Pharmacoepidemiology resources in Ireland-an introduction to pharmacy claims data

Drug-Drug Interactions and Their Association with Adverse Health Outcomes in the Older Community-Dwelling Population: A Prospective Cohort Study

BACKGROUND

Evidence on associations between drug-drug interactions (DDIs) and health outcomes in the older community-dwelling population is limited.

OBJECTIVE

We estimate potentially clinically important DDI prevalence and examine the association between DDIs and (1) adverse drug events (ADEs), (2) emergency hospital attendance and (3) health-related quality of life (HRQoL) in an older community-dwelling population in Ireland.

METHODS

This is a prospective cohort study of community-dwelling older adults (N = 904) aged ≥ 70 years from 15 general practices in Ireland recruited in 2010 (wave-1) and followed-up over 2 years (wave-2; 2012-2013), with linked national pharmacy claims data. Individuals dispensed two or more drugs (wave-1: N = 842; wave-2: N = 763) were included. DDI prevalence at baseline, follow-up and 6 months prior to each health outcome was estimated. Multi-level regression was used to model the association between DDI-exposure and health outcomes at follow-up. DDI prevalence, adjusted incidence-rate ratios (aIRR), adjusted odds ratios (aOR), β coefficients and robust standard error (RSE) from multi-level regression analyses, and 95% confidence intervals (CIs) are reported.

RESULTS

At wave-1, n = 196 (23.3% [95% CI 20.5-26.3]), individuals were potentially exposed to ≥ 1 DDI, increasing to n = 345 (45.2% [41.7-48.9]) at wave-2. At 2-year follow-up, the median number of ADEs was 3 (interquartile range [IQR 2-5]); 229 (30.1%) had ≥ 1 emergency hospital attendance, and the mean EQ-5D was 0.74 (± 0.23). Evidence for the association between DDI-exposure and emergency hospital attendance at follow-up was lacking (aOR = 1.38 [0.42-4.53]). DDI-exposure was associated with an increasing number of ADEs (aIRR = 1.26 [1.03-1.55]), and decreasing EQ-5D utility (β = - 0.07, [-0.11 to -0.04], RSE = 0.02). Aspirin-warfarin, clarithromycin-prednisolone, amiodarone-furosemide, clarithromycin-salbutamol, rosuvastatin-warfarin, amiodarone-bisoprolol, and aspirin-nicorandil were common DDIs 6 months preceding these health outcomes.

CONCLUSIONS

We found a two-fold increase in DDI prevalence between wave 1 and 2. DDI exposure was associated with increasing ADEs and declining HRQoL at 2-year follow-up. Common DDIs involved anticoagulants, cardiovascular and antimicrobial drugs, which should be targeted for medicine optimisation.

Social deprivation and diabetic kidney disease: A European view

There is a large body of literature demonstrating a social gradient in health and increasing evidence of an association between social deprivation and diabetes complications. Diabetic kidney disease (DKD) increases mortality in people with diabetes. Socioeconomic deprivation is increasingly recognized as a modifier of risk factors for kidney disease but also an independent risk factor itself for kidney disease. This may not be truly appreciated by clinicians and warrants further attention and exploration. In this review we explore the literature to date from Europe on the relationship between social deprivation and DKD. The majority of the studies showed at least an association with microalbuminuria, an early marker of DKD, while many showed an association with overt nephropathy. This was seen across many countries in Europe using a variety of different measures of deprivation. We reviewed and considered the mechanisms by which deprivation may lead to DKD. Health related behaviors such as smoking and suboptimal control of risk factors such as hypertension, hyperglycemia and elevated body mass index (BMI) accounts for some but not all of the association. Poorer access to healthcare, health literacy, and stress are also discussed as potential mediators of the association. Addressing deprivation is difficult but starting points include targeted interventions for people living in deprived circumstances, equitable roll out of diabetes technology, and flexible outpatient clinic arrangements including virtual and community-based care.

Eligibility rates and representativeness of the General Medical Services scheme population in Ireland 2016-2021: A methodological report

Background

In Ireland, the means tested General Medical Services (GMS) scheme provides access to a range of healthcare services at no or low cost to approximately one third of the population. Individuals eligible for the GMS scheme are often a focus of research, as a population that account for a large proportion of healthcare services use. The aim of this study is to describe the eligibility rates and representativeness of the GMS scheme population over time, with respect to age group, sex, and geographical area in Ireland.

Methods

Population data was obtained from the Central Statistics Office (CSO), using 2016 Census figures and projected population figures for 2017-2021. GMS eligibility figures for 2016-2021 were obtained from the HSE Primary Care Reimbursement Service (PCRS). GMS eligibility rates and relative rates of eligibility were calculated for 2016-2021 by age group and sex. Additionally, 2016 eligibility rates were calculated by geographical area.

Results

The crude eligibility rate decreased from 36.4% in 2016 to 31.2% in 2020, with a slight increase to 31.6% in 2021. In the 75+ years age group, 78.2% of the total population were eligible for the GMS scheme in 2021. The age group with the lowest rate of eligible individuals was the 25-34 age group, with 19.5% eligible in 2021. The eligibility rate was higher among females compared to males throughout the study period. The highest eligibility rate was seen in Donegal, with a crude rate of 52.8%. Dublin had the lowest rate, with a crude rate of 29.3%.

Conclusions

GMS eligibility varies greatly depending on age, sex, and geographical area, and decreased between 2016 and 2021. This study uses the most up-to-date data available to provide age group, sex and area-based figures for GMS eligibility which may inform planning and conduct of research focusing on GMS-eligible individuals.

Exploring patient medication adherence and data mining methods in clinical big data: A contemporary review

BACKGROUND

Increasingly, patient medication adherence data are being consolidated from claims databases and electronic health records (EHRs). Such databases offer an indirect avenue to gauge medication adherence in our data-rich healthcare milieu. The surge in data accessibility, coupled with the pressing need for its conversion to actionable insights, has spotlighted data mining, with machine learning (ML) emerging as a pivotal technique. Nonadherence poses heightened health risks and escalates medical costs. This paper elucidates the synergistic interaction between medical database mining for medication adherence and the role of ML in fostering knowledge discovery.

METHODS

We conducted a comprehensive review of EHR applications in the realm of medication adherence, leveraging ML techniques. We expounded on the evolution and structure of medical databases pertinent to medication adherence and harnessed both supervised and unsupervised ML paradigms to delve into adherence and its ramifications.

RESULTS

Our study underscores the applications of medical databases and ML, encompassing both supervised and unsupervised learning, for medication adherence in clinical big data. Databases like SEER and NHANES, often underutilized due to their intricacies, have gained prominence. Employing ML to excavate patient medication logs from these databases facilitates adherence analysis. Such findings are pivotal for clinical decision-making, risk stratification, and scholarly pursuits, aiming to elevate healthcare quality.

CONCLUSION

Advanced data mining in the era of big data has revolutionized medication adherence research, thereby enhancing patient care. Emphasizing bespoke interventions and research could herald transformative shifts in therapeutic modalities.

Prescription drugs with potential for misuse: protocol for a multi-indicator analysis of supply, detection and the associated health burden in Ireland between 2010 and 2020

INTRODUCTION

There is an increasing concern about the misuse of prescription drugs. Misuse refers to the intentional repurposing of prescribed drugs and/or the use of illicitly sourced prescription drugs, which may be counterfeit or contaminated. Drugs with the greatest potential for misuse are prescription opioids, gabapentinoids, benzodiazepines, Z-drugs and stimulants.

OBJECTIVE

The aim of this study is to provide a comprehensive analysis of the supply, patterns of use and health burden associated with prescription drugs with potential for misuse (PDPM) in Ireland between 2010 and 2020. Three inter-related studies will be carried out. The first study will describe trends in supply of PDPM using law enforcement drug seizures data and national prescription records from national community and prison settings. The second study aims to estimate trends in the detection of PDPM across multiple early warning systems using national forensic toxicology data. The third study aims to quantify the health burden associated with PDPM nationally, using epidemiological indicators of drug-poisoning deaths, non-fatal intentional drug overdose presentations to hospitals and drug treatment demand.

METHODS AND ANALYSIS

A retrospective observational study design, with repeated cross-sectional analyses, using negative binomial regression models or, where appropriate, joinpoint regression.

ETHICS AND DISSEMINATION

The study has received approval from the RCSI Ethics Committee (REC202202020). Results will be disseminated in peer-reviewed journals, scientific and drug policy meetings and with key stakeholders via research briefs.

Adverse drug reactions in an ageing PopulaTion (ADAPT) study: Prevalence and risk factors associated with adverse drug reaction-related hospital admissions in older patients

Background: Older people experience greater morbidity with a corresponding increase in medication use resulting in a potentially higher risk of adverse drug reactions (ADRs). Objectives: The aim of this study was to; 1) determine the prevalence and characteristics of ADR-related hospital admissions among older patients (≥65 years) in Ireland; and 2) identify the risk factors associated with ADR-related hospital admissions. Methods: A cross-sectional study of ADR prevalence in patients aged ≥65 years admitted acutely to hospital in Ireland over a 8 month period (November 2016- June 2017). A multifaceted review of each hospital admission was undertaken to assess the likelihood of an ADR being a reason for admission (cause of admission or contributing to admission) in the context of the patient's medication, clinical conditions, comorbidities and investigations. A number of decision aids were applied by two independent reviewers to assess ADR causality, avoidability and severity. A random sample of patients, determined not to have a suspected ADR on screening, were assigned to a non-ADR control group. Multivariable logistic regression was used to assess the association between potential risk factors for ADR-related admissions compared with non-ADR-related admissions. Results: In total, 3,760 hospital admission episodes (in 3,091 patients) were screened and 377 admissions were considered ADR-related (10.0%, 95% CI 9.1%, 11.0%). 219 (58.1%) ADR-related admissions were caused by an ADR, while ADRs contributed to 158 (41.9%) admissions. 268 (71.1%) of all ADR-related admissions were deemed definitely or possibly preventable/avoidable. 350 (92.8%) ADRs were classified as being of moderate severity, with 27 (7.2%) classified as severe. Antithrombotic agents, mainly aspirin and warfarin, were the drugs most frequently associated with ADR-related admissions (gastrointestinal and vascular haemorrhagic disorders). In multivariable analysis, immobility, frailty, having delirium or ulcer disease and taking anticoagulant and antiplatelet medication on admission were significantly associated with an ADR-related hospital admission. Conclusion: One in ten hospital admissions, among those aged 65 + years, were considered ADR-related, with approximately 70% potentially avoidable. Reliable and validated ADR detection and prediction tools are needed to develop prevention strategies.

Examining Quality, Use and Impact of Psychotropic (Use) in older adults with intellectual disabilities (EQUIP): study protocol

Widespread, and sometimes inappropriate use of psychotropics in adults with intellectual disability has been an international concern. These medicines have been used to treat mental health conditions, but also, controversially, some types of behaviours not necessarily associated with the diagnosis or in the absence of a relevant diagnosis. Results from the Intellectual Disability Supplement to the Irish Longitudinal Study on Ageing (IDS-TILDA) study of older adults with intellectual disability in Ireland revealed that 60% were taking psychotropics in 2010. In the intervening decade changes in regulations, policy, and increased decongregation of people with intellectual disability have taken place likely influencing the use of psychotropics. The HSE National Clinical Programme for People with Disability (NCPDD) established in the 2020 has medicines optimisation as a key priority. Existing multi-wave data from the IDS-TILDA study and the HSE national prescribing database offers an opportunity to better understand psychotropic use and prescribing patterns. This is a novel collaboration on lived experience, research, practice and policy. The aim of this research is to examine the quality and trends of psychotropic use of older adults with intellectual disability over a ten-year period in Ireland to evaluate the effects of and to inform both practice and policy to optimise medicines use and health outcomes. Health and medication data from ten years (four waves) of the IDS-TILDA study and corresponding medicines data from the HSE-PCRS prescribing database are available. Descriptive and longitudinal analysis will examine association between long-term psychotropic use, changes in trends of use, and the impact of decongregation on medicines use. This research will inform the development of national guidance on medicines optimisation for older people with intellectual disability and has the potential to change prescribing practices and improve health and wellbeing for older people with intellectual disability.

Valproate utilisation trends among women of child‐bearing potential in Ireland between 2014 and 2019: a drug utilisation study using interrupted time series

Purpose

This study aimed to examine trends in valproate use among women of childbearing potential (WCBP) aged 16–44 years in Ireland following two European‐directed regulatory interventions in December 2014 and April 2018.

Methods

This was a repeated cross‐sectional study using monthly national pharmacy claims data, to examine trend changes in the prevalence of valproate use among WCBP pre and post two separate regulatory events in December 2014 and April 2018. Annual population estimates from the Central Statistics Office were used to calculate the prevalence rate per 1000 eligible women. Segmented regression analysis of interrupted time series with negative binomial regression was used to examine rates for WCBP aged 16–44 years, and by 10‐year age groups. Prevalence ratios (PR) are presented with 95% confidence intervals (CIs).

Results

Among WCBP aged 16–44 years, there was no statistically significant change in the month‐to‐month prevalence ratio in the post‐ compared to pre‐December 2014 intervention period. A significant decline was, however, observed in the post‐, compared to pre‐April 2018 intervention period (PR = 0.998, [95% CIs: 0.996, 1.000]; p = 0.029). Among those aged 16–24 years, a significant decreasing trend in the month‐to‐month prevalence ratio was found in the post‐ compared to pre‐December 2014 intervention period (PR = 0.991, [95% CIs: 0.984, 0.998];p <0.01). A marginal effect was observed in the post‐ compared to pre‐April 2018 intervention period for those aged 25–34 years (PR = 0.996, [95% CIs: 0.992, 1.000]; p = 0.048).

Conclusion

Although no evidence of change was observed following the December 2014 intervention period, a significant decline in the prevalence ratio of valproate use was observed after the 2018 intervention, which may reflect the introduction of the most recent contraindication measures.

Prevalence and Factors Associated with Potential Drug-Drug Interactions in Older Community-Dwelling Adults: A Prospective Cohort Study

Background

Older patients are at increased risk of drug-drug interactions (DDIs) due to polypharmacy. Cardiovascular and central nervous system (CNS) drugs are commonly implicated in serious DDIs.

Objectives

This study aimed to determine the prevalence and factors associated with potential ‘severe’ cardiovascular and CNS DDIs among older (≥ 70 years) community-dwellers.

Methods

This was a prospective cohort study using linked data from a national pharmacy claims database and waves 1 and 2 of The Irish LongituDinal study on Ageing (TILDA). ‘Severe’ cardiovascular and CNS DDIs were identified using the British National Formulary 77 and Stockley’s Drug Interactions. The prevalence of ‘severe’ DDIs (any DDI vs. none) was calculated. Logistic regression was used to examine the association between sociodemographic, functional ability, and medication-related factors and the risk of DDI exposure between waves 1 and 2.

Results

A total of 1466 patients were included [mean age (standard deviation) = 78 (5.5) years; female n = 795, 54.2%]. In total, 332 community-dwellers aged ≥ 70 years [22.65%, 95% confidence interval (CI) 20.58–24.86] were potentially exposed to at least one ‘severe’ cardiovascular or CNS DDI, with more than half (54.82%) of this cohort dispensed the same DDI for a prolonged time (≥ 3 consecutive claims). Aspirin-warfarin was the most frequently dispensed (co-prescribed) DDI (n = 34, 10.24%, 95% CI 7.39–14.00), followed by atorvastatin-clarithromycin (n = 19, 5.72%, 95% CI 3.64–8.81). Polypharmacy [≥ 10 vs. < 5 drugs, odds ratio (OR) 13.40, 95% CI 8.22–21.85] and depression (depressed vs. not, OR 2.12, 95% CI 1.34–3.34) were significantly associated with these DDIs, after multivariable adjustment.

Conclusion

‘Severe’ cardiovascular and CNS DDIs are prevalent in older community-dwellers in Ireland, and those with polypharmacy and depression are at a significantly increased risk.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40266-021-00898-8.

Trends in drug poisoning deaths, by sex, in Ireland: a repeated cross-sectional study from 2004 to 2017

OBJECTIVE

To examine sex differences in age-standardised rates (ASR) of overall and drug-specific drug poisoning deaths in Ireland between 2004 and 2017.

DESIGN

Repeated cross-sectional study.

SETTING

Drug poisoning deaths in Ireland.

PARTICIPANTS

National Drug-Related Deaths Index and pharmacy claims database (Primary Care Reimbursement Service-General Medical Services) data from 2004 to 2017.

OUTCOME MEASURES

The primary outcome was trends in drug poisoning death rates by sex. The secondary outcomes were trends in drug poisoning death rates involving (1) any CNS (Central Nervous System) depressants, (2) ≥2 CNS depressants and (3) specific drugs/drug classes (eg, prescription opioids, benzodiazepines, antidepressants, alcohol, cocaine and heroin) by sex. Joinpoint regression was used to examine trends, stratified by sex, in the ASR of drug poisoning deaths (2004-2017), change points over time and average annual percentage changes (AAPCs) with 95% CI.

RESULTS

Increased ASR for all drug poisoning deaths from 6.86 (95% CI 6.01 to 7.72) per 100 000 in 2004 to 8.08 (95% CI 7.25 to 8.91) per 100 000 in 2017 was mainly driven by increasing deaths among men (AAPC 2.6%, 95% CI 0.2 to 5.1), with no significant change observed among women. Deaths involving ≥2 CNS depressants increased for both men (AAPC 5.6%, 95% CI 2.4 to 8.8) and women (AAPC 4.0%, 95% CI 1.1 to 6.9). Drugs with the highest significant AAPC increases for men were cocaine (7.7%, 95% CI 2.2 to 13.6), benzodiazepines (7.2%, 95% CI 2.9 to 11.6), antidepressants (6.1%, 95% CI 2.4 to 10.0) and prescription opioids (3.5%, 95% CI 1.6 to 5.5). For women, the highest AAPC was for antidepressants (4.2%, 95% CI 0.2 to 8.3), benzodiazepines (3.3%, 95% CI 0.1 to 6.5) and prescription opioids (3.0%, 95% CI 0.7 to 5.3).

CONCLUSION

Drugs implicated in drug poisoning deaths vary by sex. Policy response should include prescription monitoring programmes and practical harm reduction information on polydrug use, especially CNS depressant drugs.

The association between increasing oral anticoagulant prescribing and atrial fibrillation related stroke in Ireland

AIMS

Recent increases in the number of patients with atrial fibrillation (AF) prescribed oral anticoagulants (OAC) are evident in Ireland and internationally, largely due to the availability of direct oral anticoagulants (DOACs). This study aimed to determine the rate of stroke in the context of increasing anticoagulation utilisation, with a focus on AF-related ischaemic stroke (IS).

METHODS

Dispensing data for OACs were identified for the period 2010-2018 as well as hospital discharges for IS (2005-2018). Irish National Stroke Register data were used to elucidate the characteristics of patients with acute ischaemic stroke.

RESULTS

The number of patients prescribed OACs increased by 94% from 2010-2018 with a significant change from 2013 (β = 2.57, P = .038), associated with a large increase in the number of patients on DOACs. There was 3.3-fold increase in expenditure on OACs nationally from 2013 to 2018, of which 94% was DOAC related. Using the 2013 timepoint, ischaemic stroke rates until 2018 did not show a significant deviation from the previous trend (β = 0.00, P = .898). The percentage of AF-related ischaemic stroke was stable from 2013 to 2017 with a 4.5% decrease in 2018. The percentage of ischaemic stroke patients with previously diagnosed AF decreased from 2013 to 2018; however, there was an increase in the percentage of ischaemic strokes while on OAC in this cohort.

CONCLUSION

Large increases in OAC utilisation have not resulted in changes in ischaemic stroke rates at a national level. The percentage of ischaemic strokes with a previous diagnosis of AF has decreased indicating a possible benefit from greater OAC utilisation. However, the percentage presenting with an ischaemic stroke while on OAC treatment is increasing. The increase in patients presenting with stroke while treated with OAC may largely reflect the national increase in patients prescribed DOACs but the findings raise concerns about treatment failures. The real-world effectiveness of DOACs requires further examination.

Using administrative health data for palliative and end of life care research in Ireland: potential and challenges

Background: This study aims to examine the potential of currently available administrative health and social care data for palliative and end-of-life care (PEoLC) research in Ireland. Objectives include to i) identify data sources for PEoLC research ii) describe the challenges and opportunities of using these and iii) evaluate the impact of recent health system reforms and changes to data protection laws.

Methods: The 2017 Health Information and Quality Authority catalogue of health and social care datasets was cross-referenced with a recognised list of diseases with associated palliative care needs. Criteria to assess the datasets included population coverage, data collected, data dictionary and data model availability, and mechanisms for data access.

Results: Nine datasets with potential for PEoLC research were identified, including death certificate data, hospital episode data, pharmacy claims data,  one national survey, four disease registries (cancer, cystic fibrosis, motor neurone and interstitial lung disease) and a national renal transplant registry.  The ad hoc development of the health system in Ireland has resulted in i) a fragmented information infrastructure resulting in gaps in data collections particularly in the primary and community care sector where much palliative care is delivered, ii) ill-defined data governance arrangements across service providers, many of whom are not part of the publically funded health service and iii) systemic and temporal issues that affect data quality. Initiatives to improve data collections include introduction of i) patient unique identifiers, ii) health entity identifiers and iii) integration of the Eircode postcodes. Recently enacted general data protection and health research regulations will clarify legal and ethical requirements for data use.

Conclusions: Ongoing reform initiatives and recent changes to data privacy laws combined with detailed knowledge of the datasets, appropriate permissions, and good study design will facilitate future use of administrative health and social care data for PEoLC research in Ireland.

Trends in strong opioid prescribing in Ireland: A repeated cross-sectional analysis of a national pharmacy claims database between 2010 and 2019

PURPOSE

Significant increases in opioid utilisation have been reported in many countries in recent decades. This study investigated strong opioid prescribing in Irish General Medical Services (GMS) patients over a 10-year period.

METHODS

A retrospective repeated cross-sectional analysis of a national pharmacy claims database between January 2010 and December 2019 was conducted. Strong opioid prescribing in GMS patients was evaluated, including by route of administration, age (16-64 years and ≥65 years) and gender. Measures of consumption included prescribing prevalence and defined daily dose (DDD)/1000 population/day. Prevalence ratios (PRs) with 95% confidence intervals (CIs), and percentage and absolute changes were determined.

RESULTS

Strong opioid prescribing prevalence increased from 14.43% in 2010 to 16.28% in 2019, with the greatest increase in the ≥65 years age group. Tramadol was the most frequently prescribed product, constituting 63.9% of total strong opioid prescribing. The prescribing prevalence of oxycodone increased from 0.95% in 2010 to 2.68% in 2019 (PR 2.81, 95% CI 2.76, 2.87), with steep increases in oxycodone-naloxone since it became available (PR 5.23, 95% CI 4.98, 5.50). The prescribing prevalence of tapentadol increased from 0.18% to 1.58% between 2012 (first complete year available for reimbursement) and 2019 (PR 8.79, 95% CI 8.43, 9.16). Strong opioid prescribing was highest in females aged ≥65 years.

CONCLUSIONS

This study found an overall increase in strong opioid prescribing in Ireland between 2010 and 2019, particularly in older adults. Tramadol was the most frequently prescribed product, with oxycodone and tapentadol prescribing increasing markedly over the study period.

Using linked administrative health data for palliative and end of life care research in Ireland: potential and challenges

Background: This study aims to examine the potential of currently available administrative health data for palliative and end-of-life care (PEoLC) research in Ireland. Objectives include to i) identify administrative health data sources for PEoLC research ii) describe the challenges and opportunities of using these and iii) estimate  the impact of recent health system reforms and changes to data protection laws.  Methods: The 2017 Health Information and Quality Authority catalogue of health and social care datasets was cross-referenced with a recognised list of diseases with associated palliative care needs. Criteria to assess the datasets included population coverage, data collected, data dictionary and data model availability and mechanisms for data access. Results: Eight datasets with potential for PEoLC research were identified, including four disease registries, (cancer, cystic fibrosis, motor neurone and interstitial lung disease), death certificate data, hospital episode data, community prescription data and one national survey. The ad hoc development of the health system in Ireland has resulted in i) a fragmented information infrastructure resulting in gaps in data collections particularly in the primary and community care sector where much palliative care is delivered, ii) ill-defined data governance arrangements across service providers, many of whom are not part of the publically funded health service and iii) systemic and temporal issues that affect data quality. Initiatives to improve data collections include introduction of i) patient unique identifiers, ii) health entity identifiers and iii) integration of the eircode postcodes. Recently enacted general data protection and health research regulations will clarify legal and ethical requirements for data  use. Conclusions: With appropriate permissions, detailed knowledge of the datasets and good study design currently available administrative health data can be used for PEoLC research. Ongoing reform initiatives and recent changes to data privacy laws will facilitate future use of administrative health data for PEoLC research.

Study designs and statistical approaches to suicide and prevention research in real-world data

OBJECTIVE

To provide researchers, clinicians and policy makers with a primer to study designs, statistical approaches and graphical reporting methods for suicide research in real world data (RWD).

METHODS

Study designs, statistical method and graphical reporting standards are detailed with examples from the recently published literature.

RESULTS

Data sources and codes for identifying suicidal behavior are described. Study designs are described in detail for post-market surveillance, retrospective cohort studies, case control and nested case-control studies, and self-controlled (within-individual) studies including applications of marginal structural models. Graphical reporting of designs is described using an original research study.

CONCLUSIONS

Compared to RCTs, RWE studies offer larger sample sizes, greater generalizability, and real-world validity. However, these non-experimental data risk uncontrolled confounding and potential introduction of bias unless data, design and statistical approaches are rigorously aligned.

Impact of changes in controlled drugs legislation on benzodiazepine receptor agonist prescribing in Ireland: a repeated cross-sectional study

Purpose

To examine the impact of new controlled drugs legislation introduced in May 2017 on benzodiazepine receptor agonist (BZRA) prescribing in Ireland.

Methods

A repeated cross-sectional analysis was conducted using publically available monthly pharmacy claims data from the General Medical Services (GMS) database. The study population comprised all GMS-eligible individuals aged ≥ 16 years from January 2016 to September 2019. Monthly prevalence rates of individuals receiving BZRA prescriptions per 10,000 eligible population were calculated and trends examined over time. Segmented linear regression of prevalence rates was used to examine changes before and after introduction of the legislation stratified by gender and age groups. Regression coefficients (β) and 95% confidence intervals (CIs) for monthly change were calculated.

Results

Pre-legislation (January 2016 to April 2017), there was a significant monthly decline in benzodiazepine prevalence rate (β = − 1.18; 95% CI − 1.84, − 0.51; p < 0.001) but no significant change in Z-drug prescribing. Post-legislation (May 2017 to September 2019), increases in prevalence rates were observed for benzodiazepines (β = 1.04; 95% CI 0.17, 1.92; p = 0.021) and Z-drugs (β = 1.04; 95% CI 0.26, 1.83; p = 0.010). Post-legislation trends showed increases in BZRA prevalence rates among the youngest subgroup (16–44 years), with variable changes in the middle-aged subgroup (45–64 years) and no changes in the oldest subgroup (≥ 65 years).

Conclusions

This study indicates that introduction of new legislation had limited impact on BZRA prescribing on the main public health scheme in Ireland. Interventions targeting specific population subgroups may be required to achieve sustained reductions in prescribing.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00228-020-03063-z.

Effect of an Online Reimbursement Application System on Prescribing of Lidocaine 5% Medicated Plaster in the Republic of Ireland

BACKGROUND

The lidocaine 5% medicated plaster, Versatis^®, has one therapeutic indication listed on the Summary of Product Characteristics-symptomatic relief of post-herpetic neuralgia (PHN) in adults. Increased expenditure on Versatis^® suggests that there is considerable off-label use. To support the appropriate use of Versatis^®, the Health Service Executive's Primary Care Reimbursement Service (PCRS) introduced a reimbursement application system for Versatis^® from 1 September 2017.

OBJECTIVE

The aim of this study was to investigate the effect of introducing a reimbursement application system on Versatis^® prescribing under the General Medical Services (GMS) scheme.

METHODS

This study was carried out using prescription dispensing data from the PCRS pharmacy claims database. We carried out segmented linear regression to assess changes in the Versatis^® prescribing rate per 1000 GMS eligible population, before and after the introduction of the online reimbursement application system.

RESULTS

The results of the segmented regression analysis show that there was a statistically significant level (- 4.91, p < 0.001) and trend change (- 0.69, p < 0.001) in the rate of Versatis^® prescribing post-introduction of the reimbursement application system. In the year prior to the introduction of the system, 2016, the annual GMS expenditure on Versatis^® lidocaine 5% patches was over €27 million, whereas the GMS expenditure in 2018 was reduced to just over €2 million.

CONCLUSION

In our study, a substantial decrease in the dispensing of Versatis^® was seen after the implementation of a reimbursement application system. Prescribing of Versatis^® should be restricted to patients with a diagnosis of PHN not only to reduce costs, but to ensure evidence-based use of this medication.

Cognitive decline associated with anticholinergics, benzodiazepines and Z-drugs: Findings from The Irish Longitudinal Study on Ageing (TILDA)

AIMS

To estimate the association between patterns of anticholinergic, benzodiazepine and Z-drug medication use and change in cognitive function in middle-aged and older adults.

METHODS

This prospective cohort study used data from the first three waves of The Irish Longitudinal Study on Ageing (TILDA), including community-dwelling adults aged ≥50 years followed for up to 4 years (n = 7027). Cognitive function was assessed using the Mini Mental State Examination, animal naming test and word recall tests. Regular medication use was self-reported at baseline and follow-up interviews at 2 and 4 years. Pharmacy dispensing claims for a subset (n = 2905) allowed assessment of medication use between interviews and cumulative dosage. Medication use at consecutive waves of TILDA was analysed in relation to change in cognitive function between waves.

RESULTS

Strongly anticholinergic medications (Anticholinergic Cognitive Burden scale 3), benzodiazepines and Z-drugs were reported by 7.3%, 5.8% and 5.1% of participants, respectively, at any time during the study. Adjusting for potential confounders, new anticholinergic use between interviews was associated with change in recall score (-1.09, 95% confidence interval -1.64, -0.53) over 2 years compared to non-use, but not with MMSE (0.07; 95% CI -0.21, 0.34) or animal naming (-0.70; 95% CI -1.43, 0.03). The pharmacy claims analysis was consistent with this finding. Other hypothesised associations were not supported.

CONCLUSIONS

Except for new use of anticholinergic medications, no other findings supported a risk of cognitive decline over 2-year periods in this middle-aged and older cohort. Patients and prescribers should weigh this potential risk against potential benefits of commencing anticholinergic medications.

Deciphering patterns of respiratory medication use in Ireland to target interventions appropriately: a focus on COPD

BACKGROUND

Ireland has the highest rates of overnight hospitalisations for COPD in the OECD, yet lacks estimates of the prevalence of this disease or its pharmacological management. We aimed to estimate the age and sex-specific prevalence of symptomatic COPD and to identify patterns of respiratory medication use to inform interventions to improve pharmacotherapy in this condition.

METHODS

We used the national pharmacy claims database, with data on a publically insured cohort in 2016. We restricted to those aged ≥ 45 years with full eligibility for that year and examined the age and sex distribution of respiratory medications, and patterns of medication use in those suggestive of COPD.

RESULTS

In this cohort, 23% filled at least one prescription for a respiratory medication; 14% of males and 16% of females received at least one dispensing of an ICS inhaler. The proportion dispensed a long-acting muscarinic receptor antagonist (LAMA) was considerably lower. Of those newly initiated on a LAMA, 24% did not receive another within 60 days of the last covered day. The prevalence of medication use suggestive of COPD was 15% in males and 16% in females.

CONCLUSION

The prevalence of medication use consistent with the management of symptomatic COPD mirrors international prevalence estimates. Several patterns raise concern: high ICS use in older adults, under use of LAMA therapy and poor persistence of those newly initiated. We recommend the development of an intervention to assist in the implementation of new national prescribing guidelines for the management of COPD.

Trends in national pharmaceutical expenditure on diabetes in Ireland 2011-2015: a repeated cross-sectional study

OBJECTIVES

To explore trends in pharmaceutical expenditure on diabetes between 2011 and 2015, describing trends in expenditure on blood glucose-lowering medications and estimating the effect of cost-containment measures implemented during this time.

DESIGN

Repeated cross-sectional study of national pharmacy claims data in Ireland.

PARTICIPANTS

Patients' dispensed items used in the treatment or management of diabetes.

PRIMARY AND SECONDARY OUTCOMES

Total expenditure associated with diabetes was calculated by extracting data on all diabetes-related items dispensed to eligible patients. Costs were categorised into two groups. Diabetes-specific items include items used directly in diabetes treatment (WHO-Anatomical Therapeutic Chemical (ATC): A10, V07, V04) and diabetes-related include all other condition-related items (WHO-ATC: B01, C, H04, N03, N06). The impacts of two specific cost-containment measures, co-payments and reference pricing, were assessed using segmented linear regression analyses of interrupted time-series.

RESULTS

Total expenditure varied over the study period, peaking at €216 994 441 in 2012. Expenditure on diabetes-specific items increased steadily by 18% reaching €153 621 477 in 2015, with blood glucose-lowering medications accounting for 73% of this increase. During the same period, expenditure on diabetes-related items decreased by 32% to €50 835 856. The introduction of reference pricing for atorvastatin in November 2013 resulted in immediate costs savings of €2.4 million per yearly quarter (level-change p<0.001).

CONCLUSIONS

The increasing expenditure on blood glucose-lowering medications negates the effect of recent cost-containment measures, presenting a significant challenge for the provision of diabetes care. Innovative policies are required to ensure high-quality diabetes care can be provided at an equitable, affordable and sustainable rate.

Agreement between self-reported morbidity and pharmacy claims data for prescribed medications in an older community based population

Background

Studies have indicated variability around prevalence estimates of multimorbidity due to poor consensus regarding its definition and measurement. Medication-based measures of morbidity may be valuable resources in the primary-care setting where access to medical data can be limited. We compare the agreement between patient self-reported and medication-based morbidity; and examine potential patient-level predictors of discordance between these two measures of morbidity in an older (≥ 50 years) community-based population.

Methods

A retrospective cohort study was performed using national pharmacy claims data linked to The Irish LongituDinal study on Ageing (TILDA). Morbidity was measured by patient self-report (TILDA) and two medication-based measures, the Rx-Risk (< 65 years) and Rx-Risk-V (≥65 years), which classify drug claims into chronic disease classes. The kappa statistic measured agreement between self-reported and medication-based morbidity at the individual patient-level. Multivariate logistic regression was used to examine patient-level characteristics associated with discordance between measures of morbidity.

Results

Two thousand nine hundred twenty-five patients were included (< 65 years: N = 1095, 37.44%; and ≥ 65 years: N = 1830 62.56%). Hypertension and high cholesterol were the most prevalent self-reported morbidities in both age cohorts. Agreement was good or very good (κ = 0.61–0.81) for diabetes, osteoporosis and glaucoma; and moderate for high cholesterol, asthma, Parkinson’s and angina (κ = 0.44–0.56). All other conditions had fair or poor agreement. Age, gender, marital status, education, poor-delayed recall, depression and polypharmacy were significantly associated with discordance between morbidity measures.

Conclusions

Most conditions achieved only moderate or fair agreement between self-reported and medication-based morbidity. In order to improve the accuracy in prevalence estimates of multimorbidity, multiple measures of multimorbidity may be necessary. Future research should update the current Rx-Risk algorithms in-line with current treatment guidelines, and re-assess the feasibility of using these indices alone, or in combination with other methods, to yield more accurate estimates of multimorbidity.

A retrospective analysis of budget impact models submitted to the National Centre for Pharmacoeconomics in Ireland

BACKGROUND

The National Centre for Pharmacoeconomics (NCPE) is a National HTA Agency in Ireland responsible for assessment of comparative clinical effectiveness, cost-effectiveness and potential budget impact of drugs on behalf of the Health Service Executive. This research aims to assess if the budget impact models submitted to the NCPE have accurate predicted utilisation, assess if the models are consistent in the parameters included, and determine if probabilistic sensitivity analyses would aid the characterization of uncertainty.

METHODS

A retrospective analysis of budget impact models that had been submitted (January 2010-December 2017 inclusive) to the NCPE was performed. The input parameters in the budget impact model were recorded. For each drug, annual realised utilisation was compared with what had been predicted by the respective budget impact model. A probabilistic sensitivity analysis was also performed on each model.

RESULTS

A total of 12 models were included; each model pertained to one drug for one indication. Of the 12 models, six underpredicted and six overpredicted the annual realised utilisation. There were a range of different parameters included in each of the budget impact models. A probabilistic sensitivity analysis did not improve the characterization of uncertainty.

CONCLUSION

This research has demonstrated that budget impact models submitted to a national HTA agency have limited accuracy in predicting realised utilisation, and there is inconsistency among the parameters included. An electronic budget impact template for applicants has been developed, as a more systematic approach, for their submissions to the NCPE.

Longitudinal Medication Adherence in Older Adults With Multimorbidity and Association With Health Care Utilization: Results From the Irish Longitudinal Study on Ageing

BACKGROUND

The association between objectively measured nonadherence and health care utilization in multimorbid older people is unclear.

OBJECTIVE

To measure medication adherence across multiple chronic conditions, identify adherence patterns, and estimate the association between adherence and self-reported health care utilization.

METHODS

This is a retrospective cohort study of multimorbid participants aged ≥70 years in the Irish LongituDinal Study on Ageing (TILDA). Eligible participants had linked pharmacy claims data and completed TILDA wave 2 (2 years after wave 1). The RxRisk-V tool was used to identify multimorbidity. Average adherence (AA) across RxRisk-V conditions was estimated using the CMA7 function (AdhereR). Group-based trajectory models (GBTMs) identified adherence patterns in the 12 months following wave 1. Negative binomial regression was used to estimate the association between adherence and the rate of subsequent self-reported general practitioner [GP] visits, emergency department (ED) visits, outpatient visits, and hospitalizations in the 12 months following adherence measurement (reported at wave 2). Adjusted Incident Rate Ratios (aIRR) and 95% CIs are presented.

RESULTS

Higher AA (CMA7) was associated with a small significant decrease in GP visit rate (aIRR = 0.70; CI = 0.53-0.94) and outpatient visit rate (aIRR = 0.44; CI = 0.23-0.81). GBTM identified 6 adherence groups (n = 1050). Compared with high adherers, group 1 (rapid decline, modest increase) membership (aIRR = 1.72; CI = 1.09-2.73) and group 4 (high adherence, delayed decline) membership (aIRR = 1.92; CI = 1.19-3.05) significantly increased ED visit rate.

CONCLUSION AND RELEVANCE

Suboptimal medication adherence in multimorbid older adults is associated with increased health care utilization. Identification of suboptimal adherence groups for medication management interventions may help decrease the health system burden and health care costs.

Low uptake of continuous subcutaneous insulin infusion therapy in people with type 1 diabetes in Ireland: a retrospective cross-sectional study

BACKGROUND

The uptake of continuous subcutaneous insulin infusion (CSII) therapy in those with type 1 diabetes varies internationally and is mainly determined by the national healthcare reimbursement systems. The aim of this study is to estimate national and regional uptake of CSII therapy in children, adolescents and adults with type 1 diabetes in Ireland.

METHODS

A retrospective cross-sectional study was conducted utilizing the national pharmacy claims database in 2016. Individuals using CSII were identified by dispensing of infusion sets. The uptake of CSII was calculated as the percentage of people with type 1 diabetes claiming CSII sets in 2016, both in children and adolescent (age < 18 years) and adult populations (≥ 18 years). Descriptive statistics including percentages with 95% confidence intervals (CIs) are presented, stratified by age-groups and geographical regions, and chi-square tests used for comparisons.

RESULTS

Of 20,081 people with type 1 diabetes, 2111 (10.5, 95% CI: 10.1-10.9%) were using CSII in 2016. Uptake was five-fold higher in children and adolescents at 34.7% (95% CI: 32.9-36.5%) than in adults at 6.8% (95% CI: 6.4-7.2%). Significant geographical heterogeneity in CSII uptake was found, from 12.6 to 53.7% in children and adolescents (p < 0.001), and 2 to 9.6% in adults (p < 0.001).

CONCLUSIONS

Uptake of CSII in people with type 1 diabetes is low in Ireland, particularly in those ≥18 years. Identification of barriers to uptake, particularly in this group, is required.

Overcoming hurdles: measurement of health-related outcomes associated with national level medicines usage in Ireland

Public discourse on medicine provision predominantly focuses on overall expenditure. However, current literature suggests measurement of alternative indicators can provide a method to benchmark or ameliorate medicine provision. Previous research has investigated the viability of using health-related outcome metrics, such as the number of patients treated, quality-adjusted life-year gain and life-year gain, to provide macro-level estimates on medicines’ societal contributions. This editorial provides an overview of the evolving healthcare landscape surrounding medicine usage estimation and valuation in Ireland and offers recommendations on how improved methods of measuring health-related outcomes may help ameliorate efficiencies and the sustainability of a healthcare system.

Prevalence and incidence of type 1 diabetes in Ireland: a retrospective cross-sectional study using a national pharmacy claims data from 2016

OBJECTIVES

The aim of this study is to estimate the prevalence and incidence of type 1 diabetes in the Irish population using a national pharmacy claims database in the absence of a national diabetes register.

DESIGN

National, population-based, retrospective, cross-sectional study.

SETTING

Community care with data available through the Health Service Executive Pharmacy Claims Reimbursement Scheme from 2011 to 2016.

PARTICIPANTS

Individuals with type 1 diabetes were identified by coprescription of insulin and glucometer test strips without any prolonged course (>12 months) of oral hypoglycaemic agents prior to commencing insulin. Those claiming prescriptions for long-acting insulin only, without any prandial insulin, were excluded from the analysis. Incidence was estimated based on the first claim for insulin in 2016, with no insulin use in the preceding 12 months.

MAIN OUTCOME MEASURES

Prevalence of type 1 diabetes in children (<18 years) and adults (≥18 years); incidence of type 1 diabetes in children (≤14 years) and adolescents and adults (>14 years).

RESULTS

There were 20 081 prevalent cases of type 1 diabetes in 2016. The crude prevalence was 0.42% (95% CI 0.42% to 0.43%). Most prevalent cases (n=17 053, 85%) were in adults with a prevalence of 0.48% (95% CI 0.47% to 0.48%). There were 1527 new cases of type 1 diabetes in 2016, giving an incidence rate of 32 per 100 000 population/year (95% CI 30.5 to 33.7). There was a significant positive linear trend for age, for prevalence (p<0.0001) and incidence (p=0.014). The prevalence and incidence were 1.2-fold and 1.3-fold higher in men than women, respectively. Significant variations in prevalence (p<0.0001) and incidence (p<0.001) between the different geographical regions were observed.

CONCLUSIONS

This study provides epidemiological estimates of type 1 diabetes across age groups in Ireland, with the majority of prevalent cases in adults. Establishing a national diabetes register is essential to enable updated epidemiological estimates of diabetes and for planning of services in Ireland.

Fitting Marginal Structural and G-Estimation Models Under Complex Treatment Patterns: Investigating the Association Between De Novo Vitamin D Supplement Use After Breast Cancer Diagnosis and All-Cause Mortality Using Linked Pharmacy Claim and Registry Data

Studies have shown that accounting for time-varying confounding through time-dependent Cox proportional hazards models may provide biased estimates of the causal effect of treatment when the confounder is also a mediator. We explore 2 alternative approaches to addressing this problem while examining the association between vitamin D supplementation initiated after breast cancer diagnosis and all-cause mortality. Women aged 50-80 years were identified in the National Cancer Registry Ireland (n = 5,417) between 2001 and 2011. Vitamin D use was identified from linked prescription data (n = 2,570). We sought to account for the time-varying nature of vitamin D use and time-varying confounding by bisphosphonate use using 1) marginal structural models (MSMs) and 2) G-estimation of structural nested accelerated failure-time models (SNAFTMs). Using standard adjusted Cox proportional hazards models, we found a reduction in all-cause mortality in de novo vitamin D users compared with nonusers (hazard ratio (HR) = 0.84, 95% confidence interval (CI): 0.73, 0.99). Additional adjustment for vitamin D and bisphosphonate use in the previous month reduced the hazard ratio (HR = 0.45, 95% CI: 0.33, 0.63). Results derived from MSMs (HR = 0.44, 95% CI: 0.32, 0.61) and SNAFTMs (HR = 0.45, 95% CI: 0.34, 0.52) were similar. Utilizing MSMs and SNAFTMs to account for time-varying bisphosphonate use did not alter conclusions in this example.

Out of pocket or out of control: A qualitative analysis of healthcare professional stakeholder involvement in pharmaceutical policy change in Ireland

BACKGROUND

Mandatory co-payments attached to prescription medicines on the Irish public health insurance [General Medical Services (GMS)] scheme have undergone multiple iterations since their introduction in October 2010. To date, whilst patients' opinions on said co-payments have been evaluated, the perspectives of community pharmacists and general practitioners (GPs) have not.

OBJECTIVE

To explore the involvement and perceptions of community pharmacists and GPs on this pharmaceutical policy change.

METHODS

A qualitative study using purposive sampling alongside snowballing recruitment was used. Nineteen interviews were conducted in a Southern region of Ireland. Data were analysed using the Framework Approach.

RESULTS

Three major themes emerged: 1) the withered tax-collecting pharmacist; 2) concerns and prescribing patterns of physicians; and 3) the co-payment system - impact and sustainability. Both community pharmacists and GPs accepted the theoretical concept of a co-payment on the GMS scheme as it prevents moral hazard. However, there were multiple references to the burden that the current method of co-payment collection places on community pharmacists in terms of direct financial loss and reductions in workplace productivity. GPs independently suggested that a co-payment system may inhibit moral hazard by GMS patients in the utilisation of GP services. It was unclear to participants what evidence is guiding the GMS co-payment fee changes.

CONCLUSION

Interviewees accepted the rationale for the co-payment system, but reform is warranted.

Direct oral anticoagulants uptake and an oral anticoagulation paradox

Oral anticoagulation (OAC) for stroke prevention in patients with atrial fibrillation is underutilised. One of the impediments to warfarin therapy is the frequent monitoring required, usually at a specialised warfarin clinic. The advent of direct oral anticoagulants (DOACs) facilitates OAC therapy without an onerous monitoring regimen. This benefit may result in the more significant adoption of DOACs in areas without a warfarin clinic. This study analysed national administrative data for reimbursed pharmacy claims to assess OAC prescribing from 2010 to 2017 and compared the use of DOACs in areas with warfarin clinics compared to those without. Over the study period, the number of patients on OAC increased by 84%, due to a rapid increase in DOAC prescribing. The findings demonstrate that DOACs have resulted in an increase in the overall uptake of OAC therapy in Ireland. However, the increased utilisation was not evidently related to populations underserved by warfarin clinics.

Comparing Potentially Inappropriate Prescribing Tools and Their Association With Patient Outcomes

OBJECTIVE

To assess the agreement of several different measures of potentially inappropriate prescribing (PIP) in older people and compare their relationship with patient-reported outcomes.

DESIGN

Prospective cohort study including participants in The Irish Longitudinal Study on Ageing (TILDA).

SETTING

Waves 1 and 2 of TILDA, a nationally representative aging cohort study.

PARTICIPANTS

A total of 1753 community-dwelling TILDA participants with linked administrative pharmacy claims data on medications.

MEASUREMENTS

Potentially inappropriate medications were assessed using the Screening Tool of Older Persons' Potentially Inappropriate Prescriptions (STOPP) v1, American Geriatrics Society (AGS) Beers Criteria® 2012, and relevant Assessing Care of Vulnerable Elders (ACOVE) v3 indicators. Potential prescribing omissions were assessed using the Screening Tool to Alert Doctors to the Right Treatment (START) v1 and ACOVE v3 indicators. Their agreement was assessed via κ statistics, and multivariate regression was used to assess relationships with emergency department visits, general practitioner (GP) visits, quality of life, and functional decline (increased assistance needed for activities of daily living).

RESULTS

There was slight agreement between STOPP and AGS Beers Criteria® (κ = 0.20) and ACOVE indicators (κ = 0.15), while agreement between AGS Beers Criteria® and ACOVE indicators was fair (κ = 0.31). Agreement was fair between START and ACOVE indicators (κ = 0.34). All measures of inappropriate medications were significantly associated with increased GP visits. Only exposure to two or more START indicators was associated with reduced quality of life (adjusted mean difference = -1.12; 95% confidence interval [CI] = -1.92 to -0.33), and only two or more AGS Beers Criteria® were associated with functional decline (adjusted odds ratio = 2.11; 95% CI = 1.37-3.28). For omissions, both measures were associated with functional decline, but only ACOVE indicators were associated with increased GP visits.

CONCLUSION

Prevalence of PIP and relationships with outcomes can differ substantially between tools with little agreement. Choice of PIP measure for research or practice should be considered in light of the circumstances and requirements in each case. J Am Geriatr Soc 68:526-534, 2020.

Cognitive impairment and medication adherence post-stroke: A five-year follow-up of the ASPIRE-S cohort

BACKGROUND

Control of vascular risk factors is essential for secondary stroke prevention. However, adherence to secondary prevention medications is often suboptimal, and may be affected by cognitive impairment. Few studies to date have examined associations between cognitive impairment and medication adherence post-stroke, and none have considered whether adherence to secondary prevention medications might affect subsequent cognitive function. The aim of this study was to explore prospective associations between cognitive impairment and medication non-adherence post-stroke.

METHODS

A five-year follow-up of 108 stroke survivors from the Action on Secondary Prevention Interventions and Rehabilitation in Stroke (ASPIRE-S) prospective observational cohort study. Cognitive function was assessed using the Montreal Cognitive Assessment at 6 months, and a neuropsychological test battery at 5 years. Adherence to antihypertensive, antithrombotic and lipid-lowering medications was assessed using prescription refill data.

RESULTS

The prevalence of cognitive impairment at five years was 35.6%. The prevalence of non-adherence ranged from 15.1% for lipid-lowering agents to 30.2% for antithrombotics. There were no statistically significant associations between medication non-adherence in the first year post-stroke and cognitive impairment at 5 years, nor between cognitive impairment at 6 months and non-adherence at 5 years. Stroke survivors with cognitive impairment were significantly more likely to report receiving help with taking medications [OR (95% CI): 4.84 (1.17, 20.07)].

CONCLUSIONS

This is the first study to explore the potential impact of non-adherence to secondary prevention medications on cognitive impairment in stroke survivors. Findings highlight the role of family members and caregivers in assisting stroke survivors with medication administration, particularly in the context of deficits in cognitive function. Involving family members and caregivers may be a legitimate and cost-effective strategy to improve medication adherence in stroke survivors.

Association between adherence to antihypertensive medications and health outcomes in middle and older aged community dwelling adults; results from the Irish longitudinal study on ageing

PURPOSE

To examine the association between antihypertensive medication (AHTM) implementation adherence and healthcare utilisation in community-dwelling adults aged ≥ 50 years in Ireland.

METHODS

This was a prospective cohort study. The Irish Longitudinal Study on Ageing (TILDA) was linked to pharmacy claims data for participants aged ≥ 50 years. Participants were included if they had ≥ 3 pharmacy claims for one or more AHTM (ATC codes 'C02', 'C03', 'C07', 'C08' or 'C09') within the year preceding the year of self-reported healthcare utilisation outcome occurrence. Outcomes included self-reported general practitioner (GP), emergency department (ED), outpatient department visits and hospital admissions. Implementation adherence was measured using proportion of days covered (PDC), with participants classified as adherent if the average PDC ≥ 0.8. Negative binomial models were used to analyse the association between AHTM adherence and number of GP, ED, outpatient visits and hospitalisations (adjusted IRR and 95% CI are presented).

RESULTS

One thousand four hundred thirty-one participants were included. The majority of participants (72.6%) were considered adherent. Good implementation adherence to AHTM was associated with a significant decrease in self-reported GP visits (adjusted IRR 0.91, 95% CI 0.83-0.99). Adherence had no significant impact on the number of ED visits, outpatient visits or hospitalisations reported by TILDA participants.

CONCLUSIONS

Good adherence to AHTM was associated with less self-reported GP visits in this population, suggesting improved overall health status. However, the impact of medication non-adherence on the other self-reported healthcare utilisation outcomes (ED, outpatient visits and hospitalisations) was not evident in this study.

Impact of drug burden index on adverse health outcomes in Irish community-dwelling older people: a cohort study

Background

The Drug Burden Index (DBI) quantifies exposure to medications with anticholinergic and/or sedative effects. A consensus list of DBI medications available in Ireland was recently developed for use as a DBI tool. The aim of this study was to validate this DBI tool by examining the association of DBI score with important health outcomes in Irish community-dwelling older people.

Methods

This was a cohort study using data from The Irish Longitudinal Study on Ageing (TILDA) with linked pharmacy claims data. Individuals aged ≥65 years participating in TILDA and enrolled in the General Medical Services scheme were eligible for inclusion. DBI score was determined by applying the DBI tool to participants’ medication dispensing data in the year prior to outcome assessment. DBI score was recoded into a categorical variable [none (0), low (> 0 and < 1), and high (≥1)]. Outcome measures included any Activities of Daily Living (ADL) impairment, any Instrumental Activities of Daily Living (IADL) impairment, any self-reported fall in the previous 12 months, any frailty criterion met (Fried Phenotype measure), quality of life (QoL) score (CASP-19 [Control Autonomy Self-realisation Pleasure] measure), and healthcare utilisation (any hospital admission and any emergency department (ED) visit) in the previous 12 months. Statistical analyses included multivariate logistic and linear regression models controlling for potential confounders.

Results

61.3% (n = 1946) of participants received at least one DBI prescription in the year before their outcome assessment. High DBI exposure (DBI score ≥ 1) vs none was significantly associated with impaired function (ADL impairment adjusted OR 1.89, 95% CI 1.25, 2.88; IADL impairment adjusted OR 2.97, 95% CI 1.91, 4.61), self-reported falls (adjusted OR 1.50, 95%CI 1.03, 2.18), frailty (adjusted OR 1.74, 95% CI 1.14, 2.67), and reduced QoL (β = − 1.84, 95%CI -3.14, − 0.54). There was no significant association between DBI exposure and healthcare utilisation.

Conclusions

The findings validate the use of the DBI tool for predicting risk of functional impairment, falls, frailty and reduced QoL in older people in Ireland, and may be extended to other European countries. Integration of this tool into routine practice may be an appropriate step forward to improve outcomes in older people.

Prescribing pattern of antihypertensive drugs in two European cohorts: a population-based database study

Background: Antihypertensive drugs play a crucial role in reducing cardiovascular morbidity and mortality. Variability in prescribing patterns constitutes a major challenge for current healthcare systems. This study aimed to compare patterns of use of antihypertensives in general practice in two southern European populations. Methods: Observational study. Data on antihypertensive drugs consumption in primary care setting (2016) were obtained from pharmacy refill records in Campania (Italy) and Aragon (Spain). Prescribing rates and the number of defined daily doses [DDD/1,000 inhabitants/day (DID)] were calculated, and the Drug Utilization 90% (DU90%) approach used to reveal differences in prescribing patterns in both regions. Results: Antihypertensive prescribing rates in Campania and Aragon were 250.8 (95%CI: 250.2-251.3) and 201.7 (95%CI: 200.9-202.5) users/1,000 inhabitants/year. Overall consumption was of 310.1 and 256.8 DID, respectively. Spanish users, especially women and the elderly, consumed a greater volume of diuretics. Conversely, other therapeutic subgroups were more consumed in Campania. However, the most prescribed subgroups accounted for comparable proportions of the total consumption in each region. Conclusions: Both prescribing rates and intensity of antihypertensive use were higher in Campania. Pharmacy refill records in cross-country comparisons allow to know the factors influencing variability in prescribing habits with a view to improving prescribing quality.

Prevalence of potentially inappropriate prescribing in older people in primary care and its association with hospital admission: longitudinal study

OBJECTIVE

To determine whether hospital admission is associated with potentially inappropriate prescribing among older primary care patients (aged ≥65 years) and whether such prescribing was more likely after hospital admission than before.

DESIGN

Longitudinal study of retrospectively extracted data from general practice records.

SETTING

44 general practices in Ireland in 2012-15.

PARTICIPANTS

Adults aged 65 years or over attending participating practices.

EXPOSURE

Admission to hospital (any hospital admission versus none, and post-admission versus pre-admission).

MAIN OUTCOME MEASURES

Prevalence of potentially inappropriate prescribing assessed using 45 criteria from the Screening Tool for Older Persons' Prescription (STOPP) version 2, analysed both as rate of distinct potentially inappropriate prescribing criteria met (stratified Cox regression) and binary presence of potentially inappropriate prescribing (logistic regression) and adjusted for patients' characteristics. A sensitivity analysis used matching with propensity scores based on patients' characteristics and diagnoses.

RESULTS

Overall 38 229 patients were included, and during 2012 the mean age was 76.8 (SD 8.2) years and 43% (13 212) were male. Each year, 10.4-15.0% (3015/29 077 in 2015 to 4537/30 231 in 2014) of patients had at least one hospital admission. The overall prevalence of potentially inappropriate prescribing ranged from 45.3% (13 940/30 789) of patients in 2012 to 51.0% (14 823/29 077) in 2015. Independently of age, sex, number of prescription items, comorbidity, and health cover, hospital admission was associated with a higher rate of distinct potentially inappropriate prescribing criteria met; the adjusted hazard ratio for hospital admission was 1.24 (95% confidence interval 1.20 to 1.28). Among participants who were admitted to hospital, the likelihood of potentially inappropriate prescribing after admission was higher than before admission, independent of patients' characteristics; the adjusted odds ratio for after hospital admission was 1.72 (1.63 to 1.84). Analysis of propensity score matched pairs showed a slight reduction in the hazard ratio for hospital admission to 1.22 (1.18 to 1.25).

CONCLUSION

Hospital admission was independently associated with potentially inappropriate prescribing. It is important to determine how hospital admission may affect appropriateness of prescribing for older people and how potential adverse consequences of admission can be minimised.

Adherence to chronic medication in older populations: application of a common protocol among three European cohorts

PURPOSE

The purpose of this study was to evaluate and compare medication adherence to chronic therapies in older populations across different regions in Europe.

METHODS

This explorative study applied a harmonized method of data extraction and analysis from pharmacy claims databases of three European countries to compare medication adherence at a cross-country level. Data were obtained for the period between January 1, 2010, and December 31, 2011. Patients (aged ≥65 years) who newly initiated to oral antidiabetics, antihyperlipidemics, or antiosteoporotics were identified and followed for over a 12-month period. Main outcome measures were medication adherence (medication possession ratio, [MPR]; implementation) and persistence on index treatment. All country-specific data sets were prepared by employing a common data input model. Outcome measures were calculated for each country and pooled using random effect models.

RESULTS

In total, 39,186 new users were analyzed. In pooled data from the three countries, suboptimal implementation (MPR <80%) was 52.45% (95% CI: 33.43-70.79) for antihy-perlipidemics, 61.35% (95% CI: 52.83-69.22) for antiosteoporotics, and 30.33% (95% CI: 25.53-35.60) for oral antidiabetics. Similarly, rates of non-persistence (discontinuation) were 55.63% (95% CI: 35.24-74.29) for antihyperlipidemics, 60.24% (95% CI: 45.35-73.46) for antiosteoporotics, and 46.80% (95% CI: 36.40-57.4) for oral antidiabetics.

CONCLUSION

Medication adherence was suboptimal with >50% of older people non-adherent to antihyperlipidemics and antiosteoporotics in the three European cohorts. However, the degree of variability in adherence rates among the three countries was high. A harmonized method of data extraction and analysis across health-related database in Europe is useful to compare medication-taking behavior at a cross-country level.

Assessment of Medication Safety Using Only Dispensing Data

PURPOSE OF REVIEW

The purpose of this review is to provide an overview of the published studies that have been used to generate evidence on the safety of medicine use when only medication dispensing data are available.

RECENT FINDINGS

Medication dispensing databases are increasingly available for research on large populations, particularly in countries that provide universal coverage for medicines. These data are often used for drug utilisation studies to identify inappropriate medicine use at the population level that may be associated with known safety issues. Lack of coded diagnoses, to identify outcomes, and lack of data on confounders can limit use of these data in practice for medication safety assessment. To overcome these issues, studies have exploited the fact that symptoms of adverse effects of medications can be treated with other medications, for example antidepressants to treat depression or oxybutynin to treat urinary incontinence. The challenge of unmeasured confounding has been addressed by implementing self-controlled study designs that use within-person comparisons and provide inherent control for confounding. Prescription sequence symmetry analysis (SSA) is a within-person study design that has been demonstrated as a useful tool for safety signal generation in dispensing data.

SUMMARY

Using medicine initiation as a proxy for the development of adverse events can help to generate evidence of the safety of medicines when only medication dispensing data are available. Careful consideration, however, should be given to the sensitivity and specificity of the proxy medicine for the adverse event and potential for time-varying confounding due to trends in medicine utilisation. Data-mining approaches using dispensing data have the potential to improve safety assessments; however, the challenge of unmeasured confounding with these methods remains to be investigated.

Fixed-dose combination antihypertensives and risk of medication errors

OBJECTIVE

While fixed-dose combinations (FDC) can improve adherence, they may add complexity to the prescribing/dispensing process, potentially increasing risk of medication errors. This study aimed to determine if prescriptions for antihypertensive FDCs increase the risk of therapeutic duplication and drug-drug interactions (DDI).

METHODS

This retrospective observational study used administrative pharmacy claims data from the Irish Primary Care Reimbursement Service. Prescriptions dispensed to adults in 2015 were included if they contained an antihypertensive FDC, or the same drugs prescribed separately. The outcomes were therapeutic duplication and potentially serious DDI involving FDC drugs. Relative risk (RR) of these outcomes, adjusted for prescription and patient factors, was determined using generalised linear models with Poisson distributions and propensity score matching.

RESULTS

This study included 307 833 FDC prescriptions (67.0%) and 151 632 separate component prescriptions. Half of patients prescribed FDCs were female with a mean age of 67.1 (SD 12.5) years and, compared with separate component prescriptions, FDCs were less often coprescribed with other cardiovascular medications. Therapeutic duplication occurred in 0.8% of prescriptions, most often involving calcium channel blockers, and 10.6% contained a DDI (most often amlodipine and simvastatin). The RR of therapeutic duplication on FDC prescriptions compared with separate component prescriptions was 1.46 (95% CI 1.17 to 1.83) and the adjusted RR was 2.06 (95% CI 1.64 to 2.60). For DDIs, there was no significant difference between FDC and separate component prescriptions after confounder adjustment.

CONCLUSIONS

This study found FDCs were associated with increased risk of duplication. When considering prescribing FDCs, this safety consideration should be weighed against potential benefits.

De novo vitamin D supplement use post-diagnosis is associated with breast cancer survival

PURPOSE

Experimental laboratory data have indicated a protective effect of vitamin D on breast cancer progression, while epidemiological evidence is growing. Using pharmacy claims data, this study investigates the association between vitamin D supplement use initiated after a breast cancer diagnosis and associated mortality.

METHODS

Women aged 50-80 years with a record of invasive breast cancer were identified on the National Cancer Registry Ireland database (n = 5417). Initiation of de novo vitamin D post-diagnosis was identified from linked national prescription data (n = 2581, 49%). Multivariate Cox proportional hazards models were used to estimate adjusted HRs (95% CIs) for breast cancer-specific mortality.

RESULTS

There was a 20% reduction in breast cancer-specific mortality in de novo vitamin D users (modelled as a time-varying variable) compared to non-users (HR 0.80; 95% CI 0.64-0.99, p = 0.048) and the reduction was greater at 49% (HR 0.51; 95% CI 0.34-0.74, p < 0.001), if vitamin D was initiated soon after the breast cancer diagnosis (within 6 months).

CONCLUSIONS

In this large national breast cancer cohort, de novo vitamin D use post-diagnosis was found to be associated with a reduction in breast cancer-specific mortality. Vitamin D, therefore, has the potential as a non-toxic and inexpensive agent to improve survival in breast cancer patients. Findings support the need for RCTs exploring the effect of vitamin D supplementation on breast cancer survival.

An evaluation of prescribing trends and patterns of claims within the Preferred Drugs Initiative in Ireland (20112016): an interrupted time-series study

OBJECTIVE

To examine the impact of the Preferred Drugs Initiative (PDI), an Irish health policy aimed at enhancing evidence-based cost-effective prescribing, on prescribing trends and the cost of prescription medicines across seven medication classes.

DESIGN

Retrospective repeated cross-sectional study spanning the years 2011-2016.

SETTING

Health Service Executive Primary Care Reimbursement Service pharmacy claims data for General Medical Services (GMS) patients, approximately 40% of the Irish population.

PARTICIPANTS

Adults aged ≥18 years between 2011 and 2016 are eligible for the GMS scheme.

PRIMARY AND SECONDARY OUTCOMES

The percentage of PDI medications within each drug class per calendar quarter. Linear regression was used to model prescribing of the preferred drug within each medication group and to assess the impact of PDI guidelines and other relevant changes in prescribing practice. Savings in drug expenditure were estimated.

RESULTS

Between 2011 and 2016, around a quarter (23.59%) of all medications were for single-agent drugs licensed in the seven drug classes. There was a small increase in the percentage of PDI drugs, increasing from 4.64% of all medications in 2011 to 4.76% in 2016 (P<0.001). The percentage of preferred drugs within each drug class was significantly higher immediately following publication of the guidelines for all classes except urology, with the largest increases noted for lansoprazole (1.21%, 95% CI: 0.84% to 1.57%, P<0.001) and venlafaxine (0.71%, 95% CI: 0.15% to 1.27%, P=0.02). Trends in prescribing of the preferred drugs between PDI guidelines and the end of 2016 varied between drug classes. Total cost savings between 2013 and 2016 were estimated to be €2.7 million.

CONCLUSION

There has been a small increase in prescribing of PDI drugs in response to prescribing guidelines, with inconsistent changes observed across therapeutic classes. These findings are relevant where health services are seeking to develop more active prescribing interventions aimed at changing prescribing practice.

Benzodiazepine and Z-drug prescribing in Ireland: analysis of national prescribing trends from 2005 to 2015

AIMS

The aim of this study was to examine prescribing trends for benzodiazepines and Z-drugs to General Medical Services (GMS) patients in Ireland.

METHODS

A repeated cross-sectional analysis of the national pharmacy claims database was conducted for GMS patients aged ≥16 years from 2005 to 2015. Prescribing rates per 1000 eligible GMS population were calculated with 95% confidence intervals (CIs). Negative binomial regression was used to determine longitudinal trends and compare prescribing rates across years, gender and age groups. Duration of supply and rates of concomitant benzodiazepine and Z-drug prescribing were determined. Age (16-44, 45-64, ≥65 years) and gender trends were investigated.

RESULTS

Benzodiazepine prescribing rates decreased significantly from 225.92/1000 population (95% CI 224.94-226.89) in 2005 to 166.07/1000 population (95% CI 165.38-166.75) in 2015 (P < 0.0001). Z-drug prescribing rates increased significantly from 95.36/1000 population (95% CI 94.73-96.00) in 2005 to 109.11/1000 population (95% CI 108.56-109.67) in 2015 (P = 0.048). Approximately one-third of individuals dispensed either benzodiazepines or Z-drugs were receiving long-term prescriptions (>90 days). The proportion of those receiving >1 benzodiazepine and/or Z-drug concomitantly increased from 11.9% in 2005 to 15.3% in 2015. Benzodiazepine and Z-drug prescribing rates were highest for older women (≥65 years) throughout the study period.

CONCLUSIONS

Benzodiazepine prescribing to the GMS population in Ireland decreased significantly from 2005 to 2015, and was coupled with significant increases in Z-drug prescribing. The study shows that benzodiazepine and Z-drug prescribing is common in this population, with high proportions of individuals receiving long-term prescriptions. Targeted interventions are needed to reduce potentially inappropriate long-term prescribing and use of these medications in Ireland.