27th January 2017, Volume 130 Number 1449

Prasad Nishtala, Henry Ndukwe, Te-yuan Chyou, Mohammed Salahudeen, Sujita Narayan

Pharmacoepidemiology is an evolving area of research and was recognised as a distinct discipline in the early to mid-1980’s. An interface between clinical epidemiology and pharmacology, pharmacoepidemiology benefits from methodology derived through epidemiology to examine drug effects in populations. Pharmacoepidemiology encompasses themes including pharmacovigilance, drug policy, post-marketing assessment of effectiveness and safety of medicines, cost-effectiveness and rationale use of medicines among other themes of research.1

There is growing interest in pharmacoepidemiology because randomised controlled trials (RCTs) used for registration purposes are typically underpowered to reveal safety signals for rare events. Furthermore, for operational or ethical reasons, it may not be feasible to conduct RCTs in special populations such as the very old and the very young, and for a duration long enough to capture all safety end points. Advances in pharmacoepidemiology have not only overcome the boundaries of generalisability but also extended the time for follow-up to study rare adverse events.1

History of data collections and pharmacoepidemiology in New Zealand

Prior to national electronic data collections and registries in New Zealand, there was little pooled health data available on a national scale to conduct pharmacoepidemiological studies.

Before the Second World War, the Department of Statistics was accountable for compiling mortality data. The information ascertained in death registration forms (known as BDM28) was used by government for analysis and planning. In 1948 the responsibility of registering deaths was transferred to the Ministry of Health (MoH) and the mortality collection was later assigned a National Health Index (NHI) by MoH in 1988.2

Registries such as the New Zealand Cancer Registry (NZCR) is managed by New Zealand Health Information System (NZHIS) and achieved a nationwide success of over 90% coverage in 1972.2 Health Benefits Limited (HBL), through the MoH, managed procurement and supply-chain data before the introduction of Pharmaceutical Management Agency Limited (PHARMAC).3

Roberts and Norris published findings based on the HBL data, which showed variations and changes in antidepressants dispensing in New Zealand between 1993 and 1997.4 Subsequently, Nishtala and colleagues have published several drug utilisation studies analysing the pharmaceutical collections (Pharms) supplied by PHARMAC.5

Pharmacovigilance in New Zealand

Pharmacovigilance is a unique area of pharmacoepidemiology concerning monitoring and reporting of adverse effects of medicines and therapeutic devices. The New Zealand Pharmacovigilance Centre (NZPhvC) is the epicentre for monitoring adverse reactions in New Zealand. The NZPhvC currently operates two main programmes: (1) Centre for Adverse Reactions Monitoring (CARM) and (2) Intensive Vaccine Monitoring Programme (IVMP).

The Intensive Medicines Monitoring Programme (IMMP) based in Dunedin was a branch of the CARM from 1977 until 2013.6 IMMP established cohorts from prescription data, collected event information and developed a unique methodology known as Prescription Event Monitoring (PEM). PEM cohorts contained information, supplied by pharmacies, from prescriptions for medicines that have been selected for monitoring. MoH superintended IMMP coordination through CARM, however, some dispensing data entries were recorded retrospectively and affected collations as pharmacists were provided with handwritten forms to enter information on dispensed prescriptions, and these were to be returned through pre-paid postage envelopes.6 IMMP made several significant contributions to pharmacovigilance in New Zealand, performing in-depth epidemiological investigations relating to the safe use of medicines and introduced programmes such as the Medication Error Reporting Programme (MERP).7

New Zealand’s national collections

Several national collections held by governmental organisations provide significant epidemiological data useful for population studies. Big data coverage in New Zealand comprises a nationwide collection of prescription claims or reimbursement data, industry-based and wholesale or supplier warehouse databases. Data holdings can also be medical, hospital or administrative records and events or disease-based registers. Table 1 displays a list of New Zealand databases used in population-based research.

Table 1: National collections and registers relevant to pharmacoepidemiology research.

Name

Alternative nomenclature

Brief description

National claims and collections

Mortality collection

Mortality database

Confirmed deaths managed by the New Zealand Ministry of Health (MoH) and was first assigned National Health Index (NHI) in 1988.

New Zealand Health Information System data warehouse

NZHIS

Maintains datasets including Mortality data, NMDS, NZCR.

National Minimum Dataset

NMDS

NMDS is a national collection of public and private hospital discharge information, including coded clinical data for inpatients and day patients.

New Zealand Cancer Registry

NZCR

NZCR provides data for cancer incidence and survival studies, public health research, monitoring screening programmes and policy formulation.

General Medical Subsidy Claims

GMS

GMS claim holds information on subsidies paid to reduce consultation cost with general practitioners for children under 15 years of age or subsidy/community card holders.

Medicines Monitoring database

M2

Medicines and safety data held by New Zealand Medicines and Medical Devices Safety Authority (Medsafe).

Pharmaceutical Claims Collection (Pharmaceutical Management Agency (PHARMAC) collects and manages data through MoH)

Pharms Datamart

Holds dispensing data on medicines claims for prescribed dispensings that are subsidised by PHARMAC.

Primary Health Organisation enrolment database

PHO

Holds individuals’ data for primary healthcare (Primary Care visits).

Laboratory Collection

Labs

Holds data on laboratory test carried out in individuals. Labs data reporting was enforced from July 1994.

National Non-Admitted Patient Collection

NNAP

Maintains data on non-institutionalised cases in healthcare facilities.

National Maternity Collection

NMC

Holds data on maternal deliveries.

Programme for the Integration Of Mental Health Datamart

PRIMHD

Holds data on mental health status of those accessing healthcare.

Intensive Medicines Monitoring Programme (Part of New Zealand Pharmacovigilance Centre (NZPhvC) or Programme with Centre for Adverse Reaction Monitoring (CARM) including Prescription Event Monitoring)

IMMP database

CARM held dispensing data for signal detection for more than three decades (1977–2013).

Client Claims Processing System

CCPS

The CCPS holds data which accounts for disability support events in community dwelling older people.

Events and disease-based registers

Births, Deaths and Marriages Register

BDM

Held by MoH.

Virtual Diabetes Register

VDR

Specific algorithm is used to extract VDR data from NMDS database.

National Immunisation Register

NIR

Holds individual level data on immunisation.

National Cervical Screening Programme Register

NCSP

Cervical cancer screening database which holds data and some clinical information on females signed up to the Programme.

Australian & New Zealand Hip Fracture Register

Hip Fracture Register

Piloted in New Zealand.

Electronic medical (health) records

Hospital Medical Records

HMR

A keep of medical records including those prescribed for hospitalised patients.

Medication Error Reporting Programme (MERP)

MERP database

Online reporting system maintained by NZPhvC which holds core information on the medicine, medication error and detail characteristics on factors that precipitated the recorded outcome.

Administrative databases

Aotearoa New Zealand Health Tracker

ANZHT

Administrative data used in research work to estimate prevalence of common public health diseases like gout in population studies.

EpiSurv records (maintained by the Institute of Environmental Science and Research)

ESR

Mandatory reporting of notifiable diseases that have been reported to Medical Officers in the healthcare sector.

New Zealand Deprivation (NZDep) dataset

NZDep index

NZDep index is a composite deprivation score, extracted independently from information contained in census data, which compares socioeconomic positions of small area mesh-blocks between 60–110 people based on their location within a given time frame. Score of 1 (least deprived) to 10 (most deprived).8 

Methodological considerations

Pharmacoepidemiologcal studies are observational in nature, and hence suffer from confounding and bias. In the last few decades, advances in pharmacoepidemiology has encouraged novel methods of adjusting for confounding and bias.9 In New Zealand, studies have progressed from examining trends in medicine utilisation to using apposite statistical models to examine health outcomes such as mortality, hospitalisation and use of primary care services. Access to advanced statistical software has encouraged use of predictive models and propensity score matching or inverse propensity score weighting, which to some extent mimics the process of randomisation in retrospective cohort studies.

In most observational studies, confounding variables typically include socioeconomic variables, time independent (sex, ethnicity etc.) and time varying cofounders (frailty etc.) and multi-morbidity.10 Propensity scores (PS) are now increasingly being used to ensure distribution of confounding variables is the same in the exposed (treated) and the unexposed group (untreated). In propensity weighting, weights are applied to the exposed (1/PS) and the unexposed (1/(1-PS)) so that the overall distribution of the PS scores is the same in both groups. Individuals with a high PS score have covariates that favour the use of the treatment, and lower PS scores suggest that individuals are drawn away from the treatment. Therefore, in order to achieve the covariate-balancing, the PS-weighting method is designed to weight down data from over-represented individuals and vice versa for data from under-represented individuals.

Pharmacoepidemiological studies in New Zealand

Drug safety studies

Several drug safety studies have been conducted using the routinely collected national data in New Zealand, and have been able to associate medicine used with health outcomes at a population level.

Winnard et al utilised the Aotearoa New Zealand Health Tracker (ANZHT) population data to examine the overall prevalence of gout, and identified a high prevalence in the entire Aotearoa New Zealand population, particularly among Māori and Pacific people.11 Similarly, a high co-prevalence of gout, diabetes and cardiovascular disease in the same population was identified in a later study.12 A nationwide nested case-control study by Blank et al used data linkage over four years and identified that the use of a proton pump inhibitor was associated with a significantly increased risk of acute interstitial nephritis.13 Another nested case-control study using data linkage investigated the association of statin and risk of rhabdomyolysis and highlighted that the absolute risk of rhabdomyolysis in a general population of simvastatin users is very low.14 In 2009, Mehta et al examined the optimal clopidogrel therapy with the help of national health databases and concluded that clopidogrel coverage following percutaneous coronary intervention improved by 7% across sociodemographic groups after funding restrictions ceased.15

The Research in Pharmacoepidemiology (RiPE) group,16 has published copious studies covering drug safety using Pharms dataset and other datasets such as the NMDS, Primary Care visits, Mortality collection and NZDep index for older individuals aged 65 years and above. Among these were some landmark studies, which examined the impact of Drug Burden Index (DBI), a measure of exposure to anticholinergic and sedative medications, on geriatric adverse drug events including falls, and another study which compared the haemorrhage-risk associated with anticoagulants in older New Zealanders on a propensity score matched, population-level cohort.17,18 The later study concluded that the risk of any haemorrhage and intracerebral haemorrhage was lower in dabigatran users compared to warfarin users.18

Drug utilisation studies using national health collections data

Literature search in Ovid MEDLINE and EMBASE identified 19 studies that have recently been published on pharmacoepidemiology using national health collections data. These studies are listed in Supplementary Table 1. In addition to these, several other studies have examined utilisation patterns of various medicines, which can potentially guide future policy directions.

In 2005/6 Norris et al conducted a cross-sectional study using dispensing data linked with national datasets to investigate equity in statin utilisation among different ethnic groups and found that in contrast to results from other previous studies, statin use approximately matched the pattern of need.19 Another study on health services utilisation by ethnicity variations was conducted by Wheeler et al, using the NZHIS via the Patient Information Management System (PiMS also known as iPM).20 This study highlighted the variations in psychiatric inpatient representation, diagnosis and compulsory treatment across ethnic groups and socioeconomic status. Using toxicological and coronial databases between 2001 and 2005, Gallagher et al concluded that restricted access to work-related chemicals and stricter prescription/dispensing controls for antidepressants such as tricyclic antidepressants may reduce self-poisoning in New Zealand.21

In 2001, two studies reported the use of HBL data supplied by PHARMAC on subsidised dispensing of antidepressants between 1993 and 1997, and examined the regional differences4 and changes22 in the prescribing of antidepressants in New Zealand. The first study concluded that regional differences in anti-depressant prescribing are large, and later highlighted that the use of newer agents was contributing to increased overall use of anti-depressant medication and government expenditure in New Zealand. The earliest direct national estimates of diagnosed congestive heart disease prevalence (2008), by sociodemographic status was undertaken in New Zealand in 2011.23 Following this, a recent study examined the current epidemiological trends of leprosy in New Zealand for a 10-year period (2004 to 2013) to raise awareness among the health professional community and used the EpiSurv national surveillance database, administered by the Institute of Environmental Science and Research (ESR).24

Registry-based studies

Population-level registry-based studies are a newer approach to pharmacoepidemiological studies in New Zealand, in particular the Cancer Registry. The incidence trends for 18 adult cancers, by ethnicity and socioeconomic status was examined and found disparities between the different cancers among ethnic groups.25 Another recent study examined the prevalence of statin discontinuation in individuals aged 75 years and older with a diagnosis of cancer, in the last year of life.10 This study revealed that statins are more likely to be discontinued in older individuals with a diagnosis of cancer compared to those without cancer. This association hints at a plausible trend that having cancer is a more legitimate reason for discontinuing statins compared to other life-limiting conditions, knowing that the end-of-life is more predictable for individuals with cancer than individuals with other life-threatening diagnoses.

The National Cervical Screening Programme Register (NCSP-R) was used to identify women aged 20–69 years of age with an index high grade cytology report from 2009–2011, investigating pre-vaccination benchmark prevalence for Māori and non-Māori women, and identified long-term effects of vaccination to be similar in the two groups.26 Additionally, Grant et al identified associations between parental intentions and the subsequent timeliness of infant immunisation using the National Immunisation Register.27

The Virtual Diabetes Register was used to compare the prevalence of diagnosed and undiagnosed diabetes and prediabetes in New Zealand. In this study, the authors have summarised and compared regional and national diabetes prevalence surveys undertaken in New Zealand since the late 1960s.28

interRAI (international Resident Assessment Instrument)

From 1st July 2015, it has been mandatory that all residential care facilities start using the suite of health and support needs assessment tools known as interRAI, in particular the interRAI Long-Term Care Facilities Assessment System (LTCF).29 It targets people over the age of 65 years who require needs assessment for access to long-term publicly funded services. A recent study highlighted the user experience of interRAI assessment tools in New Zealand and explored clinicians’ perceptions and experience of using interRAI electronic assessment tools, highlighting some barriers that need to be overcome prior to adopting the tool.

Challenges and opportunities in New Zealand for pharmacoepidemiological research

Challenges

In 2006, Gauld & Goldfinch highlighted noteworthy limitations with the New Zealand health data collections. Concerns were raised regarding the standardisation of data collection systems, difficulties with information exchange and insufficient governance of information maintaining systems.30

While population-based datasets eliminate cost and time constraints, completeness and validity of the datasets are of a global concern. The New Zealand administrative databases are frequently utilised, and their accuracy and completeness has been tested in few observational studies.

Opportunities

The New Zealand health sector holds and maintains rich observational health care data that are valuable for evaluation of health outcomes. Health care data describe large patient populations and provide promising research opportunities for proactive, longitudinal surveillance of the safety and effectiveness of medicines. New Zealand can make major contributions to the knowledge base of medicines use and health outcomes in real-world settings. Appropriate quality and relevant data for monitoring health-related outcomes will enable opportunities for pharmacoepidemiological research and developments in policies, services, practices and quality of life for all individuals accessing healthcare in New Zealand.

Conclusions

Pharmacoepidemiology is an evolving scientific discipline that offers a number of advantages to evaluate medication safety in a large population for a long period of follow-up, with good generalisability. New Zealand is well-positioned globally to make significant contributions to the knowledge base of outcomes on the quality of medicine use in real-world settings. There is an increased recognition by various stakeholders, including Medsafe and PHARMAC, that a rapid increase in the quantity, diversity and accessibility of patient data will provide unprecedented opportunities for rapid assessment of drug safety, monitoring pharmaceutical policies, examining healthcare utilisation and decision-making relevant to all individuals accessing healthcare in New Zealand. This viewpoint article highlights the scope for methodological advances for performing pharmacoepidemiological studies in New Zealand. In conclusion, the future of pharmacoepidemiology in New Zealand requires an integrated development of models to capture outcome data, particularly on safety endpoints, and the integration of statisticians and epidemiologists on the need to improve designs and methodology in this area of study.

Summary

Pharmacoepidemiology is a combination of the sciences involved in epidemiology and pharmacology. In New Zealand there have been recent advances in pharmacoepidemiology to examine medicine use and observe adverse drug events for individuals who are often not included in trials such as older people. This paper attempts to describe the past, present and future of pharmacoepidemiology in New Zealand, with a particular focus on medicine use and safety.

Abstract

Pharmacoepidemiology is an eclectic blend of epidemiology, clinical pharmacology and biostatistics. In New Zealand there have been recent advances in pharmacoepidemiology to examine drug utilisation, monitor adverse drug events and complement pharmacovigilance. This paper attempts to describe the past, present and future of pharmacoepidemiology, particularly in the area of translational research with a particular focus on medicine use and safety. New Zealand is well-positioned globally to make significant contributions to the knowledge base of drug safety in real-world settings.

Author Information

Prasad Nishtala, School of Pharmacy, University of Otago, Dunedin; Henry Ndukwe, School of Pharmacy, University of Otago, Dunedin; Te-yuan Chyou, School of Pharmacy, University of Otago, Dunedin; Mohammed Salahudeen, School of Pharmacy, University of Otago, Dunedin; Sujita Narayan, School of Pharmacy, University of Otago, Dunedin.

Correspondence

Dr Prasad Nishtala, School of Pharmacy, University of Otago, Dunedin.

Correspondence Email

prasad.nishtala@otago.ac.nz

Competing Interests

Nil.

References

  1. Etminan M, Gill S, Fitzgerald M, Samii A. Challenges and opportunities for pharmacoepidemiology in drug-therapy decision making. J Clin Pharmacol. 2006; 46:6–9:10.1177/0091270005283285.
  2. Pearce N, Mannetje A. Occupation in Routinely Collected National Health Data: Report to the New Zealand Health Information Service. Wellington: Centre for Public Health Research, Massey University, 2004.
  3. PHARMAC Our history: Important events in PHARMAC’s history. http://www.pharmac.govt.nz/about/our-history/: Accessed 5 July 2016.
  4. Roberts E, Norris P. Regional variation in anti-depressant dispensings in New Zealand: 1993–1997. N Z Med J. 2001; 114:27–30.
  5. Narayan SW, Tordoff JM, Nishtala PS. Temporal trends in the utilisation of preventive medicines by older people: A 9-year population-based study. Arch Gerontol Geriatr. 2016; 62:103–11:10.1016/j.archger.2015.10.007.
  6. Coulter DM. The New Zealand Intensive Medicines Monitoring Programme. Pharmacoepidemiol Drug Saf. 1998; 7:79–90:10.1002/(sici)1099-1557(199803/04)7:2<79::aid-pds330>3.0.co;2-1.
  7. Kunac DL, Tatley MV, Seddon ME. A new web-based Medication Error Reporting Programme (MERP) to supplement pharmacovigilance in New Zealand--findings from a pilot study in primary care. N Z Med J. 2014; 127:69–81.
  8. Salmond CE, Crampton P. Development of New Zealand’s deprivation index (NZDep) and its uptake as a national policy tool. Can J Public Health. 2012; 103:S7–11.
  9. Brookhart MA, Sturmer T, Glynn RJ, Rassen J, Schneeweiss S. Confounding control in healthcare database research: challenges and potential approaches. Med Care. 2010; 48:S114–20:10.1097/MLR.0b013e3181dbebe3.
  10. Nishtala PS, Gnjidic D, Chyou T, Hilmer SN. Discontinuation of statins in a population of older New Zealanders with limited life expectancy. Intern Med J. 2016; 46:493–6:10.1111/imj.13024.
  11. Winnard D, Wright C, Taylor WJ, et al. National prevalence of gout derived from administrative health data in Aotearoa New Zealand. Rheumatology. 2012; 51:901–9:10.1093/rheumatology/ker361.
  12. Winnard D, Wright C, Jackson G, et al. Gout, diabetes and cardiovascular disease in the Aotearoa New Zealand adult population: co-prevalence and implications for clinical practice. The New Zealand Medical Journal (Online). 2013; 126.
  13. Blank M-L, Parkin L, Paul C, Herbison P. A nationwide nested case-control study indicates an increased risk of acute interstitial nephritis with proton pump inhibitor use. Kidney International. 2014; 86:837-44:10.1038/ki.2014.74.
  14. Parkin L, Paul C, Herbison GP. Simvastatin dose and risk of rhabdomyolysis: nested case-control study based on national health and drug dispensing data. Int J Cardiol. 2014; 174:83–9:10.1016/j.ijcard.2014.03.150.
  15. Mehta S, Wells S, Jackson R, Harrison J, Kerr A. Removal of Special Authority requirements for clopidogrel improved optimal care following percutaneous coronary intervention across sociodemographic groups. N Z Med J. 2015; 128:34–42.
  16. Research in Pharmacoepidemiology (RiPE) Group. Available at: http://www.otago.ac.nz/ripe/index.html. Accessed on: 07 June 2011.
  17. Nishtala PS, Narayan SW. Associations of Potentially Inappropriate Medicine Use with Fall-Related Hospitalisations and Primary Care Visits in Older New Zealanders: A Population-Level Study Using the Updated 2012 Beers Criteria. Drugs- Real World Outcomes. 2015.
  18. Nishtala PS, Gnjidic D, Jamieson HA, Hanger HC, Kaluarachchi C, Hilmer SN. ‘Real-world’ haemorrhagic rates for warfarin and dabigatran using population-level data in New Zealand. Int J Cardiol. 2016; 203:746–52:10.1016/j.ijcard.2015.11.067.
  19. Norris P, Horsburgh S, Becket G, et al. Equity in statin use in New Zealand. Journal of Primary Health Care. 2014; 6:17–22.
  20. Wheeler A, Moyle S, Jansen C, Robinson E, Vanderpyl J. Five-year follow-up of an acute psychiatric admission cohort in Auckland, New Zealand. The New Zealand Medical Journal (Online). 2011; 124.
  21. Gallagher LM, Kappatos D, Tisch C, Ellis PM. Suicide by poisoning in New Zealand--a toxicological analysis. The New Zealand Medical Journal (Online). 2012; 125.
  22. Roberts E, Norris P. Growth and change in the prescribing of anti-depressants in New Zealand: 1993–1997. N Z Med J. 2001; 114:25–7.
  23. Thornley S, Chan WC, Crengle S, et al. Sociodemographic differences in prevalence of diagnosed coronary heart disease in New Zealand estimated from linked national health records. The New Zealand Medical Journal (Online). 2011; 124.
  24. Yu R, Jarrett P, Holland D, Sherwood J, Pikholz C. Leprosy in New Zealand: an epidemiological update. N Z Med J. 2015; 128:9–14.
  25. Cunningham R, Shaw C, Blakely T, Atkinson J, Sarfati D. Ethnic and socioeconomic trends in breast cancer incidence in New Zealand. BMC Cancer. 2010; 10:674:10.1186/1471-2407-10-674.
  26. Kang YJ, Lewis H, Smith MA, et al. Pre-vaccination type-specific HPV prevalence in confirmed cervical high grade lesions in the Maori and non-Maori populations in New Zealand. BMC Infect Dis. 2015; 15:365:10.1186/s12879-015-1034-5.
  27. Grant CC, Chen MH, Bandara DK, et al. Antenatal immunisation intentions of expectant parents: Relationship to immunisation timeliness during infancy. Vaccine. 2016; 34:1379–88:10.1016/j.vaccine.2016.01.048.
  28. Coppell KJ, Mann JI, Williams SM, et al. Prevalence of diagnosed and undiagnosed diabetes and prediabetes in New Zealand: findings from the 2008/09 Adult Nutrition Survey. N Z Med J. 2013; 126:23–42.
  29. interRAI International Collaborative. interRAI Comprehensive Assessment System. Edition., cited 2016].Available from: http://interrai.co.nz/
  30. Gauld R, Goldfinch S. Dangerous Enthusiasms: E-Government, Computer Failure and Information System Development. Dunedin, New Zealand: Otago University Press 160 pp., 2006.