Prevalence and variation in antidepressant prescribing across Northern Ireland: a longitudinal administrative data linkage study for targeted support.

Lead Research Organisation: University of Ulster
Department Name: Sch of Psychology

Abstract

Anti-depressant prescription rates in the United Kingdom (UK) are among the highest in Western Europe and prescription rates in Northern Ireland (NI) are significantly higher than the rest of the UK. Moreover, while prescription rates are climbing annually, rates of depression are not changing, and evidence suggests that a range of socio-economic and geographical factors may be responsible. Furthermore, the majority of UK GP's often believe that alternative treatments may be more appropriate and that they would prescribe antidepressants less frequently if other options were available. Evidence seems to indicate generally that (a) there are widely different reasons why people receive anti-depressants and why GP's prescribe them, and (b) the relative importance of these reasons may vary geographically (e.g. the likelihood of receiving antidepressants may vary due to the affluence of an area or the number and proximity of available primary care services; and the associated pressures on these services).
A multi-disciplinary team of experienced academic researchers and the award winning NI charity, Aware Defeat Depression (Aware), aim to maximise the use of Census and Enhanced Prescribing Database data in NI to target, tailor and implement support for individuals in receipt of antidepressant prescriptions and to stimulate 'social prescribing' across the region (a mechanism for linking patients with non-medical sources of support within the community).
Specifically the study aims to (i) provide a detailed geographical breakdown of antidepressant prescribing in NI to identify prescription 'hot spots' at a small geographical level, (ii) use linked census data to identify demographic variables (personal, social, and economic factors) that differentiate 'hot spots', (iii) determine if the association between demographic variables and 'hot spots' differs by geographical location, (iv) determine if long-term changes in people's lives (e.g. change in deprivation, family structure, physical health) are associated with antidepressant use and (v) stimulate 'social prescribing' across the region through targeted intervention.
The project will address two main research questions highlighted in the 2015/16 ESRC strategic priority area 'Influencing Behaviour and Informing Intervention'.
On the basis of the proposed analytic strategy the project team aims to develop a nationwide map detailing the socio-economic landscape of antidepressant prescribing in NI. This map will support the non-academic partners (Aware) in their efforts to develop and implement effective and coordinated intervention programs by (i) highlighting prescription hotspots at a small geographical level (ii) generating 'risk' profiles specific to identified hotspots using personal, social and economic data and (iii) identifying 'vulnerability' at a personal, social and economic level through longitudinal change modelling.
The project has the potential to deliver effective and meaningful change. Partnership with Aware will ensure that findings can be acted upon immediately. Aware has been delivering intensive education and training programmes to thousands of adolescents and adults across NI since 1996 and has been awarded the GSK IMPACT Award in 2014. A primary focus of these programmes is to educate individuals about positive strategies that can be used in the recovery from depression. Aware, on the basis of the project findings, will tailor, target and promote 'social prescribing' in those areas and among those individuals in NI that need it most. Social prescribing may include opportunities for arts and creativity, physical activity, learning new skills, volunteering, mutual aid, befriending and self-help, as well as support with, e.g. employment, housing, debt, or parenting problems. Equipped with a detailed geographical map the project team will develop a targeted intervention strategy for delivering focussed, context sensitive social prescribing across NI.

Planned Impact

Who will benefit from this research? The project team are committed to engaging with key stakeholders in the area of community action, and at a public health, primary care, and national policy level. While the NI client base of the projects non-academic partner, Aware Defeat Depression, will be the primary beneficiaries of the research, Aware themselves and other primary care service providers will also directly benefit from the work. The proposed project will also be of interest to a wide range of academic disciplines (public health researchers, epidemiologists etc).

How will they benefit? Aware Defeat Depression (Members & Client Base): Partnership with Aware will ensure that the analytic process and findings will directly inform targeted intervention. Aware, the only charity working exclusively for those with depression in NI, is an established support network ideally positioned to deliver impact. Aware has been delivering intensive education and training programmes to thousands of adolescents and adults across NI since 1996 and has been awarded the GSK IMPACT Award in 2014. A primary focus of these programmes is to educate individuals about positive strategies that can be used in the recovery from depression. Aware, on the basis of the project findings, will be in a position to better tailor, target and promote 'social prescribing' in those areas and among those individuals in NI that need it most.
NI Primary Care Service Providers, GPs and their patients: While Aware can communicate directly with those in receipt of antidepressant medications across NI, primary care service providers and general practitioners constitute another important stakeholder. Through the project, Aware can offer a valuable opportunity to alleviate pressures in primary care (and associated prescription practices) in NI and to link GP patients with a wide range of non-medical sources of support within the community that may address their needs in more effective and diverse ways. In relation to the project report findings specifically, primary care health workers and GPs will be directly informed of (i) variations in prescription rates across NI and (ii) the socio-economic context of prescribing rates in their jurisdiction.

What will be done to ensure that they have the opportunity to benefit?
-A project website and social media account will be created. Specific consideration will be given to non-academic bodies in the design of the website. For maximum impact and utility an interactive map based on GIS principles will be designed. This application will provide users with an efficient means of navigating the project findings geographically.
-A targeted intervention strategy for delivering focussed, context sensitive social prescribing across NI will be formulated by the project team. Using the project findings Aware will target services to 'high-risk' areas across NI.
-A summary report and pathways to policy document will be compiled and sent to the ADRN, the Office of First Minister and Deputy First Minister in NI and the Royal College of General Practitioners Northern Ireland.
-On completion of the analyses the team will organise a briefing conference to ensure that project findings are effectively communicated to all stakeholders. The conference will be promoted and advertised through ARK (based at Ulster University). The primary aim of ARK is to support policy development which it achieves through its Policy Unit which has established connections with the voluntary sector, the Assembly, government departments, practitioners and academics. Importantly, this event will bring Aware and primary care service providers together to ensure that social prescribing is promoted and facilitated as effectively as possible throughout NI.
-Finally, two peer-reviewed publications will be produced, the first targeting GPs (e.g. British Journal of General Practice) and the other, broader social scientists (e.g. Social Science and Medicine).

Publications

10 25 50
 
Description Through the ADRC-NI we have linked the Northern ireland Census (2011) to the Northern Ireland Enhanced Prescribing Database (EPD). All incidences of prescriptions of antidepressants (BNF: 4.3.1: Tricyclic & Related Antidepressant Drugs, 4.3.2: Monoamine-Oxidase Inhibitors (Maois), 4.3.3: Selective Serotonin Re-Uptake Inhibitors
4.3.4: Other Antidepressant Drugs) for years 2100 2015 (inc) were identified and linked to the census data. To date we have estimated the prevalence of antidepressant prescribing for each of the 5 years. Descriptive cross-tabulations of prescribing records and socio-demographic variables from the census have been produced. For each year from 2011 to 2015 the proportions of the population in receipt of anti-depressant prescriptions was 12.45% (95%CI=12.40-12.50), 13.08% (95%CI=13.03-13.13), 13.67% (95%CI=13.62-13.72), 14.23% (95%CI=14.18-14.29), 14.93% (95%CI=14.87-14.98), and over the 5-year period the prevalence was 24.12% (95%CI=24.05-24.19).

The 5-year anti-depressant prescribing rates were used as predictors in bivariate and multivariate regression models, with census variables as predictors. The ORs from the multivariate model indicate that being prescribed an anti-depressant can be described in terms of a risk profile. In order of increasing risk the variables are
1. Self-reported mental health condition (OR=3.37).
2. Self-reported general health described as 'Fair' (OR=2.39), 'Bad' (OR=3.13), or 'Very Bad' (OR=3.22).
3. Ethnicity reported as 'White' (OR=2.14).
4. Female gender (OR=2.08).
5. Disability reported as 'Impaired a little' (OR=1.28) and 'Impaired a lot' (OR=1.34).
6. Lower rateable property value (ORs = 1.11 - 1.63).
7. Lone parent family (OR=1.17).
8. Non full-time or retired economic activity (ORs = 1.19 - 1.31).
9. Non-urban residence (ORs = 1.08 - 1.16).

Some variables were found to be protective; age 40-54 years (OR=.94), not being in a family (OR=0.95), being Catholic (OR=0.91) or No faith (OR=.95), and being a full-time (OR=0.53) or part-time student (OR=0.59).

This project was the first ESRC funded project to use the ADRC-NI infrastructure. This project acted as an 'ice-breaker' as during the course of the project new relationships and processes were implemented and evolved in order to acquire the data, gain access, and use the secure environment. The project was successful in raising awareness of the ADRC-NI, what research resources it can offer, and the processes involved in using the centre.

The process of data acquisition has been problematic at time. Data access was delayed considerably and there still ongoing delays in accessing the de-identified data at the SOA level.

This project has strengthened the association between the research team and the Institute of Social Research in Copenhagen. On the 27th November the research team (Mark Shevlin, PI; Jamie Murphy, CI; Daniel Boduszek, ECR) visited the Danish National Centre for Social Research (SFI). The SFI is responsible for conducting commissioned research in the area of welfare state policies and disseminates the results through the website, publishing reports, working papers, a quarterly magazine, and press releases.
We met with Dr Mogens Christoffersen who is an emeritus Senior Research Fellow at the centre. We originally met Dr Christoffersen while working on a Danish registry project with Professor Ask Elklit from the Psycholotraumatology Research Centre at the Southern University of Denmark. The meeting with Dr Christoffersen was incredibly helpful as he has a lot of experience in dealing with administrative data. He has worked with the Danish National Prescription Registry. He described many of the potential pitfalls in coding prescription data and offered insightful and helpful suggestions. The geographical mapping aspect of our study was something he was particularly interested in and has put us in contact with his colleague Dr Vibeke Jakobsen who has developed some interesting ways to construct 'neighbourhoods' based on geo-referenced data and administrative data. Dr Christoffersen attended the June 2018 International Conference for Administrative Data Research. The project team all attended a meeting at the Annual Administrative Data Research conference (June, 2018, Belfast) to review the project outcomes and plan for future analysis, publications, and dissemination.

The project developed research capacity. Daniel Boduszek was the project ECR. As part of the project Daniel was involved in all aspects of the design of the project, completed all necessary training and accreditation, and was instrumental in the production of all outputs. Most importantly has initiated his own research programme based on administrative data research. Daniel organized a workshop on the 5th of March 2018 at the headquarters of The Danish National Centre for Social Research (SFI) in Copenhagen. The purpose of the workshop was to establish and develop research collaboration between the SFI and Huddersfield University in the area of forensic and criminal psychology using administrative data. The workshop was chaired by Torben Tranæs, the Executive Director of Research and Professor at SFI, and was attended by Dr Christoffersen and six senior researchers. Daniel and four researchers from Huddersfield University attended. They reviewed papers from the SFI team on how registry data can be used to (1) conduct natural experiments to evaluate policy reforms, (2) statistically model neighborhood effects on youth crime, and (3) understand perceived safety in socially vulnerable neighbourhoods. Subsequent to this workshop Daniel Boduszek has initiated a number of collaborative projects. The aim is to initially use the Danish National Registry System to test competing models of recidivism and then develop a proposal to replicate the studies through the ADRC-NI.
Exploitation Route We predict that other third sector organisations and researchers will be interested in the use of linked data.
Sectors Healthcare

URL https://www.ulster.ac.uk/research/institutes/psychology/research/current-projects/-prevalence-and-variation-in-antidepressant-prescribing-across-northern-ireland
 
Description We have provided preliminary results to our non-academic partner, Tom McEneaney from AWARE DEFEAT DEPRESSION. The project team discussed the findings and Tom has initiated an AWARE team meeting to discuss the implications. On 6th March we held a dissemination and round table discussion event. This was attended by representatives from the Royal College of Psychiatrists, Department of Public Health, AWARE, as well as mental health practitioners and researchers. Discussions indicated that the project findings will be used to inform the delivery of mental health services in Northern Ireland. On 22nd January 2019 we presented the project findings to the Senior Management Team of AWARE, our non-academic partner. The implications of the results and how they could be used by AWARE to enhance their service provision was discussed. In particular, AWARE were interested in how the information on anti-depressant prescribing at small geographical areas could be most effectively used to target their services. We now have a MSc Mental Health student working on this as part of their associated research project. On the 7th February 2019 the paper "Rates and predictors of anti-depressant prescribing in Northern Ireland 2011 - 2015: A data linkage study using the Administrative Data Research Centre (NI)." was accepted for publication in the Irish Journal of Psychological Medicine. A copy of the paper has been deposited in UU's Institutional Repository and we expect significant press interest once published. The Winter 2018 Issue (#33) of Society Now, the ESRC publication, led with an article on the project entitled "Lack of alternatives increases anti-depressant prescribing". This article described the innovative methods used in the project and summarized the main project findings.
Sector Healthcare
Impact Types Societal

 
Description Informed the delivery of support by AWARE
Geographic Reach Local/Municipal/Regional 
Policy Influence Type Influenced training of practitioners or researchers
 
Title Census and BSO aggregation tools 
Description Dr Michael Rosato has developed a series of SAS macros that can be used to quickly, efficiently, and reliably link and recode the raw data from cenus and BSO. The raw data is provided in 'csv' format and the macros recode and label all the relevant variables (e.g. household size and settlement band variables are derived from separate fields in the census data). A macro for the analysis of prescribing data was also developed. This reads the data from the enhanced prescribing database and transforms it into quarterly and annual prevalence rates, and also defined daily dose estimates. 
Type Of Material Improvements to research infrastructure 
Year Produced 2018 
Provided To Others? Yes  
Impact This improves the efficiency and standardizes the coding of data from the ADRC-NI. 
 
Title Combined Census 2011 (Northern Ireland) and Enhanced Prescribing Database 
Description This new database was constructed by utilising linked administrative data in a process facilitated by the Administrative Data Research Centre in NI (ADRC-NI). The dataset was derived from the 2011 NI Census and the Enhanced Prescribing Database. These two datasets, death records for the study period (2011-2015), and healthcare registration information, were linked using anonymous one-way encryption methods by officers at the the Northern Ireland Statistics and Research Agency. Successful linkage of 92.58% of the 2011 NI Census cases with the Enhanced Prescribing Database was achieved, resulting in a sample of 1,588,355 individuals at the time of the 2011 NI Census. Data linkage methods utilised were based on the Northern Ireland Longitudinal Survey, a precursor to the ADRC-NI. The anonymous data was then made available to the research team in the secure research environment. Strict disclosure control procedures were adhered to within the secure environment. 
Type Of Material Database/Collection of data 
Year Produced 2018 
Provided To Others? No  
Impact Our non-academic partner, AWARE, have been using the data to tailor the delivery of their services. The issue of anti-depressant prescribing and alternative social prescribing options were identified in a local press story. The ECR and RA on the project are now skilled in accessing and analyzing administrative data through the ADRC-NI. 
 
Description The Danish National Centre for Social Research 
Organisation Danish National Centre for Social Research (SFI)
Country Denmark 
Sector Public 
PI Contribution We have developed and extended the research links with SFI.
Collaborator Contribution On the 27th November the research team (Mark Shevlin, PI; Jamie Murphy, CI; Daniel Boduszek, ECR) visited the Danish National Centre for Social Research (SFI). The SFI is responsible for conducting commissioned research in the area of welfare state policies and disseminates the results through the website, publishing reports, working papers, a quarterly magazine, and press releases. We met with Dr Mogens Christoffersen who is an emeritus Senior Research Fellow at the centre. We originally met Dr Christoffersen while working on a Danish registry project with Professor Ask Elklit from the Psycholotraumatology Research Centre at the Southern University of Denmark. The meeting with Dr Christoffersen was incredibly helpful as he has a lot of experience in dealing with administrative data. He has worked with the Danish National Prescription Registry. He described many of the potential pitfalls in coding prescription data and offered insightful and helpful suggestions. The geographical mapping aspect of our study was something he was particularly interested in and has put us in contact with his colleague Dr Vibeke Jakobsen who has developed some interesting ways to construct 'neighbourhoods' based on geo-referenced data and administrative data.
Impact Closer collaboration with SFI. Daniel Boduszek has already developed research links with SFI researchers in the area of criminology research using administrative linked data. Mogens Christoffersen presented "Is the Prevention Paradox Useful in Suicide Prevention Theory and Strategies?" at the International Conference for Administrative Data Research, Belfast, UK, 21-22 June 2018
Start Year 2016
 
Description ADRC-NI Research Away Day - Thursday 28 January 2016 - The Old Inn, Crawfordsburn 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact The away day was attended by ADRC-NI members, academics/researchers, NISRA staff, third sector representatives, and representatives from the local Trusts. The away day involved a review of network activity by Dr Dermot O'Reilly & Dr Frances Burns and presentations from Orla Bateson (Head of Research Support Unit, RSU) on data accessibility, Dr David Wright (RSU) described the "User Journey", and Dr Frances Burns discussed "Researcher Engagement/Training and Development". The project team gave a presentation on the project. Discussions followed with representatives from the Public Health Agency, Northern Ireland and members of the Research Support Unit from NISRA.
Year(s) Of Engagement Activity 2016
 
Description ADRC-NI Researcher Away Day - Thursday 23 March 2017 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact The annual conference for ADRC-NI reviewed all current and future projects. Mark Shevlin presented "Prevalence and variation in antidepressant prescribing across Northern Ireland: a longitudinal administrative data linkage study for targeted support".
Year(s) Of Engagement Activity 2017
 
Description ADRN Conference 2016 (London, 3 June) 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact Prof Shevlin delivered a presentation on data linkage and mental health research
Year(s) Of Engagement Activity 2016
URL https://www.adrn.ac.uk/conference/2016
 
Description Presentation at 15th Annual Psychology, Health & Medicine Conference (June1st 2018) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact Stephanie Boyle (project RA), on behalf of the project team, delivered the presentation "Utilisation of Adminstrative Data in Health Research: The Antidepressant Prescribing Project as an Example of Practice". The primary aim of this presentation was to promote the ADRC-NI as a powerful and flexible research resource for researchers in the area of health psychology in Northern Ireland. The presentation was well-attended and questions were mostly about the availability of different types of data through the ADRC-NI.
Year(s) Of Engagement Activity 2018
 
Description Press Interview 
Form Of Engagement Activity A press release, press conference or response to a media enquiry/interview
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact The findings from the roundtable project briefing event were picked up by local press (Monday 19th of March 2018). Mark Shevlin was interviewed and a brief piece was published online. Local politicians (Sinn Fein's Kevin Campbell) also contributed to the piece. This raised awareness of the issue of antidepressant prescribing. It was also picked up by Oliver Newlan who was producing a Radio 4 programme looking at rates of antidepressant prescription to children. We discussed the project and I provided him with relevant information on prescribing rates for young people.
Year(s) Of Engagement Activity 2018
URL https://www.derrynow.com/news/new-magee-research-backs-non-medical-interventions-depression/213776
 
Description Project Blog: Antidepressant Social Prescribing Project 
Form Of Engagement Activity Engagement focused website, blog or social media channel
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact In March 2017 the project bog went live. The blog was designed by the team to help raise awareness of the project and discuss associated research and news events. The blog also includes a "Monthly Must Read" section that highlights a topical research paper that has just been published and an "Ask the Expert" section where influential researchers and practitioners provide their views on the use and effectiveness of antidepressants. One contributor, Professor John Read currently a Professor of Clinical Psychology at the University of East London, worked for nearly 20 years as a Clinical Psychologist and manager of mental health services in the UK and the USA. Professor read has published findings from the largest survey to date examining the emotional and interpersonal effects of antidepressants**. We also have contributions from the project team members and statisticians from the Northern Ireland Statistics and Research Agency who are responsible for the data linkage and access.

*Read, J., Cartwright, C., & Gibson, K. (2014). Adverse emotional and interpersonal effects reported by 1829 New Zealanders while taking antidepressants. Psychiatry research, 216(1), 67-73.
Year(s) Of Engagement Activity 2017,2018
URL https://adsopp.wordpress.com/
 
Description Roundtable Project Briefing Event (Ulster University, March 6th 2018) 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact On March 6th we were pleased to release our initial results to key stakeholders in a roundtable project briefing event at Ulster University. In collaboration with Aware NI, we presented our results to representatives from the Royal College of Psychiatrists (RCP), the Public Health Agency (PHA), the Belfast Health and Social Care Trust, and academics from Ulster University and Queens University Belfast working in the areas of antidepressant use and social prescribing. The event began with Professor Mark Shevlin introducing the Administrative Data Research Network, and advocating for administrative data research. Then our headline figures were presented and discussed. Tom McEneany followed, discussing the potential for impact. Tom is our primary contact at Aware NI.

There were extensive post-presentation discussions-
Representatives from the RCP indicated that the findings were useful in providing accurate prevalence estimates for different age groups. They also stated that the information on risk factors was already well-established in the extant research literature, although it was important and useful to have regional specific information. There followed an interesting discussion on how the ADRC-NI could be used to inform the location of the new "Mental Health Hubs" based on need, and how future data sweeps could provide evidence of the efficacy of the hubs.

Representatives from the PHA were interested in how the BSO data could be disaggregated, stratifying by type of antidepressant (SSRIs, SNRIs, etc). They were interested to understand the reasons for prescribing different types, and we discussed the limitations of administrative data in this respect. Overall, they reported that the findings were useful for considering age specific mental health needs in Northern Ireland.

Tom McEneany (AWARE) described how information on the prevalence rates and predictors of antidepressant was helping AWARE target interventions. The mapping of prevalence rates was the main outcome of the project that he considered would radically help AWARE plan and deliver services as efficiently as possible.
Year(s) Of Engagement Activity 2018
 
Description Two conference presentations- International Conference for Administrative Data Research (21-22nd June 2018) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact Two conference presentations were made at the International Conference for Administrative Data Research, Queen's University Belfast, 21-22nd June 2018.
1. Mogens Christoffersen, "Is the Prevention Paradox Useful in Suicide Prevention Theory and Strategies?"
2. Mark Shevlin, "Prevalence and predictors of antidepressant prescribing in Northern Ireland"

Dr Christoffersen's presentation described the extensive Danish registries and outlined how they can be used to test complex social science hypotheses. There was an extensive discussion about how registry data is accessed and how the associated privacy concerns were addressed. He also met with post-graduate research students and discussed the possibility of collaborative research.
The discussion after Mark Shevlin's presentation was around what other data resources could be used through the ADRC-NI. These were described and those who were interested were directed to the ADRN/ADRC-NI website.
Year(s) Of Engagement Activity 2018
URL https://www.ulster.ac.uk/research/institutes/psychology/research/current-projects/-prevalence-and-va...
 
Description VCSE Data workshop: Mental Health (14 March 2017) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact The ADRC-NI hosts a series of data research workshop in partnership with DetailData, an open data project facilitated by NICVA and TheDetail.tv, the aim of which is for VCSE sector groups to learn more about both open and administrative data and how it can be used in their work, and for researchers to make connections with VCSE groups doing work at grassroots level on similar topics to their research. Each workshop has a theme; the first pilot workshop focused specifically on carers and caring. The workshop that was held on Thursday 11 May at NICVA, focused on mental health data and research. The main aim was to discuss and showcase the potential of the data sets accessible through the ADRC-NI to answer quantitative research question on relevant topics, and help VCSE researchers better understand how the available data could be used to develop a quantitative research project that would enhance their existing work and expertise. During this workshop the research team presented a detailed overview of our project.
Year(s) Of Engagement Activity 2017
URL https://www.adrn.ac.uk/public-engagement/activities-and-events/vcse-workshops/