstige Sexualhormone und ähnliche Produkte BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 G04 Urologica 2,145,487 6 G04A Urologische Antiinfektiva und Harnantiseptika 410,545 7 G04C BPH Produkte (Benigne Prostata-Hyperplasie) 715,845 8 9 G04D Harninkontinenzprodukte 390,832 10 G04E Produkte gegen Erektionsstörungen 311,659 11 G04X Alle sonstigen Urologika 316,606 12 H Hormonales System, exkl. Sexualhormone 2,875,760 13 H01 Hypophysen-/Hypothalamus-Hormone 36,944 14 15 H02 Corticosteroide Systemisch 1,575,880 16 H03 Schilddruesentherapie 1,109,934 17 H04 Sonstige HormoneFor (wachstumshormone, peer review antidiuretische Hormone, only Anti-Pa- 153,002 18 rathyroid-Hormone etc.) 19 J Antiinfektiva systemisch 8,446,623 20 J01 System.Antibakter.Prod. 5,721,173 21 J01A Tetracycline und Kombinationen 155,659 22 J01C Breitspektrum Penicilline 2,353,136 23 24 J01D Cephalosporine 897,839 25 J01F Macrolide und vergleichbare Substanzen 778,785 26 J01G Fluorchinolone 817,056 27 J01H Mittel- und Schmalspektrum-Penicilline 111,484 28 J01K Aminoglykoside 46,447 29 J01P Sonstige Beta-Lactam. Antibiot. Penicilline, Cephalosporine (Monobac- 30 69,074 tame, Peneme und Carbapeneme) 31 J01X Rifamycin und vergleichbare Substanzen, Trimethoprim, Kombinationen 491,693 32 und vergleichbare Substanzen und sonstige antibakterielle Produkte 33 J02 Antimykotika, Systemisch 591,346 34 35 J04 Tuberkulosemittel 37,320 36 J05 Antivirale Mittel, systemisch 570,063 37 J05B Antivirale Mittel, exkl. Produkte gegen HIV 236,676 http://bmjopen.bmj.com/ 38 J05C Virustatika gegen HIV 333,387 39 J05C1 Nucleoside und Nucleotide Reverse Transkriptase Hemmer 108,107 40 41 J05C2 Protease Hemmer 98,581 J05CX Nicht-Nucleoside Reverse Transkriptase Hemmer, HIV Fusionshem- 42 126,699 mer, HIV-Antivirale Integrase Hemmer und sonst.HIV-Antivirale Mittel 43 44 J06 Sera und Gammaglobulin 23,164 45 J07 Vakzine 1,406,674 46 J07B Vakzine Kombinationen 463,631 on September 28, 2021 by guest. Protected copyright. 47 J07B1 Kombinationen mit Tetanusimpfstoff 387,139 48 J07BX Mehrfach-Impfstoffe mit Masern und/oder Mumps und alle anderen 76,492 49 Kombinationen 50 J07D Bakterielle Impfstoffe 164,542 51 J07E Virale Impfstoffe 687,366 52 53 J07E1 Influenza-Impfstoff 205,391 54 J07E4 Hepatitis-Impfstoff 212,503 55 J07E9 Alle uebrigen viralen Impfstoffe, inkl. HPV, Rotaviren, Varizellen etc. 269,472 56 J07X Alle uebrigen Impfstoffe und aehnliche Produkte 91,135 57 J08B Übrige Antiinfektiva 96,883 58 59 K Infusionslösungen 24,158,749 60 K01 Intravenoese Lösungen ab 100 ML 12,064,154
4
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 62 of 68
1 2 3 K01A Elektrolytlösungen 3,372,213
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from K01B Standardlösungen 8,461,628 5 6 K01D Fettemulsionen, inkl. Produkte zur totalen parenteralen Ernährung 156,484 7 K01E Aminosäuren-Lösungen 51,140 8 K01X Kalorische Lös. über 10% und Lösungen zur Osmotherapie 22,689 9 K02 Plasma Expander 70,547 10 K03 Blut- und Blutersatzinfusionslösungen 1,817 11 12 K04 Injektionslösungen/Infusionszusätze 12,005,621 13 K04A Elektrolytlösungen 752,499 14 K04B Standardlösungen 9,981,186 15 K04C Kalorische Lösungen < 100 ML 34,050 16 K04D Sonstige Injektionslösungen/Infusionszusätze < 100 ML 1,237,886 17 For peer review only 18 K05 Spüllösungen 16,610 19 L Antineoplastika und Immunmodulatoren 1,934,950 20 L01 Antineoplastika 984,072 21 L01A Alkylisierende Substanzen 46,698 22 L01B Antimetaboliten 323,741 23 24 L01C Vinca-Alkaloide & sonstige pflanzliche Produkte 157,793 25 L01D Antineoplastisch wirkende Antibiotika 46,619 26 L01F Platinhalt.Antineoplast. 95,347 27 L01G Monoklonale Antikörper zur antineoplastischen Therapie 159,208 28 L01H Proteinkinasehemmer zur antineoplastischen Therapie 33,895 29 30 L01X Sonstige Antineoplastika 120,771 31 L02 Cytostatische Hormonetherapie 183,379 32 L02A Cytostatische Hormone 43,050 33 L02B Cytostatische Hormonantagonisten 140,329 34 L03 Immunostimulantien 79,294 35 36 L03A Immunstimulantien 40,990 37 L03B Interferone 38,304 http://bmjopen.bmj.com/ 38 L04 Immunsuppressiva 688,205 39 L04B Anti-TNF-Produkte 340,857 40 L04X Sonstige Immunsuppressiva inkl. Interleukin Inhibitoren 347,348 41 42 M Muskel-und Skelettsystem 14,787,413 43 M01 Antirheumatica 6,373,707 44 M01A Corticoidfreie Antirheumatica 5,700,326 45
M01C Antirheumatica spezifisch 673,381 on September 28, 2021 by guest. Protected copyright. 46 M02 Rheumaeinreibungen Rubefatientia 6,372,249 47 48 M03 Muskelrelaxantien 665,391 49 M03A Muskelrelaxantien peripher wirkend 85,757 50 M03B Muskelrelaxantien zentral wirkend 579,634 51 M04 Gichtmittel 368,020 52 M05 Übrigige Muskel-Skelettsystem-Präparate 1,008,046 53 54 N Nervensystem 42,690,195 55 N01 Anaesthetica 1,554,391 56 N01A Narkosemittel 1,078,794 57 N01A1 Narkosemittel Inhalationspräparate 20,582 58 N01A2 Narkosemittel Injektionspräparate 1,058,212 59 60 N01B Lokalanaesthetika 475,597
5
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 63 of 68 BMJ Open
1 2 3 N01B1 Medizinische injizierbare Lokalanästhetika 198,964
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from N01B3 Topische Lokalanästhetika 270,962 5 6 N01B9 Sonstige Lokalanästhetika und dentale injizierbare Lokalanästhetika 5,671 7 N02 Analgetica 24,080,469 8 N02A Betäubungsmittel 1,661,607 9 N02B Analgetica und Antipyretica 21,918,098 10 N02B1 Analgetica rezeptpflichtig 8,592,792 11 12 N02B2 Analgetica rezeptfrei 13,325,306 13 N02C Migränemittel 500,764 14 N02C1 Triptane 432,741 15 N02C9 Sonstige Migränemittel 68,023 16 N03 Antiepileptica 1,957,745 17 For peer review only 18 N04 Parkinsonmittel 658,710 19 N05 Psycholeptica 9,075,955 20 N05A Antipsychotika 2,002,404 21 N05A1 Atypische Antipsychotika 1,521,618 22 N05A9 Konventionelle Antipsychotika 480,786 23 24 N05B Hypnotica und Sedativa 4,462,062 25 N05B1 Hypnotica und Sedativa, barbituratfreie Reinsubstanzen 2,830,382 26 N05B2 Hypnotica und Sedativa, barbituratfreie Kombinationen 129,461 27 N05B5 Pflanzliche Hypnotica und Sedativa 1,502,219 28 N05C Tranquilizer 2,611,489 29 30 N06 Psychoanaleptica Ex.Antiadipos 4,081,575 31 N06A Antidepressiva und stimmungsstabilisierende Produkte 3,503,509 32 N06A2 Pflanzliche Antidepressiva 245,372 33 N06A3 Stimmungsstabilisierende Produkte 50,396 34 N06A4 SSRI Antidepressiva 1,319,897 35 36 N06A5 SNRI Antidepressiva 664,070 37 N06A9 Sonstige Antidepressiva 1,223,774 http://bmjopen.bmj.com/ 38 N06B Psychostimulantia 341,172 39 N06X Psycholeptica-Psychoanaleptica-Kombinationen, Nootropica und Neuro- 236,894 40 tonica sowie verschiedene andere Produkte 41 N07 Sonstige ZNS-wirksame Produkte 1,281,350 42 N07B Raucherentwöhnungsmittel 652,146 43 N07C Produkte gegen Schwindel 286,538 44 45 N07D Anti-Alzheimer Produkte 121,928 46 N07E Alkohol- und Opiatentwöhnungsmittel 126,973 on September 28, 2021 by guest. Protected copyright. 47 N07X Alle anderen ZNS-wirksamen Präparate 93,765 48 P Parasitologie 462,559 49 P01 Antiprotozoenmittel und Anthelmintika 437,402 50 51 P01B Anthelmintica, excl. Schistosomiasis-Mittel 296,327 52 P01D Malariamittel 139,242 53 P02 Sonstige Antiparasitäre Präparate inkl. Insektizide und Repellentien 26,990 54 R Respirationssystem 28,837,468 55 R01 Rhinologica 7,246,454 56 57 R01A Rhinologica topisch 6,868,431 58 R01AX Corticosteroide Rhinologika 1,067,991 59 R01A6 Antiallergische Rhinologica 230,180 60 R01A7 Abschwellende Rhinologica 4,290,641 6
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 64 of 68
1 2 3 R01A9 Sonstige Rhinologica topisch inkl. Antiinfektive Rhinologika ohne 1,279,619
4 Corticosteroide BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 5 R01B Rhinologica systemisch 378,023 6 R02 Halsschmerzmittel und Antiinfectiva 3,977,806 7 R03 Antiasthmatika und Produkte gegen chronisch-obstruktive Lungenerkran- 2,931,292 8 kungen 9 R03A B2-Stimulatoren 923,570 10 11 R03B Xanthine 16,188 12 R03C Degranulationshemmer 9,088 13 R03D Corticoide 258,075 14 R03F Komb. B2-Stimulatoren mit Corticoiden 1,172,896 15 R03G Anticholinergica rein und in Komb. mit B2-Stimulatoren 422,479 16 17 R03G3 AnticholinergicaFor peerrein, Inhalate review only 269,418 18 R03G4 Anticholinergica – Kombinationen mit Beta-2-Stimulantien, Inhalate 153,061 19 R03J Antileukotrien Antiasthmatika 90,062 20 R03X Sonstige Antiasthmatika und Produkte gegen chronisch-obstruktive Lun- 21 generkrankungen inkl. PDE4-Hemmer gegen Asthma / Chronisch-Obstruktive 38,934 22 Lungenerkrankungen 23 R04 Percutane Einreibemittel und Inhalationspräparate 1,093,346 24 R05 Husten- und Erkältungspräparate 10,899,026 25 R05A Erkältungspräparate 2,498,594 26 27 R05C Expectorantia ohne Antiinfectiva 4,499,364 28 R05D Hustensedativa 3,297,144 29 R05D1 Hustenmittel plain (Einzelsubstanzen) 1,971,766 30 R05D2 Sonstige Hustensedativa, inkl. Kombinationen 1,325,378 31 R05F Sonstige Husten- und Erkältungspräparate 603,924 32 33 R06 Antihistamine systemisch 2,687,483 34 R07 Sonstige Präparate des Respirationssystems 2,061 35 S Sinnesorgane 7,658,311 36 S01 Ophthalmologica 7,214,097 37 S01A Ophtalmologische Antiinfectiva 451,604 http://bmjopen.bmj.com/ 38 39 S01B Ophtalmologische Corticoide rein 271,663 40 S01C Ophthalm. Antiphlogistika und Antiinfektiva Kombinationen 613,682 41 S01E Myotika und Mittel zur Glaukombehandlung 1,038,592 42 S01F Mydriatika und Cykloplegika 46,547 43 S01G Ophthalmologische Antiallergika, Abschwellende Mittel, Antiseptika 1,473,188 44 45 S01H Ophtalmolog. Lokalanesthetika 40,383 46 S01K Künstliche Tränen und Netzmittel für Augen 2,503,750 on September 28, 2021 by guest. Protected copyright. 47 S01P Ophthalmologische Antineovaskularisationsprodukte 127,501 48 S01R Nonsteroidale entzündungshemmende Ophthalmologica (NSAID'S) 236,207 49 S01X Übrige Ophtalmologica inkl. Präparate für den Gebrauch von Kontaktlin- 50 sen, Augentonika und Augenvitamine, Präparate zur Verhütung von Katarakt und 411,009 51 Antikataraktogenika, Ophthalmologische Operationshilfsmittel und Ophthalmologi- 52 sche Diagnostika 53 S02 Otologica 444,214 54 T Diagnostika 43,184 55 T01 Diagnostika für bildgebende Verfahren 18,885 56 57 T02 Testdiagnostika 24,299 58 V Verschiedenes 855,707 59 V01 Allergene 6,429 60 V03 Übrige Therapeutische Präparate 820,500
7
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 65 of 68 BMJ Open
1 2 3 V03D Entgiftungspräparate für die Cytostatikatherapie 45,668
4 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from V03E Antidote 4,532 5 6 V03G Hyperkaliämie/Hyperphosphatämie-Produkte 42,259 V03X Sonstige therapeutische Präparate inkl. Entzündungshemmende En- 7 728,041 8 zyme, Eisen-Chelatbildner und Radiopharmazeutika 9 V05 Chirurgische Antiseptica 134 10 V06 Allgemeine Nährmittel 5,612 11 V07 Alle übrigen nicht therapeutischen Präparate 23,032 12 13 14 15 Source: Interpharma with data from IMS Health Schweiz, 2016. 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 46 on September 28, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
8
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 66 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 4 Item Manu- 5 No script 6 Recommendation page 7 Title and abstract 1 (a) Indicate the study’s design with a commonly used 1 8 term in the title or the abstract [yes] 9 10 (b) Provide in the abstract an informative and 1-2 11 balanced summary of what was done and what was 12 found [yes] 13 14 Introduction 15 Background/rationaleFor 2peer Explain the reviewscientific background and only rationale for the 3 16 investigation being reported [yes] 17 18 Objectives 3 State specific objectives, [yes] including any 3 19 prespecified hypotheses [no] 20 21 22 Methods 23 Study design 4 Present key elements of study design early in the 3 paper [yes] 24 25 Setting 5 Describe the setting, locations, and relevant dates, 3 26 including periods of recruitment, exposure, follow-up, 27 and data collection [yes] 28 29 Participants 6 (a) Give the eligibility criteria, and the sources and 3 30 methods of selection of participants [yes] 31 32 Variables 7 Clearly define all outcomes, exposures, predictors, 3
33 potential confounders, and effect modifiers. [yes] http://bmjopen.bmj.com/ 34 Give diagnostic criteria, if applicable [no] 35 36 Data sources/ 8* For each variable of interest, give sources of data 4 37 measurement and details of methods of assessment 38 (measurement). Describe comparability of 39 assessment methods if there is more than one group 40 [yes]
41 on September 28, 2021 by guest. Protected copyright. Bias 9 Describe any efforts to address potential sources of 4 42 bias [yes] 43 44 Study size 10 Explain how the study size was arrived at [yes] 4-5 45 46 Quantitative variables 11 Explain how quantitative variables were handled in 4-5 47 the analyses. If applicable, describe which groupings 48 were chosen and why [yes] 49 50 Statistical methods 12 (a) Describe all statistical methods, including those 5 51 used to control for confounding [yes] 52 53 (b) Describe any methods used to examine 5 54 subgroups and interactions [yes] 55 56 (c) Explain how missing data were addressed [yes] 5 57 58 (d) If applicable, describe analytical methods taking n.a. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 67 of 68 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 account of sampling strategy [no] 4 5 (e) Describe any sensitivity analyses [no] n.a. 6 7 Results 8 Participants 13* (a) Report numbers of individuals at each stage of 5 9 study—eg numbers potentially eligible, examined for 10 eligibility, confirmed eligible, included in the study, 11 completing follow-up, and analysed [no] 12 13 (b) Give reasons for non-participation at each stage 5 14 [yes] 15 For peer review only 16 (c) Consider use of a flow diagram [yes] 18 17 18 Descriptive data 14* (a) Give characteristics of study participants (eg 22-24 demographic, clinical, social) and information on 19 exposures and potential confounders [yes] 20 21 (b) Indicate number of participants with missing data 5 22 for each variable of interest [yes] 23 24 Outcome data 15* Report numbers of outcome events or summary 6 25 measures [yes] 26 27 Main results 16 (a) Give unadjusted estimates and, if applicable, 28 28 confounder-adjusted estimates and their precision 29 (eg, 95% confidence interval). Make clear which 30 confounders were adjusted for and why they were 31 included [yes] 32 b 33 ( ) Report category boundaries when continuous n.a. http://bmjopen.bmj.com/ 34 variables were categorized n.a. 35 c 36 ( ) If relevant, consider translating estimates of n.a. 37 relative risk into absolute risk for a meaningful time period [no] 38 39 Other analyses 17 Report other analyses done—eg analyses of 24 40 subgroups and interactions, and sensitivity analyses
41 [yes] on September 28, 2021 by guest. Protected copyright. 42 43 44 Discussion 45 Key results 18 Summarise key results with reference to study 8 46 objectives [yes] 47 Limitations 19 Discuss limitations of the study, taking into account 2 48 sources of potential bias or imprecision. Discuss both 49 direction and magnitude of any potential bias [yes] 50 51 Interpretation 20 Give a cautious overall interpretation of results 8-11 52 considering objectives, limitations, multiplicity of 53 analyses, results from similar studies, and other 54 relevant evidence [yes] 55 56 Generalisability 21 Discuss the generalisability (external validity) of the 6 57 study results [yes] 58 59 Other information 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 68 of 68 BMJ Open: first published as 10.1136/bmjopen-2016-013658 on 26 July 2017. Downloaded from 1 2 3 Funding 22 Give the source of funding and the role of the funders 13 4 for the present study and, if applicable, for the original 5 study on which the present article is based [yes] 6 7 8 *Give information separately for exposed and unexposed groups. 9 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological 10 background and published examples of transparent reporting. The STROBE checklist is best used in conjunction 11 12 with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of 13 Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the 14 STROBE Initiative is available at www.strobe-statement.org 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40
41 on September 28, 2021 by guest. Protected copyright. 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 Open Access Miscellaneous
Correction: Medication incidents in primary care medicine: a prospective study in the Swiss Sentinel Surveillance Network (Sentinella)
Gnädinger M, Conen D, Herzig L, et al. Medication incidents in primary care medicine: a prospective study in the Swiss Sentinel Surveillance Network (Sentinella). BMJ Open 2017;7:e013658. doi: 10.1136/bmjopen-2016-013658
The main affiliation of author Alessandro Cheschi was not included in the article. The revised affiliation list for this article is: Markus Gnädinger1, Dieter Conen2, Lilli Herzig3,4, Milo A Puhan5, Alfred Stae- helin1,4, Marco Zoller1, Alessandro Ceschi6,7,8 1. Institute of PrimaryCare, University of Zurich, Zurich, Switzerland 2. Swiss Patient Safety, Zurich, Switzerland 3. Institute of Family Medicine, University of Lausanne, Lausanne, Switzerland 4. Sentinel Surveillance Network, Swiss Federal Office of Public Health, Bern, Switzerland 5. Epidemiology,Biostatistics, and Prevention Institute, University of Zurich, Zurich, Switzerland 6. Division of Science, National Poisons Centre, Tox Info Suisse, Associated Institute of the University of Zurich, Zurich, Switzerland 7. Department of Clinical Pharmacology and Toxicology, University Hospital Zurich, Zurich, Switzerland 8. Division of Clinical Pharmacology and Toxicology, Institute of Pharmacological Sciences of Southern Switzerland, Ente Ospedaliero Cantonale, Lugano, Switzerland
Open Access This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/ © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2017. All rights reserved. No commercial use is permitted unless otherwise expressly granted.
BMJ Open 2017;7:e. doi:10.1136/bmjopen-2016-013658corr1
BMJ Open 2017;7:e. doi:10.1136/bmjopen-2016-013658corr1 1