Search

Jakob Kjølby Eika,1,2,* Kasper Bonnesen,1,2,* Lars Pedersen,1,2 Vera Ehrenstein,1,2 Henrik Toft Sørensen,1,2 Morten Schmidt1– 3 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Clinical Medicine, Aarhus University, Aarhus, Denmark; 3Department of Cardiology, Gødstrup Regional Hospital, Herning, Denmark*These authors contributed equally to this workCorrespondence: Morten Schmidt, Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark and Department of Clinical Medicine, Aarhus University, Olof Palmes Allé 43– 45, Aarhus, 8200, Denmark, Tel +45 87 16 72 12, Email morten.schmidt@clin.au.dkPurpose: Ibuprofen is used to treat acute pericarditis, but high-dose ibuprofen has also been associated with increased cardiovascular risks. We examined the cardiovascular safety of using ibuprofen for acute pericarditis.Patients and Methods: A Danish nationwide, population-based cohort study including patients ≥ 18 years with first-time acute pericarditis (n=12,381) during 1996– 2020 was conducted. Ibuprofen use was modelled in two ways: First, we considered patients exposed based on the tablet strength of their first ibuprofen filling (a proxy for an intention-to-treat analysis). Second, we considered patients exposed in a time-varying manner (a proxy for an as-treated analysis). The primary outcome of major adverse cardiovascular events (MACE) was a composite of myocardial infarction, ischemic stroke, congestive heart failure, and cardiovascular death.Results: In the intention-to-treat analysis, the 1-year risk of MACE was 1.37% (95% confidence interval [CI]: 1.03– 1.79) for ibuprofen initiators and 4.32% (95% CI: 3.89– 4.78) for non-initiators. Compared with non-initiators within 1-year follow-up, the adjusted hazard ratio for MACE was 0.75 (95% CI: 0.67– 0.85) for initiators overall, 0.38 (95% CI: 0.28– 0.52) for initiators of > 400 mg tablets, and 0.87 (95% CI: 0.76– 0.99) for initiators of ≤ 400 mg tablets. In the as-treated analysis, compared with no use, the hazard ratio associated with ibuprofen use was 0.69 (95% CI: 0.54– 0.89) for MACE, 0.82 (95% CI: 0.54– 1.26) for myocardial infarction, 0.74 (95% CI: 0.45– 1.22) for ischemic stroke, 0.67 (95% CI: 0.47– 0.96) for congestive heart failure, and 0.60 (95% CI: 0.31– 1.17) for cardiovascular death.Conclusion: Ibuprofen use for acute pericarditis was not associated with increased cardiovascular risks, supporting its safety in current practice.Keywords: Pericarditis, Anti-Inflammatory Agents, non-steroidal, Cohort Study, Epidemiology

Yuelian Sun,1– 4 Julie Werenberg Dreier,3,5 Chunsen Wu,6,7 Vera Ehrenstein,4 Jakob Christensen2– 4 1Department of Clinical Medicine, Aarhus University, Aarhus, Denmark; 2Department of Neurology, Aarhus University Hospital, Aarhus, Denmark; 3National Centre for Register-Based Research, Department of Economics and Business Economics, Aarhus University, Aarhus, Denmark; 4Department of Clinical Epidemiology, Department of Clinical Medicine, Aarhus University Hospital, Aarhus, Denmark; 5Centre for Integrated Register-based Research (CIRRAU), Aarhus University, Aarhus, Denmark; 6Department of Clinical Research, University of Southern Denmark, Odense, Denmark; 7Department of Gynaecology and Obstetrics, Odense University Hospital, Odense, DenmarkCorrespondence: Yuelian Sun, Department of Clinical Medicine, Aarhus University, Aarhus, Denmark, Email ys@clin.au.dkObjective: The Danish National Patient Registry (DNPR) is a valuable resource for medical and epidemiological Research. However, not all research articles fully described procedures they used to identify events. In this study, we compared two approaches in identifying persons with a disease diagnosis using neonatal jaundice and epilepsy as examples.Methods: A cohort of singletons born alive between the 1st January 1997 and the 30th November 2016 in Denmark was used for this purpose. Diagnostic information for a hospital contact in the registry included a primary diagnosis, secondary diagnoses, referral diagnoses, and additional information to a diagnosis (associated diagnoses), if any. Approach 1 identified patients of interest by considering all diagnostic information with exclusion of referral diagnoses only. Approach 2 identified patients of interest by additionally excluding diagnoses from a hospital contact that were coded with Z00 – Z99 of ICD-10 (for health service on examination and reproduction, etc.) as the main reason of the hospital contact. We presented the proportion of people with a diagnosis of neonatal jaundice and epilepsy by the two approaches and explored the potential explanations for the difference.Results: For the example of neonatal jaundice, the study population included N=1,186,683 persons. The proportion of children with a diagnosis of neonatal jaundice was 5.5% (n=66,736) by approach 1 and 3.9% (n=45,928) by approach 2. For the example of epilepsy, the study population included N=1,183,273 persons. The proportion of children with a diagnosis of epilepsy were 1.2% (n=14,604) by approach 1 and 0.9% (n=10,441) by approach 2.Discussion: This study demonstrated that the two approaches identified different proportion of persons with a diagnosis of neonatal jaundice and epilepsy. We advocated researchers report complete procedures of identifying patients for making research findings reproducible and comparable.Keywords: the Danish National Patient Registry, health administrative data, misclassification of diagnosis

Trine Toft Sørensen,1 Dóra Körmendiné Farkas,2 Emil Zâl Bjerregaard Riahi,2 Vera Ehrenstein,2 Victor W Henderson2– 4 1Department of Public Health, University of Copenhagen, Copenhagen, Denmark; 2Department of Clinical Epidemiology, Aarhus University, Aarhus, Denmark; 3Department of Epidemiology and Population Health, Stanford University, Stanford, CA, USA; 4Department of Neurology and Neurological Sciences, Stanford University, Stanford, CA, USACorrespondence: Trine Toft SørensenDepartment of Public Health, University of Copenhagen, Øster Farimagsgade 5, Copenhagen K 1014, DenmarkEmail skg654@ku.dkBackground: Some neurogenerative diseases have been linked to a reduced risk of cancer, but the association between motor neuron disease and cancer risk is not well understood. We hypothesized that cancer risk would be lower among those with motor neuron disease and its most common subtype, amyotrophic lateral sclerosis.Methods: We conducted a population-based cohort study of motor neuron disease and cancer risk using routinely collected data from population-based registries in Denmark. We examined cancer incidence among patients diagnosed with motor neuron disease between January 1980 and December 2013 followed through 2013. Using Danish national cancer rates for the study period, we computed standardized incidence ratios as a measure of relative risks.Results: In the cohort of 5053 patients with a motor neuron disease, the overall standardized incidence ratio of any cancer was 1.17 (95% confidence interval [CI], 1.03– 1.31); the corresponding standardized incidence ratio for amyotrophic lateral sclerosis was 1.24 (95% CI, 0.96– 1.57). The standardized incidence ratios of any cancer in the cohort with motor neuron disease was 1.52 (95% CI, 1.22– 1.87) for < 1 year of follow-up; 0.87 (95% CI, 0.68– 1.09) for years 1– 5 of follow-up; and 1.22 (95% CI, 1.01– 1.46) for > 5 years of follow-up. Beyond one year of follow-up, patients in the motor neuron disease had elevated standardized incidence ratios for lymphoid leukemia, non-Hodgkin lymphoma, and basal cell skin cancer.Conclusion: Findings fail to support the hypothesis that motor neuron disease or amyotrophic lateral sclerosis is associated with reduced cancer incidence. An elevated risk of cancer during the first year of follow-up may be attributable to heightened surveillance.Keywords: amyotrophic lateral sclerosis, cancer, cohort, motor neuron disease, neurodegeneration, risk

Eric I Benchimol,1– 3 David Moher,4,5 Vera Ehrenstein,6 Sinéad M Langan7 1Department of Paediatrics, University of Toronto, Toronto, Canada; 2Child Health Evaluative Sciences, SickKids Research Institute, and Division of Gastroenterology, Hepatology and Nutrition, The Hospital for Sick Children, Toronto, Canada; 3ICES, Toronto, Canada; 4School of Epidemiology and Public Health, University of Ottawa, Ottawa, Canada; 5Centre for Journalology, Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, Canada; 6Department of Clinical Epidemiology, Aarhus University, Aarhus, Denmark; 7Faculty of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, London, UKCorrespondence: Eric I BenchimolDivision of Gastroenterology, Hepatology and Nutrition, The Hospital for Sick Children, 555 University Avenue, Toronto, ON M5G 1X8, CanadaTel +1 416813-1500Fax +1 416813-4972Email eric.benchimol@sickkids.caIntroduction: Two recent high-profile publications (and subsequent retractions) of pharmacoepidemiology studies reporting the effectiveness and risk of hydroxychloroquine in COVID-19 patients received international media attention. Transparent and complete reporting of these studies could have provided peer reviewers and editors with sufficient information to question the methods used and the validity of results. Since these studies used routinely collected health data, the guidelines for the REporting of studies Conducted using Observational Routinely collected health Data (RECORD) should have been applied to ensure complete reporting of the research.Methods: We evaluated the two retracted articles for completeness of reporting using the RECORD for Pharmacoepidemiology (RECORD-PE) checklist, which includes the checklists for the STengthening the Reporting of OBservational studies in Epidemiology (STROBE) and RECORD. We compared the proportion of STROBE, RECORD and RECORD-PE items adequately reported using Chi-squared statistics.Results: In the article published by The Lancet, 29 of 34 STROBE items (85.3%) were adequately reported, compared with 3.5 of 13 RECORD items (26.9%) and 9.5 of 15 RECORD-PE items (63.3%)(χ2 = 14.839, P < 0.001). Similarly, the article published in NEJM reported 24 of 34 STROBE items (70.6%), two of 13 RECORD items (15.4%), and 7.5 of 15 RECORD-PE items (50.0%) (χ2 = 11.668, P = 0.003). Important aspects of the methods unique to research using routinely collected health data were not reported, including variables used to identify exposure, outcome and confounders, validation of the coding or algorithms, a description of the underlying database population and the accuracy of data linkage methods.Discussion: While STROBE items were generally adequately reported, RECORD and RECORD-PE items were not. Reporting guidelines should be effectively implemented in order for transparency and completeness of research manuscripts, allowing for adequate evaluation by editors and peer reviewers.Keywords: COVID-19, journalology, peer review, pharmacoepidemiology, reporting guidelines, routinely collected health data

Morten Schmidt1,2, Sigrun Alba Johannesdottir Schmidt1,3, Kasper Adelborg1,4, Jens Sundbøll,1 Kristina Laugesen,1 Vera Ehrenstein,1 Henrik Toft Sørensen11Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Cardiology, Regional Hospital West Jutland, Herning, Denmark; 3Department of Dermatology, Aarhus University Hospital, Aarhus, Denmark; 4Department of Clinical Biochemistry, Aarhus University Hospital, Aarhus, DenmarkAbstract: Denmark has a large network of population-based medical databases, which routinely collect high-quality data as a by-product of health care provision. The Danish medical databases include administrative, health, and clinical quality databases. Understanding the full research potential of these data sources requires insight into the underlying health care system. This review describes key elements of the Danish health care system from planning and delivery to record generation. First, it presents the history of the health care system, its overall organization and financing. Second, it details delivery of primary, hospital, psychiatric, and elderly care. Third, the path from a health care contact to a database record is followed. Finally, an overview of the available data sources is presented. This review discusses the data quality of each type of medical database and describes the relative technical ease and cost-effectiveness of exact individual-level linkage among them. It is shown, from an epidemiological point of view, how Denmark’s population represents an open dynamic cohort with complete long-term follow-up, censored only at emigration or death. It is concluded that Denmark’s constellation of universal health care, long-standing routine registration of most health and life events, and the possibility of exact individual-level data linkage provides unlimited possibilities for epidemiological research.Keywords: health care sector, political systems, population health, registries, epidemiology

Jennifer L Lund,1,2 Erzsébet Horváth-Puhó,2 Szimonetta Komjáthiné Szépligeti,2 Henrik Toft Sørensen,2 Lars Pedersen,2 Vera Ehrenstein,2 Til Stürmer1,2 1Department of Epidemiology, University of North Carolina, Chapel Hill, NC, USA; 2Department of Clinical Epidemiology, Aarhus University, Aarhus, Denmark Background: A principle of cohort studies is that cohort membership is defined by current rather than future exposure information. Pharmacoepidemiologic studies using existing databases are vulnerable to violation of this principle. We evaluated the impact of using data on future redemption of prescriptions to determine cohort membership, motivated by a published example seeking to emulate a “per-protocol” association between continuous versus never use of low-dose acetylsalicylic acid (ASA) and major bleeding (e.g., cerebral hemorrhage or gastrointestinal bleeding). Materials and methods: Danish medical registry data from 2004 to 2011 were used to construct two analytic cohorts. In Cohort 1, we used information about future redemption of low-dose ASA prescriptions to identify cohorts of continuous and never-ASA users. In Cohort 2, we identified ASA initiators and non-initiators using only contemporaneous data and censored follow-up for changes in use over time. We implemented propensity score-matched Poisson regression to evaluate associations between ASA use and major bleeding and estimated adjusted incidence rate differences (IRDs) per 1,000 person-years and ratios (IRRs) overall and stratified by time since initiation. Results: Among >6 million eligible Danish adults, we identified 403,693 low-dose ASA initiators (Cohort 2), of whom 189,150 were defined as continuous users (Cohort 1). Overall, IRDs and IRRs were similar across cohorts. However, the IRD for major bleeding in the first 90 days was substantially larger in Cohort 1 (IRD=25 per 1,000 person-years) compared with Cohort 2 (IRD=10 per 1,000 person-years). Conclusion: Using future medication redemption data to define baseline cohorts violates basic epidemiologic principles. Compared with an approach using only contemporaneous data to define cohorts, the approach based on future redemption data generated a substantially higher short-term association between low-dose ASA use and major bleeding on the absolute, but not the relative, scale possibly due to selection and immortal time biases. Keywords: pharmacoepidemiology, cohort studies, immortal time bias, selection bias

Troels Munch,1 Lotte B Christensen,1 Kasper Adelborg,1 Grethe S Tell,2 Ellen M Apalset,2,3 Anna Westerlund,3,4 Ylva T Lagerros,5,6 Johnny Kahlert,1 Fei Xue,7 Vera Ehrenstein1 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway; 3Department of Rheumatology, Bergen Group of Epidemiology and Biomarkers in Rheumatic Disease, Haukeland University Hospital, Bergen, Norway; 4Department of Medicine Solna, Centre for Pharmacoepidemiology, Karolinska Institutet, Stockholm, Sweden; 5Department of Medicine Solna, Clinical Epidemiology Unit, Karolinska Institutet, Stockholm, Sweden; 6Department of Medicine, Clinic of Endocrinology, Metabolism and Diabetes, Karolinska University Hospital, Sweden; 7Center for Observational Research, Amgen Inc., Thousand Oaks, CA, USA Background: Validation of definitions used to identify conditions of interest is imperative to epidemiologic studies based on routinely collected data. The objective of the study was thus to estimate positive predictive values (PPVs) of International Classification of Diseases, 10th Revision (ICD-10) codes to identify cases of incident acute pancreatitis leading to hospitalization and incident primary malignancy in the Scandinavian (Denmark, Norway, and Sweden) national patient registries in women with postmenopausal osteoporosis (PMO). Methods: This validation study included postmenopausal (defined as 55 years or older) women with osteoporosis, identified between 2005–2014. Potential cases were sampled based on ICD-10 codes from the three national patient registries. Cases were adjudicated by physicians, using medical record review as gold standard. PPVs with corresponding 95% CIs were computed. Results: Medical records of 286 of 325 (retrieval rate 88%) women with PMO were available for adjudication. Acute pancreatitis leading to hospitalization had a PPV of 87.6% (95% CI: 80.8%–90.2%). Incident primary malignancy had a PPV of 88.1% (95% CI: 81.3%–92.7%). The PPVs did not vary substantially across the three countries. Conclusion: ICD-10 codes to identify acute pancreatitis leading to hospitalization, and incident primary malignancy in the Scandinavian national patient registries had high PPVs among women with PMO. This allows identification of cases of acute pancreatitis and incident primary malignancy with reasonable validity and to use these as outcomes in comparative analyses. Keywords: validation, primary malignancy, acute pancreatitis, positive predictive value, postmenopausal women

Vera Ehrenstein, Henrik Nielsen, Alma B Pedersen, Søren P Johnsen, Lars Pedersen Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus N, Denmark Abstract: Routinely recorded health data have evolved from mere by-products of health care delivery or billing into a powerful research tool for studying and improving patient care through clinical epidemiologic research. Big data in the context of epidemiologic research means large interlinkable data sets within a single country or networks of multinational databases. Several Nordic, European, and other multinational collaborations are now well established. Advantages of big data for clinical epidemiology include improved precision of estimates, which is especially important for reassuring (“null”) findings; ability to conduct meaningful analyses in subgroup of patients; and rapid detection of safety signals. Big data will also provide new possibilities for research by enabling access to linked information from biobanks, electronic medical records, patient-reported outcome measures, automatic and semiautomatic electronic monitoring devices, and social media. The sheer amount of data, however, does not eliminate and may even amplify systematic error. Therefore, methodologies addressing systematic error, clinical knowledge, and underlying hypotheses are more important than ever to ensure that the signal is discernable behind the noise. Keywords: electronic health records, healthcare administrative claims, medical record linkage, multicenter studies, validation studies

Kasper Adelborg,1 Lotte Brix Christensen,1 Troels Munch,1 Johnny Kahlert,1 Ylva Trolle Lagerros,2,3 Grethe S Tell,4 Ellen M Apalset,4,5 Fei Xue,6 Vera Ehrenstein1 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus N, Denmark; 2Department of Medicine, Clinical Epidemiology Unit, Karolinska Institutet, 3Department of Medicine, Clinic of Endocrinology, Metabolism and Diabetes, Karolinska University Hospital, Stockholm, Sweden; 4Department of Global Public Health and Primary Care, University of Bergen, 5Department of Rheumatology, Haukeland University Hospital, Bergen, Norway; 6Center for Observational Research, Amgen Inc. Thousand Oaks, CA, USA Background: Clinical epidemiology research studies, including pharmacoepidemiology and pharmacovigilance studies, use routinely collected health data, such as diagnoses recorded in national health and administrative registries, to assess clinical effectiveness and safety of treatments. We estimated positive predictive values (PPVs) of International Classification of Diseases, 10th revision (ICD-10) codes for primary diagnoses of dermatologic events and hypersensitivity recorded at hospitalization or emergency room visit in the national patient registries of Denmark and Sweden among women with postmenopausal osteoporosis (PMO). Methods: This validation study included women with PMO identified from the Danish and Swedish national patient registries (2005–2014). Medical charts of the potential cases served as the gold standard for the diagnosis confirmation and were reviewed and adjudicated by physicians. Results: We obtained and reviewed 189 of 221 sampled medical records (86%). The overall PPV was 92.4% (95% confidence interval [CI], 85.1%–96.3%) for dermatologic events, while the PPVs for bullous events and erythematous dermatologic events were 52.5% (95% CI, 37.5%–67.1%) and 12.5% (95% CI, 2.2%–47.1%), respectively. The PPV was 59.0% (95% CI, 48.3%–69.0%) for hypersensitivity; however, the PPV of hypersensitivity increased to 100.0% (95% CI, 67.6%–100.0%) when restricting to diagnostic codes for anaphylaxis. The overall results did not vary by country. Conclusion: Among women with PMO, the PPV for any dermatologic event recorded as the primary diagnosis at hospitalization or at an emergency room visit was high and acceptable for epidemiologic research in the Danish and Swedish national patient registries. The PPV was substantially lower for hypersensitivity leading to hospitalization or emergency room visit. Keywords: dermatology, positive predictive value, validation, pharmacoepidemiology

Mette Nørgaard,1 Vera Ehrenstein,1 Jan P Vandenbroucke1–3 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Clinical Epidemiology, Leiden University Medical Center, The Netherlands; 3Department of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, London, United Kingdom Abstract: Population-based health care databases are a valuable tool for observational studies as they reflect daily medical practice for large and representative populations. A constant challenge in observational designs is, however, to rule out confounding, and the value of these databases for a given study question accordingly depends on completeness and validity of the information on confounding factors. In this article, we describe the types of potential confounding factors typically lacking in large health care databases and suggest strategies for confounding control when data on important confounders are unavailable. Using Danish health care databases as examples, we present the use of proxy measures for important confounders and the use of external adjustment. We also briefly discuss the potential value of active comparators, high-dimensional propensity scores, self-controlled designs, pseudorandomization, and the use of positive or negative controls. Keywords: observational studies, health care databases, confounding

John Acquavella,1 Vera Ehrenstein,1 Morten Schiødt,2 Uffe Heide-Jørgensen,1 Anders Kjellman,3,4 Svein Hansen,5 Cecilia Larsson Wexell,6,7 Bente Brokstad Herlofson,8 Sven Erik Noerholt,9 Haijun Ma,10 Katarina Öhrling,11 Rohini K Hernandez,12 Henrik Toft Sørensen1 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Oral and Maxillofacial Surgery, Rigshospitalet, Copenhagen, Denmark; 3Department of Urology, Karolinska University Hospital, 4Department of Clinical Science, Intervention, and Technology, Karolinska Institutet, Stockholm, Sweden; 5Cancer Registry of Norway, Oslo University Hospital, Oslo, Norway; 6Department of Oral and Maxillofacial Surgery, Södra Älvsborg Hospital, Borås, Sweden; 7Department of Biomaterials, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 8Department of Oral Surgery and Oral Medicine, Faculty of Dentistry, University of Oslo, Oslo, Norway; 9Department of Oral and Maxillofacial Surgery, Aarhus University Hospital, Aarhus, Denmark; 10Global Biostatistical Science, Amgen Inc., Thousand Oaks, CA, USA; 11Clinical Development, Amgen Inc., Thousand Oaks, CA, USA; 12Center for Observational Research, Amgen Inc., Thousand Oaks, CA, USA Objective: Osteonecrosis of the jaw (ONJ) is a recognized complication of potent antiresorptive therapies, especially at the doses indicated to prevent skeletal complications for cancer patients with bone metastases. This paper describes the rationale and methods for a prospective, post-authorization safety study of cancer patients treated with antiresorptive therapies. Methods: As part of a comprehensive pharmacovigilance plan, developed with regulators’ input, the study will estimate incidence of ONJ and of serious infections among adult cancer patients with bone metastases treated with denosumab (120 mg subcutaneously) or zoledronic acid (4 mg intravenously, adjusted for renal function). Patients will be identified using routinely collected data combined with medical chart review in Denmark, Sweden, and Norway. Follow-up will extend from the first administration of antiresorptive treatment to the earliest of death, loss-to-follow-up, or 5 years after therapy initiation. Results will be reported for three treatment cohorts: denosumab-naïve patients, zoledronic acid-naïve patients, and patients who switch from bisphosphonate treatment to denosumab. ONJ cases will be identified in three newly established national ONJ databases and adjudicated by the committee that functioned during the XGEVA® clinical trials program. Conclusion: This study will provide a real world counterpart to the clinical trial-estimated risks for ONJ and serious infections for cancer patients initiating denosumab or zoledronic acid. The establishment of ONJ databases in the three Scandinavian countries will have potential benefits outside this study for the elucidation of ONJ risk factors and the evaluation of ONJ treatment strategies. Keywords: cohort study, osteonecrosis of the jaw, pharmacovigilance, postmarketing drug surveillance, denosumab, zoledronic acid

Anne Gulbech Ording,1 Deirdre Cronin-Fenton,1 Vera Ehrenstein,1 Timothy L Lash,1,2 John Acquavella,1 Mikael Rørth,1 Henrik Toft Sørensen1 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Epidemiology, Rollins School of Public Health, Emory University, Atlanta, GA, USA Abstract: Clinical trials are considered the gold standard for examining drug efficacy and for approval of new drugs. Medical databases and population surveillance registries are valuable resources for post-approval observational research, which are increasingly used in studies of benefits and risk of new cancer drugs. Here, we address the challenges in translating endpoints from oncology trials to observational studies. Registry-based cohort studies can investigate real-world safety issues – including previously unrecognized concerns – by examining rare endpoints or multiple endpoints at once. In contrast to clinical trials, observational cohort studies typically do not exclude real-world patients from clinical practice, such as old and frail patients with comorbidity. The observational cohort study complements the clinical trial by examining the effectiveness of interventions applied in clinical practice and by providing evidence on long-term clinical outcomes, which are often not feasible to study in a clinical trial. Various endpoints can be included in clinical trials, such as hard endpoints, soft endpoints, surrogate endpoints, and patient-reported endpoints. Each endpoint has it strengths and limitations for use in research studies. Endpoints used in oncology trials are often not applicable in observational cohort studies which are limited by the setting of standard clinical practice and by non-standardized endpoint determination. Observational studies can be more helpful moving research forward if they restrict focus to appropriate and valid endpoints. Keywords: endpoint determination, medical oncology, treatment outcome, neoplasms, research design

Morten Schmidt,1 Sigrun Alba Johannesdottir Schmidt,1 Jakob Lynge Sandegaard,2 Vera Ehrenstein,1 Lars Pedersen,1 Henrik Toft Sørensen11Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, 2Department of Health Documentation, State Serum Institute, Copenhagen, DenmarkBackground: The Danish National Patient Registry (DNPR) is one of the world's oldest nationwide hospital registries and is used extensively for research. Many studies have validated algorithms for identifying health events in the DNPR, but the reports are fragmented and no overview exists.Objectives: To review the content, data quality, and research potential of the DNPR.Methods: We examined the setting, history, aims, content, and classification systems of the DNPR. We searched PubMed and the Danish Medical Journal to create a bibliography of validation studies. We included also studies that were referenced in retrieved papers or known to us beforehand. Methodological considerations related to DNPR data were reviewed.Results: During 1977–2012, the DNPR registered 8,085,603 persons, accounting for 7,268,857 inpatient, 5,953,405 outpatient, and 5,097,300 emergency department contacts. The DNPR provides nationwide longitudinal registration of detailed administrative and clinical data. It has recorded information on all patients discharged from Danish nonpsychiatric hospitals since 1977 and on psychiatric inpatients and emergency department and outpatient specialty clinic contacts since 1995. For each patient contact, one primary and optional secondary diagnoses are recorded according to the International Classification of Diseases. The DNPR provides a data source to identify diseases, examinations, certain in-hospital medical treatments, and surgical procedures. Long-term temporal trends in hospitalization and treatment rates can be studied. The positive predictive values of diseases and treatments vary widely (

Vera Ehrenstein,1 Rohini K Hernandez,2 Merete Lund Maegbaek,1 Johnny Kahlert,1 Mary Nguyen-Nielsen,1 Mette Nørgaard,1 Alexander Liede2 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Center for Observational Research, Amgen, Thousand Oaks, CA, USA Objective: Among patients with prostate cancer, diagnostic codes for bone metastases in the Danish National Registry of Patients have a sensitivity of 44%. In an attempt to improve the sensitivity of registry-based identification of metastases from prostate cancer, we tested a series of algorithms, combining elevated prostate-specific antigen (PSA) levels, use of antiresorptive therapy, and performed bone scintigraphy. Patients and methods: We randomly selected 212 men diagnosed with prostate cancer in 2005–2010 in the Central Denmark Region with prespecified PSA values, antiresorptive therapy, and bone scintigraphy who did not have a registry-based diagnostic code indicating presence of distant metastases. We defined three candidate algorithms for bone metastases: 1) PSA >50 µg/L and bone scintigraphy, 2) PSA >50 µg/L and antiresorptive therapy, and 3) PSA ≤50 µg/L with antiresorptive therapy or bone scintigraphy. An algorithm for distant metastasis site other than bone was defined as PSA >50 µg/L alone. Medical chart review was used as the reference standard to establish the presence or absence of metastases. Validity was expressed as a positive predictive value (PPV) or a negative predictive value, based on whether the algorithms correctly classified metastases compared with the reference standard. Results: We identified 113 men with evidence of metastases according to the candidate algorithms, and 99 men without evidence of metastases according to the candidate algorithm. The PPVs of PSA >50 µg/L were 0.10 (95% confidence interval [CI] 0.04–0.19) for bone metastases and 0.14 (95% CI 0.07–0.24) for nonbone metastases, regardless of receipt of antiresorptive therapy or presence of bone scintigraphy. The PPVs for any metastases were 0.16 (95% CI 0.06–0.32) for PSA >50 µg/L and 0.28 (95% CI 0.14–0.47) for PSA >50 µg/L with bone scintigraphy. Adding antiresorptive treatment to the algorithm did not improve PPV. All negative predictive values approached 1.00. Conclusion: Algorithms based on elevated PSA, antiresorptive therapy, or bone scintigraphy are not suitable for supplementing diagnostic codes to identify additional cases of distant metastases among men with prostate cancer. However, it is possible that in this setting, medical chart review is not a gold standard to identify metastases. Keywords: algorithm, bone metastases, prostate-specific antigen, validation, Danish databases

Sigrun Alba Johannesdottir Schmidt,1 Uffe Heide-Jørgensen,1 Angelika D Manthripragada,2 Vera Ehrenstein1 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Center for Observational Research, Amgen Inc., Thousand Oaks, CA, USA Background: With the emergence of new lipid-lowering therapies, more patients are expected to achieve substantial lowering of low-density lipoprotein cholesterol (LDL-C). However, there are limited data examining the clinical experience of patients with low (3.3 mmol/L) LDL-C. We described their demographic characteristics, entire comorbidity history, and 90-day prescription history prior to the lowest LDL-C value measured. Finally, we further restricted the analysis to individuals with very low LDL-C (

Morten Schiodt,1 Cecilia Larsson Wexell,2,3 Bente Brokstad Herlofson,4 Karen Marie Giltvedt,4 Sven Erik Norholt,5 Vera Ehrenstein61Department of Oral and Maxillofacial Surgery, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark; 2Department of Oral and Maxillofacial Surgery, Södra Älvsborg Hospital, Region Västra Götaland, Borås, 3Department of Biomaterials, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 4Department of Oral Surgery and Oral Medicine, Faculty of Dentistry, University of Oslo, Oslo, Norway; 5Department of Oral and Maxillofacial Surgery, Aarhus University and University Hospital, 6Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, DenmarkAbstract: Osteonecrosis of the jaw (ONJ) is a severe side effect associated with antiresorptive treatment. Monitoring of ONJ using routine databases in Scandinavian countries is a challenge owing to lack of valid algorithms and to heterogeneous referral practices. The aim of this paper is to describe the process of establishing a Scandinavian ONJ Cohort enrolling all ONJ cases related to antiresorptive treatment arising in Denmark, Norway, and Sweden between 2011 and 2019. The initial purpose of the cohort is to support an ongoing pharmacovigilance study of denosumab and zoledronic acid in Denmark, Norway, and Sweden. The three countries, with their 199 clinics, departments, and units of oral and maxillofacial surgery, both hospital-based and freestanding, differ somewhat in referral practices of the ONJ patients. By directly contacting all providers of care to ONJ patients in the three countries, we established a network for reporting incident cases to each country's research database directly or through a member of the Scandinavian ONJ task force as a liaison. The task force includes a Scandinavian coordinator and three national coordinators collaborating directly with the clinics. A uniform ONJ registration form has been developed, and the relevant medical community has been informed either directly or through presentations at professional meetings. A website with study information is published in each country, and data entry is ongoing. This large-scale systematic uniform registration of ONJ cases in Denmark, Norway, and Sweden, with an underlying total population of more than 20 million people, merged into the Scandinavian ONJ Cohort, will contribute to better knowledge and understanding of this challenging group of patients, and ultimately, help improve patient care. The Scandinavian ONJ Cohort as a whole and its component national ONJ research databases may offer the potential for large-scale multinational intervention and safety studies in the future.Keywords: antiresorptive, bisphosphonate, cancer, classification, denosumab, epidemiology, osteonecrosis, jaws, ONJ, osteoporosis, Scandinavia

Henrik Toft Sørensen,1 Lars Pedersen,1 Jørgen Jørgensen,2 Vera Ehrenstein11Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, 2Danish Regions, Copenhagen, DenmarkThe health care systems in many countries are facing several challenges: the aging population,the need for personalized medicine, evolving treatment modalities, quality andsafety imperatives, and unsustainable costs, to mention only a few. Population aging in the face of significant pressure to contain costs is perhaps the most immediate challenge. The proportion of people aged 65 years or older in Western Europe and North America is expected to increase to 26% in 2025. Furthermore, clinical medicine in the Western world is confronting an evolving set of diseases, as smoking becomes less prevalent and obesity more common. Diagnostics and treatment of chronic disease have improved, while the threshold for initiating preventive treatment of asymptomatic conditions has been lowered. Consequently, the number of patients with multimorbidity, that is, the coexistence of several chronic diseases, will increase dramatically.1 To treat a disease or prevent its progression, patients with several chronic diseases often take multiple drugs, each with potentially severe side effects. In patients with multiple morbidities, “polypharmacy” is a challenging clinical issue, often associated with iatrogenic harm.2 A call for innovative approaches to polypharmacy has been the focus of recent editorials in high-impact medical journals.3–5 Randomized trials rarely address multimorbidity, adherence to treatments, co-intervention (polypharmacy), or their long-term risks.6 These challenges underscore the need for population-based long-term longitudinal clinical data available for clinical care and research.

Vera Ehrenstein,1 Irene Petersen,1,2 Liam Smeeth,3 Susan S Jick,4 Eric I Benchimol,5,6 Jonas F Ludvigsson,7,8 Henrik Toft Sørensen11Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Primary Care and Population Health, University College London, London, UK; 3Department of Non-Communicable Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK; 4Boston Collaborative Drug Surveillance Program, Boston University School of Public Health, Boston, MA, USA; 5Department of Pediatrics and School of Epidemiology, Public Health and Preventive Medicine, Faculty of Medicine, University of Ottawa, Ottawa, ON, Canada; 6Institute for Clinical Evaluative Sciences, Toronto, ON, Canada; 7Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, 8Department of Pediatrics, University Hospital of Örebro, SwedenThere has been a surge of availability and use for research of routinely collected electronic health data, such as electronic health records, health administrative data, and disease registries. Symptomatic of this surge, in 2012, Pharmacoepidemiology and Drug Safety (PDS) published a supplemental issue containing several reviews of validated methods for identifying health outcomes using routine health data,1 focusing on databases feeding the US Mini-Sentinel Program.2

Ane Birgitte Telén Andersen,1 Rune Erichsen,1 Michael David Kappelman,2 Trine Frøslev,1 Vera Ehrenstein1 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Pediatrics, Division of Pediatric Gastroenterology, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Objective: The incidences of celiac disease (CD) and asthma are increasing and the two conditions are associated in individuals. Risk of asthma may be passed on to the next generation through shared risk factors. We examined whether parental CD is associated with risk of asthma in offspring. Methods: We conducted a population-based Danish nationwide cohort study, using medical databases, covering the period 1 January 1979 to 31 December 2009. For each child with a parental history of CD, we randomly sampled 100 children without this history from the children born in the same calendar year. We used Cox proportional-hazards regression to estimate incidence rate ratios for asthma, adjusting for measured covariates. Results: We identified 1,107 children with a parental history of CD and 110,700 children without this parental history. During up to 32 years of follow-up, 6,125 children received a hospital diagnosis of asthma. The adjusted incidence rate ratio for asthma associated with a parental history of CD was 1.09 (95% confidence interval: 0.86–1.39) and was similar for maternal and paternal CD. Inclusion of asthma-medication in the definition of asthma did not substantially change the results. Conclusion: There was no convincing evidence of an increased risk of asthma among offspring of parents with CD. Keywords: asthma, celiac disease, children, cohort study, parents

Aska Drljevic,1 Michael Borre,2 Morten Høyer,3 Vera Ehrenstein,4 Mary Nguyen-Nielsen4 1Department of Pharmacology, 2Department of Urology, 3Department of Oncology, 4Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark Background and aim: It is well established that cancer patients have an increased risk of venous thromboembolism (VTE). However, no previous study has examined the quality of VTE diagnoses related to cancer patients in the Danish National Registry of Patients (DNRP). To support future studies on cancer and risk of VTE, this study aimed to estimate the positive predictive value (PPV) of VTE diagnoses among prostate cancer (PC) patients registered in the DNRP. Materials and methods: We conducted a validation study using data from hospitals within the Central Denmark Region, which covers a population of 1.3 million people. Using the DNRP, we identified a total of 120 PC patients registered with VTE within the period 1995–2012. We also identified a random sample of 120 PC patients with no VTE registration within the same period. Therefore, a total of 240 patients were selected for medical chart review. We compared data from the DNRP to data collected from medical record review (ie, reference standard). We then computed PPV, sensitivity, and specificity with corresponding 95% confidence intervals (CIs) using the Jeffreys method. Results: The final study sample included 232 PC patients, of which 115 were registered with VTE and 117 had no registration of VTE in the DNRP. We found the overall PPV of VTE diagnoses in the DNRP to be 86.1% (95% CI 78.9%–91.5%). Sensitivity was 98.0% (95% CI 93.8%–99.6%), and specificity was 87.8% (95% CI 81.4%–92.6%). We also found the PPV of incident PC diagnoses in the DNRP to be 98.3% (95% CI 96.1%–99.4%). Conclusion: For PC patients, the registration of VTE diagnoses in the DNRP is associated with a high PPV. We provide evidence that data from the DNRP are valid for studies on risk of VTE among cancer patients. Keywords: positive predictive value, prostatic neoplasms, validation study, venous thromboembolism

Ane Birgitte Telén Andersen, Vera Ehrenstein, Rune Erichsen, Trine Frøslev, Henrik Toft Sørensen Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark Purpose: Inflammatory bowel disease (IBD) and autism spectrum disorders (ASD) may share genetic and environmental risk factors. We examined whether parental IBD is associated with an increased risk of ASD in offspring. Methods: We conducted a registry-based nationwide cohort study including children born alive in Denmark from January 1, 1994 to December 31, 2009, with follow-up throughout 2010. IBD in parents and ASD in offspring were identified using inpatient and outpatient hospital diagnoses. We computed risk of ASD and crude and adjusted incidence rate ratios (aIRRs) with 95% confidence intervals (CI) using Cox proportional-hazards regression. We evaluated the risk of ASD according to maternal and paternal IBD, and separately for maternal and paternal Crohn's disease (CD) and ulcerative colitis (UC). Children with parents free of IBD were the comparison cohort. Results: We identified 1,005,330 children during the study period. Among them, 11,888 (1.2%) had a parent with IBD and 8,087 (0.8%) had a diagnosis of ASD during up to 17 years of follow-up. The 10-year risks of ASD were 0.7% among children of parents with IBD and 0.9% among children of parents without IBD. The aIRR for ASD among children with parental IBD was 0.8 (95% CI: 0.6–1.0), and results were similar regardless of parent of IBD origin or whether a parent had CD or UC. The estimates were similar for different ASD subtypes. Conclusion: We found no evidence of an increased risk of ASD among children born to parents with IBD. Keywords: autistic disorder, children, cohort study, epidemiology, inflammatory bowel disease, parents

Henrik Gammelager,1 Claus Sværke,1 Sven Erik Noerholt,2 Bjarne Neumann-Jensen,3 Fei Xue,4 Cathy Critchlow,4 Johan Bergdahl,5 Ylva Trolle Lagerros,5 Helle Kieler,5 Grethe S Tell,6 Vera Ehrenstein11Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Department of Oral and Maxillofacial Surgery, Aarhus University Hospital, Aarhus, Denmark; 3Department of Oral and Maxillofacial Surgery, Aalborg University Hospital, Aalborg, Denmark; 4Center for Observational Research, Amgen Inc., Thousand Oaks, CA, USA; 5Center for Pharmacoepidemiology, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden; 6Department of Global Public Health and Primary Care, University of Bergen, Bergen, NorwayBackground: Osteonecrosis of the jaw (ONJ) is an adverse effect of drugs that suppress bone turnover – for example, drugs used for the treatment of postmenopausal osteoporosis. The Danish National Registry of Patients (DNRP) is potentially valuable for monitoring ONJ and its prognosis; however, no specific code for ONJ exists in the International Classification of Diseases 10th revision (ICD-10), which is currently used in Denmark. Our aim was to estimate the positive predictive value (PPV) of an algorithm to capture ONJ cases in the DNRP among women with postmenopausal osteoporosis.Methods: We conducted this cross-sectional validation study in the Central and North Denmark Regions, with approximately 1.8 million inhabitants. In total, 54,956 women with postmenopausal osteoporosis were identified from June 1, 2005 through May 31, 2010. To identify women potentially suffering from ONJ, we applied an algorithm based on ICD-10 codes in the DNRP originating from hospital-based departments of oral and maxillofacial surgery (DOMS). ONJ was adjudicated by chart review and defined by the presence of exposed maxillofacial bone for 8 weeks or more, in the absence of recorded history of craniofacial radiation therapy. We estimated the PPV for the overall algorithm and for each separate ICD-10 code used in the algorithm.Results: Charts were obtained and reviewed for all 60 women with an ICD-10 code potentially representing ONJ. Nineteen potential ONJ cases were confirmed, corresponding to an overall PPV of 32% (95% confidence interval: 20%–45%).Conclusion: Among women with postmenopausal osteoporosis, only about one-third of the potential ONJ cases identified by our ICD-10 based algorithm were confirmed by medical chart review, despite the restriction to patients treated at DOMS. To capture true ONJ cases among women with postmenopausal osteoporosis, alternative approaches are needed.Keywords: bisphosphonate-associated osteonecrosis of the jaw, International Classification of Diseases, osteonecrosis of the jaw, osteoporosis, registries, validity

Anne-Mette B Bjørn,1 Rikke B Nielsen,1 Mette Nørgaard,1 Ellen A Nohr,2 Vera Ehrenstein11Department of Clinical Epidemiology, 2Section for Epidemiology, Department of Public Health, Aarhus University Hospital, Aarhus, DenmarkBackground: The purpose of this nested case-control study in Denmark was to study the association between use of corticosteroids and risk of miscarriage.Methods: We identified prescriptions for corticosteroids before the miscarriage/index date. We estimated odds ratios (ORs) for miscarriage and for early (

Mary Nguyen-Nielsen,1 Elisabeth Svensson,1 Ida Vogel,2 Vera Ehrenstein,1 Lone Sunde21Department of Clinical Epidemiology, Aarhus University Hospital, Denmark; 2Department of Clinical Genetics, Aarhus University Hospital, DenmarkAbstract: Denmark has an extensive collection of national and regional medical registries. There are many advantages to registry-based research when investigating genetic diseases which, due to their rarity, can be difficult to identify. In this study, we aimed to provide an updated overview of Danish registries for medical genetic conditions and describe how data linkage across registries can be used to collect data on genetic diseases at the individual level and at the family level. We present a list of medical genetic registries in Denmark at the national level, data sources from the departments of clinical genetics and other specialized centers, and project-specific data sources. We also summarize key general registries, such as the Danish National Registry of Patients, the Danish Medical Birth Registry, and the Civil Registration System, which are renowned for their comprehensive and high quality data, and are useful supplemental data sources for genetic epidemiology research. We describe the potential for data linkage across multiple registries, which allows for access to medical histories with follow-up time spanning birth to death. Finally, we provide a brief introduction to the Danish epidemiological research setting and legalities related to data access. The Danish collection of medical registries is a valuable resource for genetic epidemiology research.Keywords: epidemiology, registry-based research, genetic disorders

Johan Bergdahl,1 Fredrik Jarnbring,2 Vera Ehrenstein,3 Henrik Gammelager,3 Fredrik Granath,1 Helle Kieler,1 Madeleine Svensson,1 Grethe S Tell,4 Ylva Trolle Lagerros11Centre for Pharmacoepidemiology, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden; 2Department of Oral and Maxillofacial Surgery Solna, Karolinska University Hospital, Stockholm, Sweden; 3Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 4Department of Public Health and Primary Health Care, University of Bergen, Bergen, NorwayBackground: Osteonecrosis of the jaw (ONJ) is a medical condition associated with antiresorptive drugs, among others, used to treat osteoporosis and bone metastasis. Currently, there is no consensus regarding the definition of ONJ, and no ONJ-specific International Classification of Diseases-10 code exists. Therefore, register-based studies of this condition may be troublesome.Purpose: To evaluate an algorithm ascertaining ONJ cases in an attempt to facilitate future assessments of ONJ in clinical and epidemiological studies.Methods: By means of the Patient Register and the Prescribed Drug Register, we identified all postmenopausal female residents in Sweden from 2005 through 2009. To identify potential cases of ONJ, we employed an algorithm including the following conditions: periapical abscess with sinus, inflammatory conditions of jaws, alveolitis of jaws, idiopathic aseptic necrosis of bone, osteonecrosis due to drugs, osteonecrosis due to previous trauma, other secondary osteonecrosis, other osteonecrosis, and unspecified osteonecrosis. Women seen at departments of oral and maxillofacial surgery, with at least one of the conditions, were classified as potential cases of ONJ. Conditions in anatomic sites other than the jaw were excluded. Validation was performed through medical record review. Case confirmation was based on the ONJ definition by the American Association of Oral and Maxillofacial Surgeons. The algorithm was evaluated by positive predictive values (PPVs) stratified by diagnosis.Results: For the 87 potential cases identified through our algorithm, the medical records were obtained for 83. The overall PPV was 18% (95% confidence interval (CI) 10%–28%). The highest PPV was observed in osteonecrosis due to drugs (83%, 95% CI 36%–100%). Several diagnoses had a PPV of 0 or were not used at all (periapical abscess with sinus, alveolitis of jaws, idiopathic aseptic necrosis of bone, osteonecrosis due to previous trauma, other secondary osteonecrosis, other osteonecrosis, and unspecified osteonecrosis).Conclusion: It was possible to ascertain cases of ONJ from the Swedish registers using this algorithm; however, the PPV was low. Thus, further refinements of the algorithm are necessary.Keywords: bisphosphonate-associated osteonecrosis of the jaw, epidemiology, methods, ONJ, registries, validation

Sigrun Alba Johannesdottir, Erzsébet Horváth-Puhó, Vera Ehrenstein, Morten Schmidt, Lars Pedersen, Henrik Toft SørensenDepartment of Clinical Epidemiology, Institute of Clinical Medicine, Aarhus University Hospital, Aarhus, DenmarkAbstract: The Danish health care system provides partial reimbursement of most prescription medications in Denmark. The dispensation of prescription medications is registered in administrative databases. Each time a prescription is redeemed at a pharmacy, an electronic record is generated with information related to the user, prescriber, the pharmacy, and the dispensed drug. The National Health Service gathers this information for administration of the drug reimbursement plan. Recently, this information became the basis for the establishment of a new research database, the Danish National Database of Reimbursed Prescriptions (DNDRP). In this paper, we review the content, coverage, quality, linkage, access, and research possibilities of this new database. The database encompasses the reimbursement records of all reimbursed drugs sold in community pharmacies and hospital-based outpatient pharmacies in Denmark since 2004. On average, approximately 3.5 million users are recorded in the database each year. During the coverage period, the number of annual prescription redemptions increased by 15%. Most dispensed prescriptions are in the categories “alimentary tract and metabolism”, “cardiovascular system”, “nervous system”, and “respiratory system”. Individuals are identified by the unique central personal registration (CPR) number assigned to all persons born in or immigrating to Denmark. The new database fully complies with Denmark’s Act on Processing of Personal Data, while avoiding additional restrictions imposed on data use at the Danish National Prescription Registry, administered by Statistics Denmark. Most importantly, CPR numbers are reversibly encrypted, which allows re-identification of drug users; furthermore, the data access is possible outside the servers of Statistics Denmark. These features open additional opportunities for international collaboration, validation studies, studies on adverse drug effects requiring review of medical records, studies involving contact to general practitioners, and linkage of prescription data to other clinical and research databases. The DNDRP thus is a valuable data source for pharmacoepidemiological research.Keywords: Denmark, patient registration, pharmacoepidemiology, registry-based research

Malene Schou Nielsson,1 Rune Erichsen,1 Trine Frøslev,1 Aliki Taylor,2 John Acquavella,2 Vera Ehrenstein11Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; 2Center for Observational Research International, Amgen Ltd Thousand Oaks, CA, USA Background: The purpose of this study was to estimate the positive predictive value (PPV) of the coding for bisphosphonate treatment in selected cancer patients from the Danish National Patient Registry (DNPR).Methods: Through the DNPR, we identified all patients with recorded cancer of the breast, prostate, lung, kidney, and with multiple myeloma. We restricted the study sample to patients with bisphosphonate treatment recorded during an admission to Aalborg Hospital, Denmark, from 2005 through 2009. We retrieved and reviewed medical records of these patients from the initial cancer diagnosis onwards to confirm or rule out bisphosphonate therapy. We calculated the PPV of the treatment coding as the proportion of patients with confirmed bisphosphonate treatment.Results: We retrieved and reviewed the medical records of 60 cancer patients with treatment codes corresponding to bisphosphonate therapy. Recorded code corresponded to treatment administered intravenously for 59 of 60 patients, corresponding to a PPV of 98.3% (95% confidence interval 92.5–99.8). In the remaining patient, bisphosphonate treatment was also confirmed but was an orally administered bisphosphonate; thus, the treatment for any bisphosphonate regardless of administration was confirmed for all 60 patients (PPV of 100%, 95% confidence interval 95.9–100.0).Conclusion: The PPV of bisphosphonate treatment coding among cancer patients in the DNPR is very high and the recorded treatment nearly always corresponds to intravenous administration.Keywords: bisphosphonate, neoplasm metastases, predictive value of tests, validation studies

Ane Birgitte Telén Andersen, Dóra Körmendiné Farkas, Frank Mehnert, Vera Ehrenstein, Rune ErichsenDepartment of Clinical Epidemiology, Aarhus University Hospital, DenmarkPurpose: Use of paracetamol during pregnancy may increase the risk of asthma in offspring. The association between prenatal exposure to maternal use of paracetamol and risk of asthma was investigated.Methods: A cohort study of 197,060 singletons born in northern Denmark in 1996–2008 was conducted, with follow-up until the end of 2009. Maternal paracetamol use during pregnancy was defined as a redeemed prescription. Asthma in offspring was defined as at least two prescriptions of both a β-agonist and an inhaled glucocorticoid and/or a hospital diagnosis of asthma during follow-up. Absolute risk of asthma in offspring was estimated using the Kaplan–Meier method and incidence rate ratios adjusted for known risk factors were estimated using Cox proportional-hazards regression.Results: Overall, 976 (0.5%) children were exposed prenatally to maternal use of prescription paracetamol. During follow-up, 24,506 (12.4%) children developed asthma. Absolute risk of asthma was 7.5% after 2 years and 14.4% after 10 years among the unexposed children. Corresponding risks were 12.7% and 21.6% among the exposed children. The adjusted incidence rate ratio was 1.35 (95% confidence interval: 1.17–1.57) for exposure in any trimester of pregnancy. A similar association was present for paracetamol exposure in each of the trimesters and for maternal use of prescription nonsteroidal anti-inflammatory drugs. Furthermore, maternal prescription use in the year following the relevant delivery also showed similar associations.Conclusion: A robust association was found between prenatal exposure to maternal use of prescription paracetamol and the risk of asthma; however, noncausal explanations could not be ruled out for such association.Keywords: acetaminophen, asthma, children, paracetamol, pregnancy

Sigrun Alba Johannesdottir, Merete Lund Mægbæk, Jens Georg Hansen, Timothy L Lash, Lars Pedersen, Vera EhrensteinDepartment of Clinical Epidemiology, Institute of Clinical Medicine, Aarhus University Hospital, Aarhus, DenmarkAbstract: Epidemiologic studies often rely on drug dispensation records to measure medication intake. We aimed to estimate correspondence between general practitioner (GP)-reported treatment and timing of prescription dispensation. From seven GPs in northern Denmark, we obtained 317 prescription records for 286 patients treated with ten commonly prescribed medication types for chronic diseases. We linked the GP-reported information to the regional prescription database to retrieve patients’ prescription records both prospectively and retrospectively in relation to the GP-reported date of treatment (index date, August 20, 2008 for all patients). We computed overall and medication-specific correspondence between GP-reported treatment and the timing of dispensation. We computed correspondence based on both exact medication and therapeutic subgroup agreement. The correspondence for dispensation within ±90 days of GP-reported treatment was 0.81 (95% confidence interval = 0.76–0.85) with variation by medication type, ranging from 0.55 for ACE-inhibitors to 1.00 for oral glucose-lowering agents. The correspondence was greater when analyzed within therapeutic groups than when analyzed for exact medications within these groups.Keywords: pharmacoepidemiology, predictive value, drug prescriptions, primary health care

Anne-Mette Bay Bjørn1, Mette Nørgaard1, Heidi Holmager Hundborg1, Ellen Aagaard Nohr2, Vera Ehrenstein11Department of Clinical Epidemiology, Aarhus University Hospital, 2Department of Epidemiology, Institute of Public Health, University of Aarhus, DenmarkPurpose: To describe patterns of prescribed drug use over time among primiparous women in Denmark.Methods: Through the Danish Medical Birth Registry, we identified all primiparous women giving live birth or stillbirth at ≥ 22 gestational weeks in northern Denmark, from 1999 to 2009. From the Aarhus University Prescription Database we obtained information on the women's prescriptions for reimbursed drugs filled from 30 days before conception until delivery.Results: Among 85,710 primiparous women, 47,982 (56.0%) redeemed at least one prescription from 30 days before conception until delivery. Women aged 35 years and older had the highest overall prevalence of prescription drug use (61.1%). Age-standardized prevalence of drug use was 54.7% in 1999 and 61.2% in 2009, prevalence ratio (PR) of 1.13 (95% confidence interval 1.10; 1.16), adjusted for age and smoking.Conclusion: Over the 11-year period from 1999 to 2009, we found a modest increase in overall use of drugs by primiparous women in Denmark. This increase was not, however, explained by an increasing proportion of older first-time mothers. We noted changes in patterns of use of anti-infective drugs and antidepressants.Keywords: drug utilization, epidemiology, pregnancy