Chris Schilling1, Duncan Mortimer2, Kim Dalziel3, Emma Heeley4, John Chalmers4, Philip Clarke3. 1. Centre for Health Policy, School of Population and Global Health, University of Melbourne, Melbourne, VIC, 3051, Australia. chris.schilling@unimelb.edu.au. 2. Centre for Health Economics, Monash Business School, Monash University, Melbourne, VIC, 3800, Australia. 3. Centre for Health Policy, School of Population and Global Health, University of Melbourne, Melbourne, VIC, 3051, Australia. 4. The George Institute for Global Health, The University of Sydney and the Royal Prince Alfred Hospital, Sydney, NSW, 2050, Australia.
Abstract
BACKGROUND AND OBJECTIVE: Many guidelines for clinical decisions are hierarchical and nonlinear. Evaluating if these guidelines are used in practice requires methods that can identify such structures and thresholds. Classification and regression trees (CART) were used to analyse prescribing patterns of Australian general practitioners (GPs) for the primary prevention of cardiovascular disease (CVD). Our aim was to identify if GPs use absolute risk (AR) guidelines in favour of individual risk factors to inform their prescribing decisions of lipid-lowering medications. METHODS: We employed administrative prescribing information that is linked to patient-level data from a clinical assessment and patient survey (the AusHeart Study), and assessed prescribing of lipid-lowering medications over a 12-month period for patients (n = 1903) who were not using such medications prior to recruitment. CART models were developed to explain prescribing practice. Out-of-sample performance was evaluated using receiver operating characteristic (ROC) curves, and optimised via pruning. RESULTS: We found that individual risk factors (low-density lipoprotein, diabetes, triglycerides and a history of CVD), GP-estimated rather than Framingham AR, and sociodemographic factors (household income, education) were the predominant drivers of GP prescribing. However, sociodemographic factors and some individual risk factors (triglycerides and CVD history) only become relevant for patients with a particular profile of other risk factors. The ROC area under the curve was 0.63 (95% confidence interval [CI] 0.60-0.64). CONCLUSIONS: There is little evidence that AR guidelines recommended by the National Heart Foundation and National Vascular Disease Prevention Alliance, or conditional individual risk eligibility guidelines from the Pharmaceutical Benefits Scheme, are adopted in prescribing practice. The hierarchy of conditional relationships between risk factors and socioeconomic factors identified by CART provides new insights into prescribing decisions. Overall, CART is a useful addition to the analyst's toolkit when investigating healthcare decisions.
BACKGROUND AND OBJECTIVE: Many guidelines for clinical decisions are hierarchical and nonlinear. Evaluating if these guidelines are used in practice requires methods that can identify such structures and thresholds. Classification and regression trees (CART) were used to analyse prescribing patterns of Australian general practitioners (GPs) for the primary prevention of cardiovascular disease (CVD). Our aim was to identify if GPs use absolute risk (AR) guidelines in favour of individual risk factors to inform their prescribing decisions of lipid-lowering medications. METHODS: We employed administrative prescribing information that is linked to patient-level data from a clinical assessment and patient survey (the AusHeart Study), and assessed prescribing of lipid-lowering medications over a 12-month period for patients (n = 1903) who were not using such medications prior to recruitment. CART models were developed to explain prescribing practice. Out-of-sample performance was evaluated using receiver operating characteristic (ROC) curves, and optimised via pruning. RESULTS: We found that individual risk factors (low-density lipoprotein, diabetes, triglycerides and a history of CVD), GP-estimated rather than Framingham AR, and sociodemographic factors (household income, education) were the predominant drivers of GP prescribing. However, sociodemographic factors and some individual risk factors (triglycerides and CVD history) only become relevant for patients with a particular profile of other risk factors. The ROC area under the curve was 0.63 (95% confidence interval [CI] 0.60-0.64). CONCLUSIONS: There is little evidence that AR guidelines recommended by the National Heart Foundation and National Vascular Disease Prevention Alliance, or conditional individual risk eligibility guidelines from the Pharmaceutical Benefits Scheme, are adopted in prescribing practice. The hierarchy of conditional relationships between risk factors and socioeconomic factors identified by CART provides new insights into prescribing decisions. Overall, CART is a useful addition to the analyst's toolkit when investigating healthcare decisions.
Authors: Mona Razavian; Emma L Heeley; Vlado Perkovic; Sophia Zoungas; Andrew Weekes; Anushka A Patel; Craig S Anderson; John P Chalmers; Alan Cass Journal: Nephrol Dial Transplant Date: 2011-11-02 Impact factor: 5.992
Authors: Joanne E McKenzie; Denise A O'Connor; Matthew J Page; Duncan S Mortimer; Simon D French; Bruce F Walker; Jennifer L Keating; Jeremy M Grimshaw; Susan Michie; Jill J Francis; Sally E Green Journal: Implement Sci Date: 2010-11-10 Impact factor: 7.327
Authors: Joanne E McKenzie; Simon D French; Denise A O'Connor; Duncan S Mortimer; Colette J Browning; Grant M Russell; Jeremy M Grimshaw; Martin P Eccles; Jill J Francis; Susan Michie; Kerry Murphy; Fiona Kossenas; Sally E Green Journal: Implement Sci Date: 2013-08-19 Impact factor: 7.327
Authors: Simon D French; Joanne E McKenzie; Denise A O'Connor; Jeremy M Grimshaw; Duncan Mortimer; Jill J Francis; Susan Michie; Neil Spike; Peter Schattner; Peter Kent; Rachelle Buchbinder; Matthew J Page; Sally E Green Journal: PLoS One Date: 2013-06-13 Impact factor: 3.240
Authors: Heiner K Berthold; Ioanna Gouni-Berthold; Michael Böhm; Wilhelm Krone; Kurt P Bestehorn Journal: Cardiovasc Diabetol Date: 2009-05-13 Impact factor: 9.951
Authors: Theresa L Walunas; Jiancheng Ye; Jennifer Bannon; Ann Wang; Abel N Kho; Justin D Smith; Nicholas Soulakis Journal: Implement Sci Date: 2021-03-31 Impact factor: 7.327
Authors: Kirk A Rockett; Thomas N Williams; Carolyne M Ndila; Vysaul Nyirongo; Alexander W Macharia; Anna E Jeffreys; Kate Rowlands; Christina Hubbart; George B J Busby; Gavin Band; Rosalind M Harding Journal: Wellcome Open Res Date: 2021-09-22
Authors: Marlise E A van Eersel; Sipke T Visser; Hanneke Joosten; Ron T Gansevoort; Joris P J Slaets; Gerbrand J Izaks Journal: BMC Neurol Date: 2020-06-12 Impact factor: 2.474