Suchergebnisse

IntroductionBeta-blockers, renin-angiotensin system inhibitors and statins are evidence-based pharmacotherapies for preventing recurrent acute coronary syndrome (ACS), however their use in older people may depend on clinical factors such as comorbidity and frailty. Deriving comorbidity information from linked administrative data is well established. The recent development of the Hospital Frailty Risk Score (HFRS) now allows ascertainment of frailty from linked administrative data.
Objectives and ApproachWe determined the ability of the HFRS to discriminate between groups of older patients initiated on these three cardio-protective medicines following hospitalisation for an incident ACS. We used a 15-year look back to identify incident ACS cases between 2005 and 2008 from Western Australian hospital data and linked them to national pharmaceutical dispensing records. The study was limited to patients aged ≥65 years who had not received these medicines in the last two years and were discharged alive from their index ACS admission. Separate competing risk regression models assessed the association between HFRS and initiation of each medicine, controlling for comorbidities and other variables available from the linked datasets.
ResultsOverall, increasing levels of HFRS were associated with lower subdistribution hazards of initiation of each medicine examined, however these varied by age group and sex. For example, the subdistribution hazard ratios (SHR) for beta-blocker initiation among men in the oldest age group (≥85 years) were 0.65 (95% CI: 0.44-0.95) and 0.51 (95% CI: 0.29-0.90) in the intermediate and high HFRS groups respectively compared to the low frailty group. The corresponding SHR for women were 1.01 (95% CI: 0.76-1.35) and 0.89 (95% CI: 0.63-1.27), respectively.
Conclusion / ImplicationsThe HFRS applied to linked administrative data discriminates between groups of older people who received cardio-protective medicines following an incident ACS. There is potential for it to be used in other chronic disease conditions.

IntroductionHospital administrative data is a valuable source to measure myocardial infarction (MI) rates. However, admission counts are susceptible to over-inflation if the patient is transferred multiple times during a single episode of care, and variables denoting transfers may not be reliable. To obtain an accurate number of events, hospital transfers need to be correctly identified.
Objectives and ApproachWe assessed multivariable logistic regression and various machine-learning models to predict transfers in hospital administrative data. Using Western Australian linked hospital data, we identified records from 2000-2016 with a principal discharge diagnosis of MI. Our standard method to compare against was a 24-hour look-back to identify a transfer using just admission and separation dates from the current and previous records for the same patient. Multivariable logistic regression and decision trees with various boosting algorithms were used to predict if a single record was a transfer, using variables recorded in the admission (e.g. age, sex, type of hospital, admitted from, emergency/elective admission). The performance of each model was calculated using metrics including area under the curve (AUC).
ResultsRecords in the training, validation and testing samples had similar characteristics: mean age=68.9 years, 66% were male and 58% admitted to tertiary hospitals. Gradient Boosting Decision Tree (AUC=0.887, 95%CI: 0.886-0.887) outperformed multivariable logistic regression (AUC=0.875; 95% CI: 0.869-0.881) and random forest models (AUC=0.859; 95% CI: 0.853-0.865).
Conclusion / ImplicationsMultivariable logistic regression and machine-learning models are able to identify transfers in a single record from existing variables. They can be used in unlinked hospital administrative data where records belonging to the same patient cannot be identified.

IntroductionLinked data are increasingly used in pharmacoepidemiology studies to enhance value beyond that which can be achieve from stand-alone pharmaceutical data. The complexity of pharmaceutical data can make any linked data analysis challenging and it is imperative that this is matched by the human capacity to perform this work.
Objectives and ApproachResearch is needed to understand the state of the current pharmacoepidemiology workforce and to prioritise its capacity building needs. We aim to profile the Australian pharmacoepidemiology workforce to explore views, needs, priority areas and perspectives relevant to capacity building. Participants are the regular pharmacoepidemiology workforce (Group 1) and senior medicines stakeholders (Group 2). Following a literature review and consultation with a group of key informants, we developed survey and interview instruments for each group. We piloted the instruments in February 2018 and study data collection is planned for March 2018. We will use a mixed-methods approach to analyse the data.
ResultsWe conducted a review of existing literature and identified workforce views, needs and priorities at four levels: personal, team, organisation and wider community. During the consultative process, the informants highlighted the multidisciplinary nature of the pharmacoepidemiology workforce including many with non-health related backgrounds. They also raised concerns about attracting applicants with suitable skills and experience, job satisfaction, career progression and workforce retention. We developed instruments to (i) further explore these issues, (ii) ascertain their experience with linked health data, (iii) determine their training needs, and, (iv) learn about their future intentions. We will present findings on issues pertinent to the Australian pharmacoepidemiology landscape and suggest priorities for building workforce capacity.
Conclusion/ImplicationsThis study will provide empirical evidence to support and prioritise capacity building in the Australian pharmacoepidemiology workforce to improve their ability to work with linked data. The instruments that we developed and findings may be relevant to phamacoepidemiology workforce in other countries and other emerging fields that use linked data.

ABSTRACTObjectivesNational statistics for hospital admissions for acute CHD based on unlinked administrative data are inflated because of inter/intra-hospital transfers or related readmissions for further investigations or procedures. Our objective was to estimate the inflation of CHD inpatient counts using multiple approaches based initially on Western Australian data that can be applied to future National studies.
ApproachWe used a linked hospital morbidity dataset from the Western Australian Data Linkage System to determine hospitalisations for each CHD subcategory from 1990-2010. Transfers were defined as contiguous admissions separated by ≤1 day. Episodes-of-care (EOC) were defined as admissions (with/without transfers) that were within 28 days of the initial CHD admission. As the principal diagnosis may vary between hospitals involved in transfers or admissions within an EOC, we explored four approaches for allocating a diagnosis:i. Hierarchical diagnosis: selection of diagnosis based on clinical severity (ST-elevation myocardial infarction (STEMI)>non-STEMI>unstable angina>stable angina>other CHD>chest pain); ii. Hospital hierarchy: diagnosis based on highest hospital level (tertiary>private>other metropolitan non-tertiary>rural);iii/iv. Temporal order of diagnosis: diagnosis based on first or last record in transfer/EOC.
ResultsThe proportion of cases that were transferred varied according to disease severity and time: 13% (1990) to 27% (2010) for STEMI; 5% to 7% for stable angina and unchanged at 4% for chest pain. Compared to transfer-level data using the first approach, unlinked data overestimated STEMI counts by 3% (1990) to 11% (2010), stable angina by 3% to 5% and chest pain by 6% to 6%. Similarly for EOC-level data, the overestimates were 5% (1990) to 12% (2010) for STEMI, 13% to 19% for stable angina and 20% to 14% for chest pain. The four approaches for allocating a diagnosis produced differing counts with the difference being larger for more clinically severe diagnoses than for less clinically severe diagnoses. For example, using transfer-level data, the differences between approaches i and iv in 2010 were 12%, 2% and 1% for STEMI, stable angina and chest pain respectively.
ConclusionThere is a potential to overestimate counts of CHD in inpatient data if transfers and readmissions are not taken into account, and this inaccuracy can differ across disease subcategories and approach used. This has important implications where higher disease severity, such as myocardial infarction, is an indicator of population health. Transfer- or EOC-level data are more likely to reflect true CHD hospitalisation counts than unlinked-level data, and are more appropriate for epidemiological studies of CHD rates.

ABSTRACTObjectivesAlthough outcomes for the majority of women diagnosed with primary breast cancer are good, with five-year survival exceeding 90%, some women will experience cancer recurrence and ultimately die from the disease. It is important for patients, clinicians and health service planners to know the risk of recurrence once initial treatment for primary breast cancer is completed. However, none of Australia's State or Territory cancer registries routinely report on cancer recurrence which could be used to evaluate this issue. To address this absence of direct reporting, we aimed to determine the incidence of cancer recurrence in Australian clinical practice after completion of treatment for primary breast cancer, using a range of linked health data sources.
ApproachWe performed a retrospective cohort study using linked health data from New South Wales (NSW), Australia. Data were linked from six data collections: i) Cancer Registry, ii) Admitted Patient Data Collection, iii) Pharmaceutical Benefits Scheme claims, iv) Medicare (outpatient) claims, v) Death Registry; and the vi) NSW 45 and Up Study. We identified 2416 women diagnosed with primary invasive breast cancer during 2003-2008 in NSW who had not had a recurrence by 18 months post-diagnosis. Unit-level hospital, pharmacy and outpatient claims were used to identify services indicative of recurrence. Incidence of recurrence was calculated and multivariate Cox regression used to identify baseline and active treatment characteristics predictive of cancer recurrence up to six years post-diagnosis.
ResultsA total of 217 women (9.0%) had a hospital, pharmacy or outpatient claim indicating breast cancer recurrence between 18 months and six years post-diagnosis. Overall annual cumulative incidence of recurrence was 3.3%. Recurrence was significantly higher for women with node-positive (4.8% vs. 2.5% annually, adjHR=1.7, 95%CI=1.3-2.3) or hormone receptor-negative (3.8% vs. 3.1% annually, adjHR=1.3, 95%CI=1.0-1.7) tumours. Women with tumours >2cm at diagnosis were more likely to experience recurrence within six years compared with those with a smaller initial tumour (4.8% vs. 2.7 annually, adjHR=1.5; 95%CI=1.1-2.0).
ConclusionsWomen with breast cancer in this Australian cohort experienced recurrence at 3.3%pa in the years following completion of treatment. Those at greatest risk of recurrence were women with node-positive or hormone-receptor negative tumours, or tumours >2cm at initial diagnosis, consistent with international findings. This method for ascertaining breast cancer recurrence can be used to assess population-level changes over time and to investigate the impact of specific treatments on outcomes in the absence of available Cancer Registry data.

ObjectivesCardiovascular disease is the largest contributor of increased mortality in patients with gout. Acute inflammation as seen with gout attacks may have a mechanistic role in developing Major Adverse Cardiovascular Events (MACE). We examined the temporal relationship between admission to hospital with acute gout and MACE.
ApproachLinked inpatient and mortality data from the Western Australian Rheumatic Disease Epidemiology Registry were used. We identified patients with an incident acute gout (index) hospitalisation and admission or death records due to MACE (composite of acute coronary syndrome, stroke, heart failure, cardiovascular death). The risk of MACE during the index post-discharge period (1-30 days after index admission) and control period (365 days prior to index admission and 365 days post-discharge) was determined using a self-controlled case-series (SCCS) design. Conditional fixed-effects Poisson regression was used to obtain incidence rate ratios (IRR). Sensitivity analyses were performed excluding deaths and 180-day events.
ResultsWe identified 962 patients (mean age=76.2 years [SD=12.2]; 66.8% male) with incident acute gout admission and documented MACE during the control and/or index post-discharge periods. 917 (95.3%) patients experienced MACE during the control period and 114 (11.9%) during the index post-discharge period. The rate of MACE during the control and post-discharge periods were 0.84 and 1.44 events per person-year, respectively, with an IRR=1.67 (95% CI: 1.38-2.02) for the post-discharge period compared with the control period from regression analysis. Sensitivity analyses excluding deaths and 180-day events were IRR=1.68 (95% CI=1.29-2.20) and IRR=1.66 (95% CI=1.34-2.07) respectively.
ConclusionOur self-controlled case-series study using linked administrative data found an increased risk of MACE during the 30 days after discharge for index gout hospitalisation. This suggests a temporal association between acute inflammation and MACE.