Updated core competencies in pharmacoepidemiology to inform contemporary curricula and training for academia, government, and industry.

PURPOSE: The first paper to specify the core content of pharmacoepidemiology as a profession was published by an ISPE (International Society for Pharmacoepidemiology) workgroup in 2012 (Jones JK et al. PDS 2012; 21[7]:677-689). Due to the broader and evolving scope of pharmacoepidemiology, ISPE considers it important to proactively identify, update and expand the list of core competencies to inform curricula of education programs; thus, better positioning pharmacoepidemiologists across academic, government (including regulatory), and industry positions. The aim of this project was to update the list of core competencies in pharmacoepidemiology. METHODS: To ensure applicability of findings to multiple areas, a working group was established consisting of ISPE members with positions in academia, industry, government, and other settings. All competencies outlined by Jones et al. were extracted from the initial manuscript and presented to the working group for review. Expert-based judgments were collated and used to identify consensus. It was noted that some competencies could contribute to multiple groups and could be directly or indirectly related to a group. RESULTS: Five core domains were proposed: (1) Epidemiology, (2) Clinical Pharmacology, (3) Regulatory Science, (4) Statistics and data science, and (5) Communication and other professional skills. In total, 55 individual competencies were proposed, of which 25 were new competencies. No competencies from the original work were dropped but aggregation or amendments were made where considered necessary. CONCLUSIONS: While many core competencies in pharmacoepidemiology have remained the same over the past 10 years, there have also been several updates to reflect new and emerging concepts in the field.

Tree-temporal scan statistics for safety signal detection in vaccine clinical trials.

The evaluation of safety is critical in all clinical trials. However, the quantitative analysis of safety data in clinical trials poses statistical difficulties because of multiple potentially overlapping endpoints. Tree-temporal scan statistic approaches address this issue and have been widely employed in other data sources, but not to date in clinical trials. We evaluated the performance of three complementary scan statistical methods for routine quantitative safety signal detection: the self-controlled tree-temporal scan (SCTTS), a tree-temporal scan based on group comparison (BGTTS), and a log-rank based tree-temporal scan (LgRTTS). Each method was evaluated using data from two phase III clinical trials, and simulated data (simulation study). In the case study, the reference set was adverse events (AEs) in the Reference Safety Information of the evaluated vaccine. The SCTTS method had higher sensitivity than other methods, and after dose 1 detected 80 true positives (TP) with a positive predictive value (PPV) of 60%. The LgRTTS detected 49 TPs with 69% PPV. The BGTTS had 90% of PPV with 38 TPs. In the simulation study, with simulated reference sets of AEs, the SCTTS method had good sensitivity to detect transient effects. The LgRTTS method showed the best performance for the detection of persistent effects, with high sensitivity and expected probability of type I error. These three methods provide complementary approaches to safety signal detection in clinical trials or across clinical development programmes. All three methods formally adjust for multiple testing of large numbers of overlapping endpoints without being excessively conservative.

Methods for drug safety signal detection using routinely collected observational electronic health care data: A systematic review.

PURPOSE: Signal detection is a crucial step in the discovery of post-marketing adverse drug reactions. There is a growing interest in using routinely collected data to complement established spontaneous report analyses. This work aims to systematically review the methods for drug safety signal detection using routinely collected healthcare data and their performance, both in general and for specific types of drugs and outcomes. METHODS: We conducted a systematic review following the PRISMA guidelines, and registered a protocol in PROSPERO. MEDLINE, EMBASE, PubMed, Web of Science, Scopus, and the Cochrane Library were searched until July 13, 2021. RESULTS: The review included 101 articles, among which there were 39 methodological works, 25 performance assessment papers, and 24 observational studies. Methods included adaptations from those used with spontaneous reports, traditional epidemiological designs, methods specific to signal detection with real-world data. More recently, implementations of machine learning have been studied in the literature. Twenty-five studies evaluated method performances, 16 of them using the area under the curve (AUC) for a range of positive and negative controls as their main measure. Despite the likelihood that performance measurement could vary by drug-event pair, only 10 studies reported performance stratified by drugs and outcomes, in a heterogeneous manner. The replicability of the performance assessment results was limited due to lack of transparency in reporting and the lack of a gold standard reference set. CONCLUSIONS: A variety of methods have been described in the literature for signal detection with routinely collected data. No method showed superior performance in all papers and across all drugs and outcomes, performance assessment and reporting were heterogeneous. However, there is limited evidence that self-controlled designs, high dimensional propensity scores, and machine learning can achieve higher performances than other methods.

Visualizations throughout pharmacoepidemiology study planning, implementation, and reporting.

Transparency is increasingly promoted to instill trust in nonrandomized studies using real-world data. Graphics and data visualizations support transparency by aiding communication and understanding, and can inform study design and analysis decisions. However, other than graphical representation of a study design and flow diagrams (e.g., a Consolidated Standards of Reporting Trials [CONSORT] like diagram), specific standards on how to maximize validity and transparency with visualization are needed. This paper provides guidance on how to use visualizations throughout the life cycle of a pharmacoepidemiology study-from initial study design to final report-to facilitate rationalized and transparent decision-making about study design and implementation, and clear communication of study findings. Our intent is to help researchers align their practices with current consensus statements on transparency.

Modeling the factors that influence exposure to SARS-CoV-2 on a subway train carriage.

We propose the Transmission of Virus in Carriages (TVC) model, a computational model which simulates the potential exposure to SARS-CoV-2 for passengers traveling in a subway rail system train. This model considers exposure through three different routes: fomites via contact with contaminated surfaces; close-range exposure, which accounts for aerosol and droplet transmission within 2 m of the infectious source; and airborne exposure via small aerosols which does not rely on being within 2 m distance from the infectious source. Simulations are based on typical subway parameters and the aim of the study is to consider the relative effect of environmental and behavioral factors including prevalence of the virus in the population, number of people traveling, ventilation rate, and mask wearing as well as the effect of model assumptions such as emission rates. Results simulate generally low exposures in most of the scenarios considered, especially under low virus prevalence. Social distancing through reduced loading and high mask-wearing adherence is predicted to have a noticeable effect on reducing exposure through all routes. The highest predicted doses happen through close-range exposure, while the fomite route cannot be neglected; exposure through both routes relies on infrequent events involving relatively few individuals. Simulated exposure through the airborne route is more homogeneous across passengers, but is generally lower due to the typically short duration of the trips, mask wearing, and the high ventilation rate within the carriage. The infection risk resulting from exposure is challenging to estimate as it will be influenced by factors such as virus variant and vaccination rates.

Measuring the Effectiveness of Real-World Evidence to Ensure Appropriate Impact.

The value of real-world evidence (RWE) in medicines regulation and health technology assessment has been increasingly emphasized. Nevertheless, although RWE is increasingly used, there has been limited systematic evidence of its value. A recent study that examined the role and impact of RWE in regulatory assessments conducted through the European Medicines Agency provided such evidence. Results of the study demonstrated RWE was important to decision making, particularly for certain questions such as the quantification of adverse events, the evaluation of risk minimization measures, and the assessment of product usage. The study suggested, however, that in many of the assessments further RWE would have been valuable and concluded that RWE has, as yet, played a limited role in hypothesis generation and in the assessment of medication effectiveness. This study had been possible only because of the transparency of the European Medicines Agency decision making. Ensuring transparency of RWE evidence collection, study design and conduct, and of decision making based on this evidence will facilitate further development of the uses and value of RWE. Keywords: benefit-risk assessment; medicines regulation; real-world evidence; regulatory decision making.

Preytaxis and Travelling Waves in an Eco-epidemiological Model.

Preytaxis is the attraction (or repulsion) of predators along prey density gradients and a potentially important mechanism for predator movement. However, the impact preytaxis has on the spatial spread of a predator invasion or of an epidemic within the prey has not been investigated. We investigate the effects preytaxis has on the wavespeed of several different invasion scenarios in an eco-epidemiological system. In general, preytaxis cannot slow down predator or disease invasions and there are scenarios where preytaxis speeds up predator or disease invasions. For example, in the absence of disease, attractive preytaxis results in an increased wavespeed of predators invading prey, whereas repulsive preytaxis has no effect on the wavespeed, but the wavefront is shallower. On top of this, repulsive preytaxis can induce spatiotemporal oscillations and/or chaos behind the invasion front, phenomena normally only seen when the (non-spatial) coexistence steady state is unstable. In the presence of disease, the predator wave can have a different response to attractive susceptible and attractive infected prey. In particular, we found a case where attractive infected prey increases the predators' wavespeed by a disproportionately large amount compared to attractive susceptible prey since a predator invasion has a larger impact on the infected population. When we consider a disease invading a predator-prey steady state, we found some counter-intuitive results. For example, the epidemic has an increased wavespeed when infected prey attract predators. Likewise, repulsive susceptible prey can also increase the infection wave's wavespeed. These results suggest that preytaxis can have a major effect on the interactions of predators, prey and diseases.

Lessons from meta-analyses of randomized clinical trials for analysis of distributed networks of observational databases.

Networks of constellations of longitudinal observational databases, often electronic medical records or transactional insurance claims or both, are increasingly being used for studying the effects of medicinal products in real-world use. Such databases are frequently configured as distributed networks. That is, patient-level data are kept behind firewalls and not communicated outside of the data vendor other than in aggregate form. Instead, data are standardized across the network, and queries of the network are executed locally by data partners, and summary results provided to a central research partner(s) for amalgamation, aggregation, and summarization. Such networks can be huge covering years of data on upwards of 100 million patients. Examples of such networks include the FDA Sentinel Network, ASPEN, CNODES, and EU-ADR. As this is a new emerging field, we note in this paper the conceptual similarities and differences between the analysis of distributed networks and the now well-established field of meta-analysis of randomized clinical trials (RCTs). We recommend, wherever appropriate, to apply learnings from meta-analysis to help guide the development of distributed network analyses of longitudinal observational databases.

Do FDA label changes work? Assessment of the 2010 class label change for proton pump inhibitors using the Sentinel System's analytic tools.

PURPOSE: To pilot use of the US Food and Drug Administration's (FDA's) Sentinel System data and analytic tools by a non-FDA stakeholder through the Innovation in Medical Evidence Development and Surveillance system of the Reagan Udall Foundation. We evaluated the US FDA 2010 proton pump inhibitor (PPI) class label change that warned of increased risk of bone fracture, to use PPIs for the lowest dose and shortest duration, and to manage bone status for those at risk for osteoporosis. METHODS: The cohort consisted of adults aged 18 years or older prescribed PPIs without fracture risk factors. We evaluated incident and prevalent uses of the 8 PPIs noted in the label change. Outcomes evaluated before and after label change were PPI dispensing patterns, incident fractures, and osteoporosis screening or interventions. Consistent with FDA use of descriptive tools, we did not include direct comparisons or statistical testing. RESULTS: There were 1 488 869 and 2 224 420 incident PPI users in the before [PRE] and after [POST] periods, respectively. Users with 1 year or more of exposure decreased (8.4% vs 7.5%), as did mean days supplied/user (130.4 to 113.7 d among all users and 830.8 to 645.4 d among users with 1 y or more of exposure). Osteoporosis screening and interventions did not appear to increase, but the proportion of patients with fractures decreased (4.4% vs 3.1%). Prevalent user results were similar. CONCLUSIONS: This analysis demonstrated the ability to use Sentinel tools to assess the effectiveness of a label change and accompanying communication at the population level and suggests an influence on subsequent dispensing behavior.

Assessing performance of sequential analysis methods for active drug safety surveillance using observational data.

The routine use of sequential methods is well established in clinical studies. Recently, there has been increasing interest in applying these methods to prospectively monitor the safety of newly approved drugs through accrual of real-world data. However, the application to marketed drugs using real-world data has been limited and work is needed to determine which sequential approaches are most suited to such data. In this study, the conditional sequential sampling procedure (CSSP), a group sequential method, was compared with a log-linear model with Poisson distribution (LLMP) through a SAS procedure (PROC GENMOD) combined with an alpha-spending function on two large longitudinal US administrative health claims databases. Relative performance in identifying known drug-outcome associations was examined using a set of 50 well-studied drug-outcome pairs. The study finds that neither method correctly identified all pairs but that LLMP often provides better ability and shorter time for identifying the known drug-outcome associations with superior computational performance when compared with CSSP, albeit with more false positives. With the features of flexible confounding control and ease of implementation, LLMP may be a good alternative or complement to CSSP.

Evidence generation from healthcare databases: recommendations for managing change.

There is an increasing reliance on databases of healthcare records for pharmacoepidemiology and other medical research, and such resources are often accessed over a long period of time so it is vital to consider the impact of changes in data, access methodology and the environment. The authors discuss change in communication and management, and provide a checklist of issues to consider for both database providers and users. The scope of the paper is database research, and changes are considered in relation to the three main components of database research: the data content itself, how it is accessed, and the support and tools needed to use the database. Copyright © 2016 John Wiley & Sons, Ltd.

Risk of acute liver injury associated with use of antibiotics. Comparative cohort and nested case-control studies using two primary care databases in Europe.

PURPOSE: To assess the impact of varying study designs, exposure and outcome definitions on the risk of acute liver injury (ALI) associated with antibiotic use. METHODS: The source population comprised of patients registered in two primary care databases, in the UK and in Spain. We identified a cohort consisting of new users of antibiotics during the study period (2004-2009) and non-users during the study period or in the previous year. Cases with ALI were identified within this cohort and classified as definite or probable, based on recorded medical information. The relative risk (RR) of ALI associated with antibiotic use was computed using Poisson regression. For the nested case-control analyses, up to five controls were matched to each case by age, sex, date and practice (in CPRD) and odds ratios (OR) were computed with conditional logistic regression. RESULTS: The age, sex and year adjusted RRs of definite ALI in the current antibiotic use periods was 10.04 (95% CI: 6.97-14.47) in CPRD and 5.76 (95% CI: 3.46-9.59) in BIFAP. In the case-control analyses adjusting for life-style, comorbidities and use of medications, the OR of ALI for current users of antibiotics was and 5.7 (95% CI: 3.46-9.36) in CPRD and 2.6 (95% CI: 1.26-5.37) in BIFAP. CONCLUSION: Guided by a common protocol, both cohort and case-control study designs found an increased risk of ALI associated with the use of antibiotics in both databases, independent of the exposure and case definitions used. However, the magnitude of the risk was higher in CPRD compared to BIFAP.

Do case-only designs yield consistent results across design and different databases? A case study of hip fractures and benzodiazepines.

BACKGROUND: The case-crossover (CXO) and self-controlled case series (SCCS) designs are increasingly used in pharmacoepidemiology. In both, relative risk estimates are obtained within persons, implicitly controlling for time-fixed confounding variables. OBJECTIVES: To examine the consistency of relative risk estimates of hip/femur fractures (HFF) associated with the use of benzodiazepines (BZD) across case-only designs in two databases (DBs), when a common protocol was applied. METHODS: CXO and SCCS studies were conducted in BIFAP (Spain) and CPRD (UK). Exposure to BZD was divided into non-use, current, recent and past use. For CXO, odds ratios (OR; 95%CI) of current use versus non-use/past were estimated using conditional logistic regression adjusted for co-medications (AOR). For the SCCS, conditional Poisson regression was used to estimate incidence rate ratios (IRR; 95%CI) of current use versus non/past-use, adjusted for age. To investigate possible event-exposure dependence the relative risk in the 30 days prior to first BZD exposure was also evaluated. RESULTS: In the CXO current use of BZD was associated with an increased risk of HFF in both DBs, AORBIFAP = 1.47 (1.29-1.67) and AORCPRD = 1.55 (1.41-1.70). In the SCCS, IRRs for current exposure was 0.79 (0.72-0.86) in BIFAP and 1.21 (1.13-1.30) in CPRD. However, when we considered separately the 30-day pre-exposure period, the IRR for current period was 1.43 (1.31-1.57) in BIFAP and 1.37 (1.27-1.47) in CPRD. CONCLUSIONS: CXO designs yielded consistent results across DBs, while initial SCCS analyses did not. Accounting for event-exposure dependence, estimates derived from SCCS were more consistent across DBs and designs.

Prevalence of antibiotic use: a comparison across various European health care data sources.

PURPOSE: There is widespread concern about increases in antibiotic use, but comparative data from different European countries on rates of use are lacking. This study was designed to measure and understand the variation in antibiotic utilization across five European countries. METHODS: Seven European healthcare databases with access to primary care data from Denmark, Germany, the Netherlands, Spain and the UK were used to measure and compare the point and 1-year-period prevalence of antibiotic use between 2004 and 2009. Descriptive analyses were stratified by gender, age and type of antibiotic. Separate analyses were performed to measure the most common underlying indications leading to the prescription of an antibiotic. RESULTS: The average yearly period prevalence of antibiotic use varied from 15 (Netherlands) to 30 (Spain) users per 100 patients. A higher prevalence of antibiotic use by female patients, the very young (0-9 years) and old (80+ years), was observed in all databases. The lowest point prevalence was recorded in June and September and ranged from 0.51 (Netherlands) to 1.47 (UK) per 100 patients per day. Twelve percent (Netherlands) to forty-nine (Spain) percent of all users were diagnosed with a respiratory tract infection, and the most common type of antibiotic prescribed were penicillin. CONCLUSION: Using identical methodology in seven EU databases to assess antibiotic use allowed us to compare drug usage patterns across Europe. Our results contribute quantitatively to the true understanding of similarities and differences in the use of antibiotic agents in different EU countries.

Hip/femur fractures associated with the use of benzodiazepines (anxiolytics, hypnotics and related drugs): a methodological approach to assess consistencies across databases from the PROTECT-EU project.

BACKGROUND: Results from observational studies may be inconsistent because of variations in methodological and clinical factors that may be intrinsically related to the database (DB) where the study is performed. OBJECTIVES: The objectives of this paper were to evaluate the impact of applying a common study protocol to study benzodiazepines (BZDs) (anxiolytics, hypnotics, and related drugs) and the risk of hip/femur fracture (HFF) across three European primary care DBs and to investigate any resulting discrepancies. METHODS: To measure the risk of HFF among adult users of BZDs during 2001-2009, three cohort and nested case control (NCC) studies were performed in Base de datos para la Investigación Farmacoepidemiológica en Atención Primaria (BIFAP) (Spain), Clinical Practice Research Datalink (CPRD) (UK), and Mondriaan (The Netherlands). Four different models (A-D) with increasing levels of adjustment were analyzed. The risk according to duration and type of BZD was also explored. Adjusted hazard ratios (cohort), odds ratios (NCC), and their 95% confidence intervals were estimated. RESULTS: Adjusted hazard ratios (Model C) were 1.34 (1.23-1.47) in BIFAP, 1.66 (1.54-1.78) in CPRD, and 2.22 (1.55-3.29) in Mondriaan in cohort studies. Adjusted odds ratios (Model C) were 1.28 (1.16-1.42) in BIFAP, 1.60 (1.49-1.72) in CPRD, and 1.48 (0.89-2.48) in Mondriaan in NCC studies. A short-term effect was suggested in Mondriaan, but not in CPRD or BIFAP. All DBs showed an increased risk with the concomitant use of anxiolytic and hypnotic drugs. CONCLUSIONS: Applying similar study methods to different populations and DBs showed an increased risk of HFF in BZDs users but differed in the magnitude of the risk, which may be because of inherent differences between DBs.

Exposure to benzodiazepines (anxiolytics, hypnotics and related drugs) in seven European electronic healthcare databases: a cross-national descriptive study from the PROTECT-EU Project.

PURPOSE: Studies on drug utilization usually do not allow direct cross-national comparisons because of differences in the respective applied methods. This study aimed to compare time trends in BZDs prescribing by applying a common protocol and analyses plan in seven European electronic healthcare databases. METHODS: Crude and standardized prevalence rates of drug prescribing from 2001-2009 were calculated in databases from Spain, United Kingdon (UK), The Netherlands, Germany and Denmark. Prevalence was stratified by age, sex, BZD type [(using ATC codes), i.e. BZD-anxiolytics BZD-hypnotics, BZD-related drugs and clomethiazole], indication and number of prescription. RESULTS: Crude prevalence rates of BZDs prescribing ranged from 570 to 1700 per 10,000 person-years over the study period. Standardization by age and sex did not substantially change the differences. Standardized prevalence rates increased in the Spanish (+13%) and UK databases (+2% and +8%) over the study period, while they decreased in the Dutch databases (-4% and -22%), the German (-12%) and Danish (-26%) database. Prevalence of anxiolytics outweighed that of hypnotics in the Spanish, Dutch and Bavarian databases, but the reverse was shown in the UK and Danish databases. Prevalence rates consistently increased with age and were two-fold higher in women than in men in all databases. A median of 18% of users received 10 or more prescriptions in 2008. CONCLUSION: Although similar methods were applied, the prevalence of BZD prescribing varied considerably across different populations. Clinical factors related to BZDs and characteristics of the databases may explain these differences.

Modelling the impact and control of an infectious disease in a plant nursery with infected plant material inputs.

The ornamental plant trade has been identified as a key introduction pathway for plant pathogens. Establishing effective biosecurity measures to reduce the risk of plant pathogen outbreaks in the live plant trade is therefore important. Management of invasive pathogens has been identified as a weakest link public good, and thus is reliant on the actions of individual private agents. This paper therefore provides an analysis of the impact of the private agents' biosecurity decisions on pathogen prevention and control within the plant trade. We model the impact that an infectious disease has on a plant nursery under a constant pressure of potentially infected input plant materials, like seeds and saplings, where the spread of the disease reduces the value of mature plants. We explore six scenarios to understand the influence of three key bioeconomic parameters; the disease's basic reproductive number, the loss in value of a mature plant from acquiring an infection and the cost-effectiveness of restriction. The results characterise the disease dynamics within the nursery and explore the trade-offs and synergies between the optimal level of efforts on restriction strategies (actions to prevent buying infected inputs), and on removal of infected plants in the nursery. For diseases that can be easily controlled, restriction and removal are substitutable strategies. In contrast, for highly infectious diseases, restriction and removal are often found to be complementary, provided that restriction is cost-effective and the optimal level of removal is non-zero.

Implementation of a Taxonomy-Based Framework for the Selection of Appropriate Drugs and Outcomes for Real-World Data Signal Detection Studies.

INTRODUCTION: For signal detection studies investigating either drug safety or method evaluation, the choice of drug-outcome pairs needs to be tailored to the planned study design and vice versa. While this is well understood in hypothesis-testing epidemiology, it should be as important in signal detection, but this has not widely been considered. There is a need for a taxonomy framework to provide guidance and a systematic reproducible approach to the selection of appropriate drugs and outcomes for signal detection studies either investigating drug safety or assessing method performance using real-world data. OBJECTIVE: The aim was to design a general framework for the selection of appropriate drugs and outcomes for signal detection studies given a study design of interest. As a motivating example, we illustrate how the framework is applied to build a reference set for a study aiming to assess the performance of the self-controlled case series with active comparators. METHODS: We reviewed criteria presented in two published studies which aimed to provide practical advice for choosing the appropriate signal evaluation methodology, and assessed their relevance for signal detection. Further characteristics specific to signal detection were added. The final framework is based on: the application of study design requirements, the database(s) of interest, and the clinical importance of the drug(s) and outcome(s) under consideration. This structure was applied by selecting drug-outcome pairs as a reference set (i.e. list of drug-outcome pairs classified as positive or negative controls) for which the method is expected to work well for a signal detection study aiming to assess the performance of self-controlled case series. Eight criteria were used, related to the application of self-controlled case series assumptions, choice of active comparators, coverage in the database of interest and clinical importance of the outcomes. RESULTS: After application of the framework, two classes of antibiotics (seven drugs) were selected for the study, and 28 outcomes from all organ classes were chosen from the drug labels, out of the 273 investigated. In total, this corresponds to 104 positive controls (drug-outcome pairs) and 58 negative controls. CONCLUSIONS: We proposed and applied a framework for the selection of drugs and outcomes for both drug safety signal detection and method assessment used in signal detection to optimise their performance given a study design. This framework will eliminate part of the bias relating to drugs and outcomes not being suited to the method or database. The main difficulty lies in the choice of the criteria and their application to ensure systematic selection, especially as some information remains unknown in signal detection, and clinical judgement was needed on occasions. The same framework could be adapted for other methods.