Data 1
Track 2
| Friday, October 27, 2023 |
| 10:35 AM - 12:30 PM |
| Pyrmont theatre |
Speaker
A/Prof Christopher Pearce
Director Of Research
Aurora Primary Care Research Institute
Data-driven patient-centred general practice: A collaborative integrated approach
10:35 AM - 10:50 AMSummary
Aim
To develop and implement a partnership-based data strategy for general practice capable of delivering priority outcomes for the Australian public.
Content
GPs are the gateway to most realms of Australian healthcare. Analysis of the data recorded during GP consultations is a powerful source of knowledge and evidence. Capture, curation and distillation of this data at an aggregated regional level provides much earlier and deeper insight into those at highest risk of hospitalisation and deterioration compared to any other data source.
Advances in digital health promise dramatic changes in healthcare, not just in how care providers operate but in how individual patients are managed to achieve optimal care. Digital innovation in systems to link, deploy data and compute predictive analytics can be applied to deliver benefits to patients and health services, for example, reducing hospital admissions, health system costs and provision of the right care at the right time and place.
The program presented here is an action research-based integrated health intelligence platform which is capable of supporting GPs at the point of care through real-time data analytics, prompts and insights while simultaneously contributing to the national health picture by curating de-identified data snapshots for research, policy debate and health system evaluation.
This partnership program is ongoing, notable achievements to date include:
- Avoidable hospitalisation risk calculator;
- Point of care support to make evidence-based care easier for GPs;
- Supporting early diagnosis of chronic conditions;
- Increasing adherence to interventions amongst patients with complex health needs;
- Improving primary health care data linkage with a range of linkage partners including hospitals, clinical registries and the National Disability Insurance Agency.
For the past seven years, our program has demonstrated that pursuing a partnership-based integrated patient-centred data strategy for general practice enhances care and improves patient outcomes.
To develop and implement a partnership-based data strategy for general practice capable of delivering priority outcomes for the Australian public.
Content
GPs are the gateway to most realms of Australian healthcare. Analysis of the data recorded during GP consultations is a powerful source of knowledge and evidence. Capture, curation and distillation of this data at an aggregated regional level provides much earlier and deeper insight into those at highest risk of hospitalisation and deterioration compared to any other data source.
Advances in digital health promise dramatic changes in healthcare, not just in how care providers operate but in how individual patients are managed to achieve optimal care. Digital innovation in systems to link, deploy data and compute predictive analytics can be applied to deliver benefits to patients and health services, for example, reducing hospital admissions, health system costs and provision of the right care at the right time and place.
The program presented here is an action research-based integrated health intelligence platform which is capable of supporting GPs at the point of care through real-time data analytics, prompts and insights while simultaneously contributing to the national health picture by curating de-identified data snapshots for research, policy debate and health system evaluation.
This partnership program is ongoing, notable achievements to date include:
- Avoidable hospitalisation risk calculator;
- Point of care support to make evidence-based care easier for GPs;
- Supporting early diagnosis of chronic conditions;
- Increasing adherence to interventions amongst patients with complex health needs;
- Improving primary health care data linkage with a range of linkage partners including hospitals, clinical registries and the National Disability Insurance Agency.
For the past seven years, our program has demonstrated that pursuing a partnership-based integrated patient-centred data strategy for general practice enhances care and improves patient outcomes.
Takeaways
1.General practice data is the earliest available source of patient data 2. Addressing reversible health risk factors, managing chronic conditions and minimising their impacts are important primary health care priorities
3. An integrated data approach in the Australian context makes it easier for health professionals to make evidence-based care decisions which improve patient outcomes.
3. An integrated data approach in the Australian context makes it easier for health professionals to make evidence-based care decisions which improve patient outcomes.
Biography
Adam McLeod started his career as a Registered Nurse, moving into community based Drug and Alcohol and Hospital in the Home. After working in leadership positions across HARP, and Community care Adam’s passion for digital health emerged and he went on to manage a range IT projects, including programs for the Victorian Office of the CIO, Community Health Services, Monash and Eastern Health. Adam innovates in a range of data, telehealth and digital health programs. Adam holds a Masters in Health Informatics and is the architect of the POLAR platform.
Prof Michael Green
Professor
Queen's University
The Ecology of Health Services in Ontario, Canada-2023
10:50 AM - 11:05 AMSummary
Aim and objectives Ontario is the most populous province in Canada, with 14 million residents representing approximately 40% of the total Canadian population. Using standard health administrative measures, we examine the average number of health care services accessed each day in Ontario in a 24 hour period.
Content and Approach: This cohort study uses linked health administrative data sets in conjunction with standard measures of health care utilization. This includes: family physician visits, specialist visits, Emergency Department (ED) visits, hospitalizations, hip and knee replacements, prescription authorization and chronic conditions. Data for the pre-pandemic period of March 1, 2019 to Feb 28, 2020 will be included to avoid pandemic-related irregularities in service provision. Services to be captured include: all health care services, diabetes, hypertension, congestive heart failure, chronic obstructive pulmonary disease, mental health. We will report the average number of visits in 24 hours to family physicians, specialists, ED visits, hospital admissions, CT/MRI scans, and hip and knee replacements.
Goals and Results Average number of visits to family physicians, specialists, Emergency Departments, hospitalizations and hip and knee replacements in a 24 hour period will be reported and data visualized. Full results to be included in presentation.
Conclusion: Province wide utilization of health care services by the 14 million Ontarians predominantly occurs within primary care. Health policy planners and decision makers need to prioritize primary care service provision as the area where most health care occurs on a daily basis.
Content and Approach: This cohort study uses linked health administrative data sets in conjunction with standard measures of health care utilization. This includes: family physician visits, specialist visits, Emergency Department (ED) visits, hospitalizations, hip and knee replacements, prescription authorization and chronic conditions. Data for the pre-pandemic period of March 1, 2019 to Feb 28, 2020 will be included to avoid pandemic-related irregularities in service provision. Services to be captured include: all health care services, diabetes, hypertension, congestive heart failure, chronic obstructive pulmonary disease, mental health. We will report the average number of visits in 24 hours to family physicians, specialists, ED visits, hospital admissions, CT/MRI scans, and hip and knee replacements.
Goals and Results Average number of visits to family physicians, specialists, Emergency Departments, hospitalizations and hip and knee replacements in a 24 hour period will be reported and data visualized. Full results to be included in presentation.
Conclusion: Province wide utilization of health care services by the 14 million Ontarians predominantly occurs within primary care. Health policy planners and decision makers need to prioritize primary care service provision as the area where most health care occurs on a daily basis.
Takeaways
1. Be knowledgable about the extent of primary care provision in Ontario, Canada.
2. Be knowledgable on why health policy planners and decision makers need to prioritize primary care service provision.
3. Know what types of health services are provided in a 24 hour period in Ontario, Canada.
2. Be knowledgable on why health policy planners and decision makers need to prioritize primary care service provision.
3. Know what types of health services are provided in a 24 hour period in Ontario, Canada.
Biography
Dr Mike Green is the Head of Queens Dept of Family Medicine
E/Prof Siaw-Teng Liaw
Emeritus Professor
UNSW Sydney
Addressing Quality and Interoperability of data in general practice
11:05 AM - 11:20 AMSummary
Coupled with evidence-based guidelines and well-designed algorithms, digital real-world data (RWD) from health information systems (HIS) will support personalised medicine and continuity of care through the linkage of good quality information from multiple health facilities in primary and secondary care. The key challenges to data-driven innovation are interoperability and quality of digital data and tools. Interoperability standards and data quality benchmarks exist but the quality and fidelity of compliance to them remain significant unknowns. Many studies, however, suggest compliance is low and that RWD from general practice HIS are not fit for purpose.
Aim: To assess the compliance of MedicineInsight, a national repository of RWD from 600+ Australian general practices to interoperability standards and quality benchmarks.
Method: We mapped MedicineInsight data to a widely-used Common Data Model (CDM) and assessed its ability to aggregate 600+ disparate GP datasets to conduct "big data" research using 4 scenarios: opioid use for cancer & non-cancer pain, cardiovacular risk assessment and heart failure, surveillance of infectious disease and vaccine use, and youth mental health. Shared open-source data analytics & quality assessment tools were used in a secure digital health research environment at UNSW Sydney.
Outcomes: Significant preparatory technical work was required to address architecture design, operating procedures, security and governance challenges in preparing the datasets. This was compounded by a change of MedicineInsight data custodian. The syntax (CDM) and semantic (SNOMED-CT) mapping was relatively uneventful. The final results from the 4 scenarios will be discussed for methods robustness and fitness-for-purpose of RWD. The implications for good clinical practice and documentation for continuity-of-care in integrated primary care settings will discussed. Working through the Australian Health Research Alliance, a CDM approach will enable data linkage of RWD from registries, biobanks, social media and wearables to advance the potential of personalised medicine at individual and population levels.
Aim: To assess the compliance of MedicineInsight, a national repository of RWD from 600+ Australian general practices to interoperability standards and quality benchmarks.
Method: We mapped MedicineInsight data to a widely-used Common Data Model (CDM) and assessed its ability to aggregate 600+ disparate GP datasets to conduct "big data" research using 4 scenarios: opioid use for cancer & non-cancer pain, cardiovacular risk assessment and heart failure, surveillance of infectious disease and vaccine use, and youth mental health. Shared open-source data analytics & quality assessment tools were used in a secure digital health research environment at UNSW Sydney.
Outcomes: Significant preparatory technical work was required to address architecture design, operating procedures, security and governance challenges in preparing the datasets. This was compounded by a change of MedicineInsight data custodian. The syntax (CDM) and semantic (SNOMED-CT) mapping was relatively uneventful. The final results from the 4 scenarios will be discussed for methods robustness and fitness-for-purpose of RWD. The implications for good clinical practice and documentation for continuity-of-care in integrated primary care settings will discussed. Working through the Australian Health Research Alliance, a CDM approach will enable data linkage of RWD from registries, biobanks, social media and wearables to advance the potential of personalised medicine at individual and population levels.
Takeaways
1. Good clinical documentation is essential for good research data
2. Good data are complete, conform to standards and are plausible
3. General practice requires a common data model that reflects good clinical practice.
2. Good data are complete, conform to standards and are plausible
3. General practice requires a common data model that reflects good clinical practice.
Biography
Emeritus Prof Teng Liaw is a GP, clinical academic and health informatics expert based at UNSW Sydney.
Dr Julie Gordon
University of Sydney
Collection of data in general practice in Australia, for use in research
11:20 AM - 11:35 AMSummary
General practice is an essential part of the Australian health care system, and usually a patient’s first contact with the health system is a general practitioner (GP). The role of the GP includes chronic condition management, multimorbidity, population health and preventive health care, and GPs manage patients across the lifespan. The COVID-19 pandemic highlighted the importance of general practice within Australia’s model of health care. The production of valid and reliable data to understand what occurs in general practice is vital to: evaluate quality of care; health outcomes; to undertake research; and to inform policy and practice.
The aim of this presentation is to provide an overview of models that have been used in Australia to collect data from general practice for research and evaluation purposes. The content of the presentation will describe the advantages and limitations of different approaches, including active data collection methods (where the data are collected specifically for research and evaluation purposes) and passive data collection (where data collected for direct patient care, often through clinical information systems, are re-used for research and evaluation). The role of clinical classifications and terminologies will be described, with particular reference to Wonca’s International Classification of Primary Care, Version 2 (ICPC-2) and the Australian clinical terminology classified to ICPC-2, called ICPC-2 PLUS. The presentation will outline the importance of the data structures used in clinical information systems to re-use of the data collected, focussing on the need for standardised data element names and definitions, and required linkages between data elements. Emerging developments in Australian general practice data collection will be described.
The aim of this presentation is to provide an overview of models that have been used in Australia to collect data from general practice for research and evaluation purposes. The content of the presentation will describe the advantages and limitations of different approaches, including active data collection methods (where the data are collected specifically for research and evaluation purposes) and passive data collection (where data collected for direct patient care, often through clinical information systems, are re-used for research and evaluation). The role of clinical classifications and terminologies will be described, with particular reference to Wonca’s International Classification of Primary Care, Version 2 (ICPC-2) and the Australian clinical terminology classified to ICPC-2, called ICPC-2 PLUS. The presentation will outline the importance of the data structures used in clinical information systems to re-use of the data collected, focussing on the need for standardised data element names and definitions, and required linkages between data elements. Emerging developments in Australian general practice data collection will be described.
Takeaways
At the conclusion of this presentation attendees will understand:
1. how data are collected and used in Australia for general practice research and evaluation
2. how the data structures used within clinical information systems can influence the reliability and validity of data
3. the importance of clinical classifications and terminologies in general practice research.
1. how data are collected and used in Australia for general practice research and evaluation
2. how the data structures used within clinical information systems can influence the reliability and validity of data
3. the importance of clinical classifications and terminologies in general practice research.
Biography
Julie Gordon is a health information manager and health services researcher with over twenty years' experience. Her work focuses on the development of clinical classifications and terminologies, and their application in both clinical settings and for research.
Prof Mark Morgan
Professor of General Practice
Bond University
GP data for GPs: the effectiveness of real-time computer decision support in Australia
11:35 AM - 11:50 AMSummary
AIM: To evaluate a software program that analyses routinely collected electronic clinical information before returning real-time alerts and prompts.
METHOD: After development, testing and implementation across a region of Australia an evaluation of impact was made. Response rates in which patient care altered were recorded for the various medication-safety alerts and care prompts.
RESULTS: Primary Sense software was installed in 87 GP practices in the Gold Coast region of Australia. Medication-safety alerts resulted in changed patient care for 45% of alerts (30-69% between different safety alerts). Risk-stratified care prompts were followed 26% of the time (12%-57% between different prompt categories). Response rates were maintained over two years of data collection.
DISCUSSION: This research has demonstrated a remarkable impact that far exceeds other interventions including educational, audit with feedback and nudge letters.
CONCLUSION: Real-time computer decision support is a valuable tool that can be delivered at scale. As the software is rolled out nationally, there is further scope to combine risk information held in multiple practices to further improve patient safety.
METHOD: After development, testing and implementation across a region of Australia an evaluation of impact was made. Response rates in which patient care altered were recorded for the various medication-safety alerts and care prompts.
RESULTS: Primary Sense software was installed in 87 GP practices in the Gold Coast region of Australia. Medication-safety alerts resulted in changed patient care for 45% of alerts (30-69% between different safety alerts). Risk-stratified care prompts were followed 26% of the time (12%-57% between different prompt categories). Response rates were maintained over two years of data collection.
DISCUSSION: This research has demonstrated a remarkable impact that far exceeds other interventions including educational, audit with feedback and nudge letters.
CONCLUSION: Real-time computer decision support is a valuable tool that can be delivered at scale. As the software is rolled out nationally, there is further scope to combine risk information held in multiple practices to further improve patient safety.
Takeaways
1. Routinely recorded general practice data can be analysed to provide an additional layer of prescribing safety in the form of real-time alerts a
2. Comprehensive analysis of demographic, clinical and billing data can improve care for high-risk patients by providing real-time prompts
3. Computer decision support designed by GPs for GPs is highly impactful and can be scaled across whole regions
2. Comprehensive analysis of demographic, clinical and billing data can improve care for high-risk patients by providing real-time prompts
3. Computer decision support designed by GPs for GPs is highly impactful and can be scaled across whole regions
Biography
Over the last 20 years Prof Morgan has worked as a GP and academic in rural, regional and urban Australia. In that time, he has led a body of research to improve service delivery for people living with multiple medical conditions. As a senior leader within the Royal Australian College of General Practitioners, he chairs the development of national guidelines and resources for our 40,000 members.
His leadership, research and experience in knowledge translation allows him to contribute to many national strategic initiatives and government committees. This is supported by a strong teaching background in general practice for 20 years and a position at the University of Adelaide where he led the GP program. His move to Bond in 2017 has seen an extension of his skills to become the Phase Lead for the clinical component of the Medical Program and Associate Dean of External Engagement for the Faculty of Health Sciences and Medicine. In 2020, he was appointed Professor of General Practice.
Prof Nigel Stocks
University Of Adelaide
Research outputs from MedicineInsight: an Australian primary care Electronic Medical Record database: 2017-2022
11:50 AM - 11:55 AMSummary
Background: Electronic Medical Record (EMR) databases are common across Europe and North America. Over many years they have provided valuable research results for clinical practice, policy and use of medications, pathology tests and other investigations in general practice/primary care. In Australia, the Bettering the Evaluation and Care of Health (BEACH) study of general practitioner clinical activity 1998-2016, was used by researchers, industry and government to inform health system planning and practice. More recently NPS MedicineWise was funded by the Federal Government to develop a large national general practice database containing de-identified EMRs from more than 650 general practices (8.2% of all practices in the country) and over 2,700 GPs from all Australian states and regions. Unlike BEACH, MedicineInsight allows the collection of longitudinal data for patients which can be linked to other data sets. Data has been available to researchers since 2017. This study investigates the research outputs and reports arising from this access until the end of 2022.
Methods: Publicly available data from the MedicineInsight website will be accessed, searches of the literature using PUBMED, SCOPUS and other electronic databases undertaken to ascertain the number of publications, citations and their impact. Contact with researchers, Government, and employees of NPS Medicinewise will be used to ascertain translational outcomes.
Results: Preliminary analysis of data from 2017-2022 indicates that 66 papers have been published with a total of 532 cites and an H index of 12. The most cited clinical papers explored musculoskeletal epidemiology (42), use of SGLT2 inhibitors for diabetes and risk of infection (38) and sex disparities in the management of coronary heart disease (26). Additional results including researcher disciplines, number of reports and qualitative assessment of translational outcomes will be presented.
Conclusion: EMR databases are valuable research tools, their full potential should be used to improve health outcomes.
Methods: Publicly available data from the MedicineInsight website will be accessed, searches of the literature using PUBMED, SCOPUS and other electronic databases undertaken to ascertain the number of publications, citations and their impact. Contact with researchers, Government, and employees of NPS Medicinewise will be used to ascertain translational outcomes.
Results: Preliminary analysis of data from 2017-2022 indicates that 66 papers have been published with a total of 532 cites and an H index of 12. The most cited clinical papers explored musculoskeletal epidemiology (42), use of SGLT2 inhibitors for diabetes and risk of infection (38) and sex disparities in the management of coronary heart disease (26). Additional results including researcher disciplines, number of reports and qualitative assessment of translational outcomes will be presented.
Conclusion: EMR databases are valuable research tools, their full potential should be used to improve health outcomes.
Takeaways
Understand the potential of general practice electronic medical record databases for research
Be more aware of how GP data can be used to inform practice and policy
Be able to share the success of MedicineInsight with colleagues
Be more aware of how GP data can be used to inform practice and policy
Be able to share the success of MedicineInsight with colleagues
Biography
Prof Stocks is Head of the Discipline of General Practice at the University of Adelaide. He is a general practitioner (FRACGP) and public health physician (FAFPHM). He is an experienced primary health care researcher who has conducted both qualitative and quantitative studies including RCT’s, cohort and observational studies, systematic reviews and meta-analyses. His main interests are in cardiovascular health (including CVD risk), infectious disease, prevention and quality of life with an emphasis on clinical and health services research. He has successfully recruited patients to large (ASPREE and WISDOM) and many smaller randomised controlled trials. For several years he was the RACGP SA/NT Faculty Chair, a member and Chair of their national Council and member of their research committee. He now supports the College as Chair of their National Awards Committee and member of their Research Foundation.
Dr Geannagail Anuran
Medical Specialist Ii
Department Of Family And Community Medicine, Philippine General Hospital
Feedback, workshop and random monitoring as QA interventions in improving data entries of resident physicians in EMR for COVID-19 teleconsultations
11:55 AM - 12:00 PMSummary
Background: Medical records provide a repository of patient information, physical examination, laboratory findings, and the outcomes of interventions. The completeness of data contained in the electronic medical record (EMR) is an important factor leading to health service improvement.
Objective: To determine the effectiveness of quality assurance (QA) interventions (feedback, workshop and random monitoring system) in improving data entries in the EMR of resident physicians for COVID-19 teleconsultations.
Methods: This was a before-and-after study involving EMR entries of physician trainees on health care workers (HCWs) during the 2022 COVID-19 pandemic. All medical records of HCWs who sought consultations at the employees’ clinic for COVID-19 symptoms and/or exposure were included. A chart audit was conducted against a checklist of criteria for six months before and three months after the study interventions (one-day QA workshop, provision of feedback, and random monitoring). The change in the level of completeness from pre- to post-intervention was computed, and the percentage of charts meeting the minimum standards was likewise determined.
Results: A total of 2358 and 862 chart entries were audited before and after the QA interventions, respectively. The percentage of charts meeting five out of fifteen criteria post-intervention increased by 13-61 percentage points. The improvement in one criterion was adequate to reach the minimum standard for chart completeness set at 90%. Ninety-three to ninety-nine percent of charts met the minimum standards for 10 out of the 15 criteria for chart completeness, while only 23-66% of audited charts met the other five criteria.
Conclusion: The QA interventions of conducting a workshop, provision of feedback on performance, and random monitoring of chart entries of resident physicians led to an increasing trend on the percentage of charts meeting the minimum standards for completeness.
Objective: To determine the effectiveness of quality assurance (QA) interventions (feedback, workshop and random monitoring system) in improving data entries in the EMR of resident physicians for COVID-19 teleconsultations.
Methods: This was a before-and-after study involving EMR entries of physician trainees on health care workers (HCWs) during the 2022 COVID-19 pandemic. All medical records of HCWs who sought consultations at the employees’ clinic for COVID-19 symptoms and/or exposure were included. A chart audit was conducted against a checklist of criteria for six months before and three months after the study interventions (one-day QA workshop, provision of feedback, and random monitoring). The change in the level of completeness from pre- to post-intervention was computed, and the percentage of charts meeting the minimum standards was likewise determined.
Results: A total of 2358 and 862 chart entries were audited before and after the QA interventions, respectively. The percentage of charts meeting five out of fifteen criteria post-intervention increased by 13-61 percentage points. The improvement in one criterion was adequate to reach the minimum standard for chart completeness set at 90%. Ninety-three to ninety-nine percent of charts met the minimum standards for 10 out of the 15 criteria for chart completeness, while only 23-66% of audited charts met the other five criteria.
Conclusion: The QA interventions of conducting a workshop, provision of feedback on performance, and random monitoring of chart entries of resident physicians led to an increasing trend on the percentage of charts meeting the minimum standards for completeness.
Takeaways
At the conclusion of my presentation, attendees will take away:
1. The completeness of chart entries in electronic medical records is important for ensuring patient safety.
2. Quality assurance activities are conducted to ensure continuous health service improvement.
3. Feedback, workshop, and random monitoring are helpful in improving chart entries of physician trainees.
1. The completeness of chart entries in electronic medical records is important for ensuring patient safety.
2. Quality assurance activities are conducted to ensure continuous health service improvement.
3. Feedback, workshop, and random monitoring are helpful in improving chart entries of physician trainees.
Biography
Geannagail Anuran is a clinician, educator, manager, and researcher. Her interests include personnel development, quality improvement, and operational excellence.
Dr J Martinez Gutierrez
Research Fellow
University Of Melbourne
Using primary care data to identify patients at risk of cancer in primary care
12:00 PM - 12:05 PMSummary
Background:
Up to 50% of patients with anaemia and thrombocytosis are not followed up in primary care. Inadequate follow-up of abnormal test results can lead to missed opportunities for early cancer diagnosis. Primary care data, stored in general practice electronic medical records, can be used to identify patients at risk of undiagnosed cancer. We aimed to evaluate the follow-up of patients with abnormal test results such as anaemia or thrombocytosis in an Australian primary care dataset. Our goal was to use this information to develop a clinical decision-support tool to identify patients at risk of cancer and improve early cancer diagnosis in general practice.
Methods:
We retrospectively analysed 10 years of data (2010-2019) from 79 general practice sites and a cohort of 598,075 unique patients captured in the Patron primary care database. We calculated prevalence and inadequate follow-up for patients with anaemia and thrombocytosis.
Results:
Prevalence of anaemia and thrombocytosis was 17.6% and 4.6% respectively. Patients with one abnormal test result, had no evidence of follow-up in 60.5% and 75.6% of cases of anaemia and thrombocytosis respectively. This increased to more than 85% for patients with two abnormal test results. Using this data, we have developed an algorithm for early cancer detection.
Conclusion:
Most patients with anaemia and thrombocytosis are not followed up in Australian general practices increasing the risk of delayed diagnosis of cancer. Primary care data can be used to develop tools to facilitate timely cancer detection in primary care.
Up to 50% of patients with anaemia and thrombocytosis are not followed up in primary care. Inadequate follow-up of abnormal test results can lead to missed opportunities for early cancer diagnosis. Primary care data, stored in general practice electronic medical records, can be used to identify patients at risk of undiagnosed cancer. We aimed to evaluate the follow-up of patients with abnormal test results such as anaemia or thrombocytosis in an Australian primary care dataset. Our goal was to use this information to develop a clinical decision-support tool to identify patients at risk of cancer and improve early cancer diagnosis in general practice.
Methods:
We retrospectively analysed 10 years of data (2010-2019) from 79 general practice sites and a cohort of 598,075 unique patients captured in the Patron primary care database. We calculated prevalence and inadequate follow-up for patients with anaemia and thrombocytosis.
Results:
Prevalence of anaemia and thrombocytosis was 17.6% and 4.6% respectively. Patients with one abnormal test result, had no evidence of follow-up in 60.5% and 75.6% of cases of anaemia and thrombocytosis respectively. This increased to more than 85% for patients with two abnormal test results. Using this data, we have developed an algorithm for early cancer detection.
Conclusion:
Most patients with anaemia and thrombocytosis are not followed up in Australian general practices increasing the risk of delayed diagnosis of cancer. Primary care data can be used to develop tools to facilitate timely cancer detection in primary care.
Takeaways
1. Almost half of the patients with anaemia and thrombocytosis are not followed-up in primary care and this can lead to missed opportunities for early cancer diagnosis
2. Primary care data, stored in general practice electronic medical records, can be used to identify patients at risk of undiagnosed cancer
3. Using the information in a primary care dataset, we developed a clinical decision support tool to identify patients at risk of an undiagnosed cancer
2. Primary care data, stored in general practice electronic medical records, can be used to identify patients at risk of undiagnosed cancer
3. Using the information in a primary care dataset, we developed a clinical decision support tool to identify patients at risk of an undiagnosed cancer
Biography
Javiera Martinez Gutierrez is an academic, primary care researcher, and clinician. She is currently a research fellow at the Center for Cancer Research in the Department of General Practice at the University of Melbourne, Australia. She also holds an academic appointment at Pontificia Universidad Catolica de Chile in the Department of Family Medicine. Her research focus is on cancer prevention, screening, and early detection. She is interested in community participation and the use of mobile technologies to improve cancer detection and care in general practice.
