Search

Derrick Lopez,1 Juan Lu,1,2 Frank M Sanfilippo,1 Judith M Katzenellenbogen,1 Tom Briffa,1 Lee Nedkoff1,3 1Cardiovascular Epidemiology Research Centre, School of Population and Global Health, The University of Western Australia, Crawley, Western Australia, Australia; 2Harry Perkins Institute of Medical Research, Murdoch, Western Australia, Australia; 3Victor Chang Cardiac Research Institute, Sydney, New South Wales, AustraliaCorrespondence: Derrick Lopez, School of Population and Global Health (M431), The University of Western Australia, 35 Stirling Highway, Crawley, Western Australia, 6009, Australia, Email Derrick.Lopez@uwa.edu.auPurpose: Measures of disease burden using hospital administrative data are susceptible to over-inflation if the patient is transferred during their episode of care. We aimed to identify and compare measures of coronary heart disease (CHD) and myocardial infarction (MI) episodes using six algorithms that account for transfers.Patient and Methods: We used person-linked hospitalisations for CHD and MI for 2000– 2016 in Western Australia based on the interval between discharge and subsequent admission (date, datetime algorithms), pathway (admission source, discharge destination) and any combination to generate machine learning models (random forest [RF], gradient boosting machine [GBM]). The date and datetime algorithms used deidentified patient identifiers to identify records belonging to the same individual. We calculated counts, age-standardised rates (ASR) and age-adjusted trends for CHD and MI for each algorithm.Results: Counts of CHD increased from 11,733 in 2000 to 13,274 in 2016, while MI increased from 2605 to 4480 using the date algorithm. Correspondingly ASR for CHD decreased from 2086.2 to 1463.1 while MI increased from 468.2 to 498.1 per 100,000 person-years. ASR for CHD and MI for datetime algorithm were consistently 1– 2% higher than the date algorithm. Differences in ASR of CHD and MI counts increased over time with the admission source, RF and GBM algorithms relative to the date algorithm. Age-adjusted trends in CHD and MI episode rates using RF and GBM differed significantly from all other algorithms. Only 86.7% and 87.6% of MI episodes identified by the date algorithm were identified by the admission source and discharge destination algorithms, respectively.Conclusion: The date and datetime algorithms produced the most valid measures of CHD and MI episodes. Findings underscore the importance of identifying admission and discharge dates/times belonging to the same individual in enumerating these episodes.Keywords: patient transfer, rates, trends, Western Australia

Keira Robinson,1 Judith M Katzenellenbogen,1 Timothy J Kleinig,2 Joosup Kim,3,4 Charley A Budgeon,1 Amanda G Thrift,4 Lee Nedkoff1,5 1School of Population and Global Health, The University of Western Australia, Perth, Western Australia, Australia; 2Department of Neurology, Royal Adelaide Hospital, Adelaide, South Australia, Australia; 3Department of Medicine, The Florey Institute of Neuroscience and Mental Health, The University of Melbourne, Melbourne, Victoria, Australia; 4Department of Medicine, School of Clinical Sciences at Monash Health, Monash University, Melbourne, Victoria, Australia; 5Victor Chang Cardiac Research Institute, Sydney, New South Wales, AustraliaCorrespondence: Keira Robinson, School of Population and Global Health, The University of Western Australia, M431, 35 Stirling Hwy, Crawley, WA, 6009, Australia, Email keira.robinson@uwa.edu.auPurpose: People with cardiac disease have 2– 4 times greater risk of stroke than the general population. We measured stroke incidence in people with coronary heart disease (CHD), atrial fibrillation (AF) or valvular heart disease (VHD).Methods: We used a person-linked hospitalization/mortality dataset to identify all people hospitalized with CHD, AF or VHD (1985– 2017), and stratified them as pre-existing (hospitalized 1985– 2012 and alive at October 31, 2012) or new (first-ever cardiac hospitalization in the five-year study period, 2012– 2017). We identified first-ever strokes occurring from 2012 to 2017 in patients aged 20– 94 years and calculated age-specific and age-standardized rates (ASR) for each cardiac cohort.Results: Of the 175,560 people in the cohort, most had CHD (69.9%); 16.3% had multiple cardiac conditions. From 2012– 17, 5871 first-ever strokes occurred. ASRs were greater in females than males in single and multiple condition cardiac groups, largely driven by rates in females aged ≥ 75 years, with stroke incidence in this age group being at least 20% greater in females than males in each cardiac subgroup. In females aged 20– 54 years, stroke incidence was 4.9-fold greater in those with multiple versus single cardiac conditions. This differential declined with increasing age. Non-fatal stroke incidence was greater than fatal stroke in all age groups except in the 85– 94 age group. Incidence rate ratios were up to 2-fold larger in new versus pre-existing cardiac disease.Conclusion: Stroke incidence in people with cardiac disease is substantial, with older females, and younger patients with multiple cardiac conditions, at elevated risk. These patients should be specifically targeted for evidence-based management to minimize the burden of stroke.Keywords: stroke incidence, coronary heart disease, valvular heart disease, atrial fibrillation, cardiac disease, risk factors

D Bond-Smith,1 R Seth,1 N de Klerk,1,2 L Nedkoff,1 M Anderson,3 J Hung,1 J Cannon,1,2 K Griffiths,4,5 JM Katzenellenbogen1,2 1School of Population and Global Health, The University of Western Australia, Perth, Australia; 2Telethon Kids Institute, Perth, Australia; 3Queensland Health, Brisbane, Australia; 4Centre for Big Data Research, The University of New South Wales, Sydney, Australia; 5Menzies School of Health Research, Charles Darwin University, Darwin, AustraliaCorrespondence: D Bond-Smith Email daniela.bond-smith@uwa.edu.auBackground: Previous research has raised substantial concerns regarding the validity of the International Statistical Classification of Diseases and Related Health Problems (ICD) codes (ICD-10 I05–I09) for rheumatic heart disease (RHD) due to likely misclassification of non-rheumatic valvular disease (non-rheumatic VHD) as RHD. There is currently no validated, quantitative approach for reliable case ascertainment of RHD in administrative hospital data.Methods: A comprehensive dataset of validated Australian RHD cases was compiled and linked to inpatient hospital records with an RHD ICD code (2000– 2018, n=7555). A prediction model was developed based on a generalized linear mixed model structure considering an extensive range of demographic and clinical variables. It was validated internally using randomly selected cross-validation samples and externally. Conditional optimal probability cutpoints were calculated, maximising discrimination separately for high-risk versus low-risk populations.Results: The proposed model reduced the false-positive rate (FPR) from acute rheumatic fever (ARF) cases misclassified as RHD from 0.59 to 0.27; similarly for non-rheumatic VHD from 0.77 to 0.22. Overall, the model achieved strong discriminant capacity (AUC: 0.93) and maintained a similar robust performance during external validation (AUC: 0.88). It can also be used when only basic demographic and diagnosis data are available.Conclusion: This paper is the first to show that not only misclassification of non-rheumatic VHD but also of ARF as RHD yields substantial FPRs. Both sources of bias can be successfully addressed with the proposed model which provides an effective solution for reliable RHD case ascertainment from hospital data for epidemiological disease monitoring and policy evaluation.Keywords: rheumatic heart disease, international classification of diseases, prediction, case ascertainment, acute rheumatic fever, non-rheumatic valvular heart disease, administrative data, validation, discrimination, receiver operating curve, Australia

Judith M Katzenellenbogen,1,2,* Daniela Bond-Smith,1,* Rebecca J Seth,1 Karen Dempsey,3 Jeffrey Cannon,2 Lee Nedkoff,1 Frank M Sanfilippo,1 Nicholas de Klerk,1,2 Joe Hung,1 Elizabeth Geelhoed,4 Daniel Williamson,5 Rosemary Wyber,2,6,7 Anna P Ralph,3 Dawn Bessarab1 On behalf of the ERASE Collaboration Study Group1School of Population and Global Health, The University of Western Australia, Perth, WA, Australia; 2Group A Streptococcus Research Group, Telethon Kids Institute, Perth, WA, Australia; 3Global and Tropical Health, Menzies School of Health Research, Charles Darwin University, Darwin, NT, Australia; 4School of Allied Health, The University of Western Australia, Perth, WA, Australia; 5Aboriginal and Torres Strait Islander Health Branch, Queensland Health, Brisbane, QLD, Australia; 6Office of the Chief Scientist, The George Institute for Global Health, Sydney, NSW, Australia; 7The University of Sydney, Sydney, NSW, Australia*These authors contributed equally to this workCorrespondence: Judith M KatzenellenbogenSchool of Population and Global Health, The University of Western Australia, M431 35 Stirling Highway, Crawley, Perth, WA 6009, AustraliaTel +61 8 6488 1001Fax +61 8 6488 1188Email Judith.katzenellenbogen@uwa.edu.auPurpose: Acute rheumatic fever (ARF) and rheumatic heart disease (RHD) persist as public health issues in developing countries and among disadvantaged communities in high-income countries, with rates in Aboriginal and Torres Strait Islander peoples in Australia among the highest recorded globally. A robust evidence base is critical to support policy recommendations for eliminating RHD, but available data are fragmented and incomplete. The End RHD in Australia: Study of Epidemiology (ERASE) Project aims to provide a comprehensive database of ARF and RHD cases in Australia as a basis for improved monitoring and to assess prevention and treatment strategies. The objective of this paper is to describe the process for case ascertainment and profile of the study cohort.Patients and methods: The ERASE database has been built using linked administrative data from RHD registers, inpatient hospitalizations, and death registry data from 2001 to 2017 (mid-year). Additional linked datasets are available. The longitudinal nature of the data is harnessed to estimate onset and assess the progression of the disease. To accommodate systematic limitations in diagnostic coding for RHD, hospital-only identified RHD has been determined using a purposefully developed prediction model.Results: Of 132,053 patients for whom data were received, 42,064 are considered true cases of ARF or RHD in the study period. The patient population under 60 years in the compiled dataset is more than double the number of patients identified in ARF/RHD registers (12,907 versus 5049). Non-registered patients were more likely to be older, non-Indigenous, and at a later disease stage.Conclusion: The ERASE Project has created an unprecedented linked administrative database on ARF and RHD in Australia. These data provide a critical baseline for efforts to end ARF/RHD in Australia. The methodological work conducted to compile this database resulted in significant improvements in the robustness of epidemiological estimates and entails valuable lessons for ARF/RHD research globally.Keywords: rheumatic fever, rheumatic heart disease, indigenous, epidemiology, linked data

Isabelle Arnet,1 Melanie Greenland,2 Matthew W Knuiman,2 Jamie M Rankin,3 Joe Hung,4 Lee Nedkoff,2 Tom G Briffa,2 Frank M Sanfilippo2 1Department of Pharmaceutical Sciences, Pharmaceutical Care Research Group, University of Basel, Basel, Switzerland; 2School of Population and Global Health, Faculty of Health and Medical Sciences, The University of Western Australia, Perth, WA, Australia; 3Cardiology Department, Fiona Stanley Hospital Murdoch, WA, Australia; 4School of Medicine, Sir Charles Gairdner Hospital Unit, The University of Western Australia, Perth, WA, Australia Background: Electronic health care data contain rich information on medicine use from which adherence can be estimated. Various measures developed with medication claims data called for transparency of the equations used, predominantly because they may overestimate adherence, and even more when used with multiple medications. We aimed to operationalize a novel calculation of adherence with polypharmacy, the daily polypharmacy possession ratio (DPPR), and validate it against the common measure of adherence, the medication possession ratio (MPR) and a modified version (MPRm). Methods: We used linked health data from the Australian Pharmaceutical Benefits Scheme and Western Australian hospital morbidity dataset and mortality register. We identified a strict study cohort from 16,185 patients aged ≥65 years hospitalized for myocardial infarction in 2003–2008 in Western Australia as an illustrative example. We applied iterative exclusion criteria to standardize the dispensing histories according to previous literature. A SAS program was developed to calculate the adherence measures accounting for various drug parameters. Results: The study cohort was 348 incident patients (mean age 74.6±6.8 years; 69% male) with an admission for myocardial infarction who had cardiovascular medications over a median of 727 days (range 74 to 3,798 days) prior to readmission. There were statins (96.8%), angiotensin converting enzyme inhibitors (88.8%), beta-blockers (85.6%), and angiotensin receptor blockers (13.2%) dispensed. As expected, observed adherence values were higher with mean MPR (median 89.2%; Q1: 73.3%; Q3: 104.6%) than mean MPRm (median 82.8%; Q1: 68.5%; Q3: 95.9%). DPPR values were the most narrow (median 83.8%; Q1: 70.9%; Q3: 96.4%). Mean MPR and DPPR yielded very close possession values for 37.9% of the patients. Values were similar in patients with longer observation windows. When the traditional threshold of 80% was applied to mean MPR and DPPR values to signify the threshold for good adherence, 11.6% of patients were classified as good adherers with the mean MPR relative to the DPPR. Conclusion: In the absence of transparent and standardized equations to calculate adherence to polypharmacy from refill databases, the novel DPPR algorithm represents a valid and robust method to estimate medication possession for multi-medication regimens. Keywords: medication adherence, claims database, DPPR, medication possession ratio, ­algorithm, administrative data