1. |
- Belsti, Yitayeh, et al.
(författare)
-
Comparison of machine learning and conventional logistic regression-based prediction models for gestational diabetes in an ethnically diverse population : the Monash GDM Machine learning model
- 2023
-
Ingår i: International Journal of Medical Informatics. - : Elsevier. - 1386-5056 .- 1872-8243. ; 179
-
Tidskriftsartikel (refereegranskat)abstract
- Background: Early identification of pregnant women at high risk of developing gestational diabetes (GDM) is desirable as effective lifestyle interventions are available to prevent GDM and to reduce associated adverse outcomes. Personalised probability of developing GDM during pregnancy can be determined using a risk prediction model. These models extend from traditional statistics to machine learning methods; however, accuracy remains sub-optimal.Objective: We aimed to compare multiple machine learning algorithms to develop GDM risk prediction models, then to determine the optimal model for predicting GDM.Methods: A supervised machine learning predictive analysis was performed on data from routine antenatal care at a large health service network from January 2016 to June 2021. Predictor set 1 were sourced from the existing, internationally validated Monash GDM model: GDM history, body mass index, ethnicity, age, family history of diabetes, and past poor obstetric history. New models with different predictors were developed, considering statistical principles with inclusion of more robust continuous and derivative variables. A randomly selected 80% dataset was used for model development, with 20% for validation. Performance measures, including calibration and discrimination metrics, were assessed. Decision curve analysis was performed.Results: Upon internal validation, the machine learning and logistic regression model's area under the curve (AUC) ranged from 71% to 93% across the different algorithms, with the best being the CatBoost Classifier (CBC). Based on the default cut-off point of 0.32, the performance of CBC on predictor set 4 was: Accuracy (85%), Precision (90%), Recall (78%), F1-score (84%), Sensitivity (81%), Specificity (90%), positive predictive value (92%), negative predictive value (78%), and Brier Score (0.39).Conclusions: In this study, machine learning approaches achieved the best predictive performance over traditional statistical methods, increasing from 75 to 93%. The CatBoost classifier method achieved the best with the model including continuous variables.
|
|
2. |
- Teede, Helena J, et al.
(författare)
-
Recommendations from the 2023 International Evidence-based Guideline for the Assessment and Management of Polycystic Ovary Syndrome.
- 2023
-
Ingår i: Fertility and sterility. - 1556-5653. ; 120:4, s. 767-793
-
Tidskriftsartikel (refereegranskat)abstract
- What is the recommended assessment and management of those with polycystic ovary syndrome (PCOS), based on the best available evidence, clinical expertise, and consumer preference?International evidence-based guidelines address prioritized questions and outcomes and include 254 recommendations and practice points, to promote consistent, evidence-based care and improve the experience and health outcomes in PCOS.The 2018 International PCOS Guideline was independently evaluated as high quality and integrated multidisciplinary and consumer perspectives from six continents; it is now used in 196 countries and is widely cited. It was based on best available, but generally very low to low quality, evidence. It applied robust methodological processes and addressed shared priorities. The guideline transitioned from consensus based to evidence-based diagnostic criteria and enhanced accuracy of diagnosis, whilst promoting consistency of care. However, diagnosis is still delayed, the needs of those with PCOS are not being adequately met, evidence quality was low and evidence-practice gaps persist.The 2023 International Evidence-based Guideline update reengaged the 2018 network across professional societies and consumer organizations with multidisciplinary experts and women with PCOS directly involved at all stages. Extensive evidence synthesis was completed. Appraisal of Guidelines for Research and Evaluation-II (AGREEII)-compliant processes were followed. The Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework was applied across evidence quality, feasibility, acceptability, cost, implementation and ultimately recommendation strength and diversity and inclusion were considered throughout.This summary should be read in conjunction with the full Guideline for detailed participants and methods. Governance included a six-continent international advisory and management committee, five guideline development groups, and paediatric, consumer, and translation committees. Extensive consumer engagement and guideline experts informed the update scope and priorities. Engaged international society-nominated panels included paediatrics, endocrinology, gynaecology, primary care, reproductive endocrinology, obstetrics, psychiatry, psychology, dietetics, exercise physiology, obesity care, public health and other experts, alongside consumers, project management, evidence synthesis, statisticians and translation experts. Thirty-nine professional and consumer organizations covering 71 countries engaged in the process. Twenty meetings and five face-to-face forums over 12 months addressed 58 prioritized clinical questions involving 52 systematic and 3 narrative reviews. Evidence-based recommendations were developed and approved via consensus across five guideline panels, modified based on international feedback and peer review, independently reviewed for methodological rigour, and approved by the Australian Government National Health and Medical Research Council (NHMRC).The evidence in the assessment and management of PCOS has generally improved in the past five years, but remains of low to moderate quality. The technical evidence report and analyses (∼6000 pages) underpins 77 evidence-based and 54 consensus recommendations, with 123 practice points. Key updates include: i) further refinement of individual diagnostic criteria, a simplified diagnostic algorithm and inclusion of anti-Müllerian hormone (AMH) levels as an alternative to ultrasound in adults only; ii) strengthening recognition of broader features of PCOS including metabolic risk factors, cardiovascular disease, sleep apnea, very high prevalence of psychological features, and high risk status for adverse outcomes during pregnancy; iii) emphasizing the poorly recognized, diverse burden of disease and the need for greater healthcare professional education, evidence-based patient information, improved models of care and shared decision making to improve patient experience, alongside greater research; iv) maintained emphasis on healthy lifestyle, emotional wellbeing and quality of life, with awareness and consideration of weight stigma; and v) emphasizing evidence-based medical therapy and cheaper and safer fertility management.Overall, recommendations are strengthened and evidence is improved, but remain generally low to moderate quality. Significantly greater research is now needed in this neglected, yet common condition. Regional health system variation was considered and acknowledged, with a further process for guideline and translation resource adaptation provided.The 2023 International Guideline for the Assessment and Management of PCOS provides clinicians and patients with clear advice on best practice, based on the best available evidence, expert multidisciplinary input and consumer preferences. Research recommendations have been generated and a comprehensive multifaceted dissemination and translation programme supports the Guideline with an integrated evaluation program.This effort was primarily funded by the Australian Government via the National Health Medical Research Council (NHMRC) (APP1171592), supported by a partnership with American Society for Reproductive Medicine, Endocrine Society, European Society for Human Reproduction and Embryology, and the Society for Endocrinology. The Commonwealth Government of Australia also supported Guideline translation through the Medical Research Future Fund (MRFCRI000266). HJT and AM are funded by NHMRC fellowships. JT is funded by a Royal Australasian College of Physicians (RACP) fellowship. Guideline development group members were volunteers. Travel expenses were covered by the sponsoring organizations. Disclosures of interest were strictly managed according to NHMRC policy and are available with the full guideline, technical evidence report, peer review and responses (www.monash.edu/medicine/mchri/pcos). Of named authors HJT, CTT, AD, LM, LR, JBoyle, AM have no conflicts of interest to declare. JL declares grant from Ferring and Merck; consulting fees from Ferring and Titus Health Care; speaker's fees from Ferring; unpaid consultancy for Ferring, Roche Diagnostics and Ansh Labs; and sits on advisory boards for Ferring, Roche Diagnostics, Ansh Labs, and Gedeon Richter. TP declares a grant from Roche; consulting fees from Gedeon Richter and Organon; speaker's fees from Gedeon Richter and Exeltis; travel support from Gedeon Richter and Exeltis; unpaid consultancy for Roche Diagnostics; and sits on advisory boards for Roche Diagnostics. MC declares travels support from Merck; and sits on an advisory board for Merck. JBoivin declares grants from Merck Serono Ltd.; consulting fees from Ferring B.V; speaker's fees from Ferring Arzneimittell GmbH; travel support from Organon; and sits on an advisory board for the Office of Health Economics. RJN has received speaker's fees from Merck and sits on an advisory board for Ferring. AJoham has received speaker's fees from Novo Nordisk and Boehringer Ingelheim. The guideline was peer reviewed by special interest groups across our 39 partner and collaborating organizations, was independently methodologically assessed against AGREEII criteria and was approved by all members of the guideline development groups and by the NHMRC.
|
|
3. |
- Teede, Helena J., et al.
(författare)
-
Recommendations from the 2023 international evidence-based guideline for the assessment and management of polycystic ovary syndrome
- 2023
-
Ingår i: European Journal of Endocrinology. - 0804-4643 .- 1479-683X. ; 189
-
Tidskriftsartikel (refereegranskat)abstract
- Study question: What is the recommended assessment and management of those with polycystic ovary syndrome (PCOS), based on the best available evidence, clinical expertise, and consumer preference? Summary answer: International evidence-based guidelines address prioritized questions and outcomes and include 254 recommendations and practice points, to promote consistent, evidence-based care and improve the experience and health outcomes in PCOS. What is known already: The 2018 International PCOS Guideline was independently evaluated as high quality and integrated multidisciplinary and consumer perspectives from 6 continents; it is now used in 196 countries and is widely cited. It was based on best available, but generally very low-to low-quality, evidence. It applied robust methodological processes and addressed shared priorities. The guideline transitioned from consensus-based to evidence-based diagnostic criteria and enhanced accuracy of diagnosis, whilst promoting consistency of care. However, diagnosis is still delayed, the needs of those with PCOS are not being adequately met, the evidence quality was low, and evidence-practice gaps persist. Study design, size, and duration: The 2023 International Evidence-based Guideline update re-engaged the 2018 network across professional societies and consumer organizations with multidisciplinary experts and women with PCOS directly involved at all stages. Extensive evidence synthesis was completed. Appraisal of Guidelines for Research and Evaluation II (AGREEII)-compliant processes were followed. The Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework was applied across evidence quality, feasibility, acceptability, cost, implementation, and ultimately recommendation strength, and diversity and inclusion were considered throughout. Participants/materials, setting, and methods: This summary should be read in conjunction with the full guideline for detailed participants and methods. Governance included a 6-continent international advisory and management committee, 5 guideline development groups, and paediatric, consumer, and translation committees. Extensive consumer engagement and guideline experts informed the update scope and priorities. Engaged international society-nominated panels included paediatrics, endocrinology, gynaecology, primary care, reproductive endocrinology, obstetrics, psychiatry, psychology, dietetics, exercise physiology, obesity care, public health, and other experts, alongside consumers, project management, evidence synthesis, statisticians, and translation experts. Thirty-nine professional and consumer organizations covering 71 countries engaged in the process. Twenty meetings and 5 face-to-face forums over 12 months addressed 58 prioritized clinical questions involving 52 systematic and 3 narrative reviews. Evidence-based recommendations were developed and approved via consensus across 5 guideline panels, modified based on international feedback and peer review, independently reviewed for methodological rigour, and approved by the Australian Government National Health and Medical Research Council. Main results and the role of chance: The evidence in the assessment and management of PCOS has generally improved in the past 5 years but remains of low to moderate quality. The technical evidence report and analyses (∼6000 pages) underpin 77 evidence-based and 54 consensus recommendations, with 123 practice points. Key updates include the following: (1) further refinement of individual diagnostic criteria, a simplified diagnostic algorithm, and inclusion of anti-Müllerian hormone levels as an alternative to ultrasound in adults only; (2) strengthening recognition of broader features of PCOS including metabolic risk factors, cardiovascular disease, sleep apnoea, very high prevalence of psychological features, and high risk status for adverse outcomes during pregnancy; (3) emphasizing the poorly recognized, diverse burden of disease and the need for greater healthcare professional education, evidence-based patient information, improved models of care, and shared decision-making to improve patient experience, alongside greater research; (4) maintained emphasis on healthy lifestyle, emotional well-being, and quality of life, with awareness and consideration of weight stigma; and (5) emphasizing evidence-based medical therapy and cheaper and safer fertility management. Limitations and reasons for caution: Overall, recommendations are strengthened and evidence is improved but remains generally low to moderate quality. Significantly greater research is now needed in this neglected, yet common condition. Regional health system variation was considered and acknowledged, with a further process for guideline and translation resource adaptation provided. Wider implications of the findings: The 2023 International Guideline for the Assessment and Management of PCOS provides clinicians and patients with clear advice on best practice, based on the best available evidence, expert multidisciplinary input, and consumer preferences. Research recommendations have been generated, and a comprehensive multifaceted dissemination and translation programme supports the guideline with an integrated evaluation programme.
|
|