Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
133 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Towards Clinical Prediction with Transparency: An Explainable AI Approach to Survival Modelling in Residential Aged Care (2312.00271v3)

Published 1 Dec 2023 in cs.LG

Abstract: Background: Accurate survival time estimates aid end-of-life medical decision-making. Objectives: Develop an interpretable survival model for elderly residential aged care residents using advanced machine learning. Setting: A major Australasian residential aged care provider. Participants: Residents aged 65+ admitted for long-term care from July 2017 to August 2023. Sample size: 11,944 residents across 40 facilities. Predictors: Factors include age, gender, health status, co-morbidities, cognitive function, mood, nutrition, mobility, smoking, sleep, skin integrity, and continence. Outcome: Probability of survival post-admission, specifically calibrated for 6-month survival estimates. Statistical Analysis: Tested CoxPH, EN, RR, Lasso, GB, XGB, and RF models in 20 experiments with a 90/10 train/test split. Evaluated accuracy using C-index, Harrell's C-index, dynamic AUROC, IBS, and calibrated ROC. Chose XGB for its performance and calibrated it for 1, 3, 6, and 12-month predictions using Platt scaling. Employed SHAP values to analyze predictor impacts. Results: GB, XGB, and RF models showed the highest C-Index values (0.714, 0.712, 0.712). The optimal XGB model demonstrated a 6-month survival prediction AUROC of 0.746 (95% CI 0.744-0.749). Key mortality predictors include age, male gender, mobility, health status, pressure ulcer risk, and appetite. Conclusions: The study successfully applies machine learning to create a survival model for aged care, aligning with clinical insights on mortality risk factors and enhancing model interpretability and clinical utility through explainable AI.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (60)
  1. Relationship between cancer patients’ predictions of prognosis and their treatment preferences. Jama, 279(21):1709–1714, 1998.
  2. Extent and determinants of error in doctors’ prognoses in terminally ill patients: prospective cohort studycommentary: Why do doctors overestimate? commentary: Prognoses should be based on proved indices not intuition. Bmj, 320(7233):469–473, 2000.
  3. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. Jama, 300(14):1665–1673, 2008.
  4. Prediction of institutionalization in the elderly. a systematic review. Age and ageing, 39(1):31–38, 2010.
  5. The language of dying: Communication about end-of-life in residential aged care. Death Studies, 46(3):684–694, 2022.
  6. Length of stay for older adults residing in nursing homes at the end of life. Journal of the American Geriatrics Society, 58(9):1701–1706, 2010.
  7. Do the elderly have a voice? advance care planning discussions with frail and older individuals: a systematic literature review and narrative synthesis. British Journal of General Practice, 63(615):e657–e668, 2013.
  8. A comprehensive review of machine learning algorithms and their application in geriatric medicine: present and future. Aging Clinical and Experimental Research, pages 1–35, 2023.
  9. Personalized medicine: will it work for decreasing age-related morbidities? In Aging, pages 683–700. Elsevier, 2023.
  10. Application of machine learning approaches in predicting clinical outcomes in older adults–a systematic review and meta-analysis. BMC geriatrics, 23(1):561, 2023.
  11. Responsive and minimalist app based on explainable ai to assess palliative care needs during bedside consultations on older patients. Sustainability, 13(17):9844, 2021.
  12. Prediction of mortality in geriatric traumatic brain injury patients using machine learning algorithms. Brain Sciences, 13(1):94, 2023.
  13. A new random forest algorithm-based prediction model of post-operative mortality in geriatric patients with hip fractures. Frontiers in Medicine, 9, 2022. ISSN 2296-858X. doi:10.3389/fmed.2022.829977. URL https://www.frontiersin.org/articles/10.3389/fmed.2022.829977.
  14. Machine Learning Approaches to Predict 6-Month Mortality Among Patients With Cancer. JAMA Network Open, 2(10):e1915997–e1915997, 10 2019. ISSN 2574-3805. doi:10.1001/jamanetworkopen.2019.15997. URL https://doi.org/10.1001/jamanetworkopen.2019.15997.
  15. Validation of a Machine Learning Algorithm to Predict 180-Day Mortality for Outpatients With Cancer. JAMA Oncology, 6(11):1723–1730, 11 2020. ISSN 2374-2437. doi:10.1001/jamaoncol.2020.4331. URL https://doi.org/10.1001/jamaoncol.2020.4331.
  16. A comparison of machine learning methods for survival analysis of high-dimensional clinical data for dementia prediction. Scientific reports, 10(1):20410, 2020.
  17. Development and validation of a deep learning algorithm for mortality prediction in selecting patients with dementia for earlier palliative care interventions. JAMA network open, 2(7):e196972–e196972, 2019.
  18. Development and external validation of a mortality prediction model for community-dwelling older adults with dementia. JAMA Internal Medicine, 182(11):1161–1170, 2022.
  19. Helicobacter pylori (h. pylori) risk factor analysis and prevalence prediction: a machine learning-based approach. BMC Infectious Diseases, 22(1):655, 2022.
  20. Machine learning algorithms for identifying predictive variables of mortality risk following dementia diagnosis: a longitudinal cohort study. Scientific Reports, 13(1):9480, 2023.
  21. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (tripod) the tripod statement. Circulation, 131(2):211–219, 2015.
  22. The validity of the rx-risk comorbidity index using medicines mapped to the anatomical therapeutic chemical (atc) classification system. BMJ open, 8(4):e021122, 2018.
  23. URL https://catalog.interrai.org/content/interrai-long-term-care-facilities-ltcf-assessment-form-and-users-manual-australian-edition.
  24. Use of the interrai chess scale to predict mortality among persons with neurological conditions in three care settings. PloS one, 9(6):e99066, 2014.
  25. Minimum data set changes in health, end-stage disease and symptoms and signs scale: a revised measure to predict mortality in nursing home residents. Journal of the American Geriatrics Society, 66(5):976–981, 2018.
  26. Validation of the interrai cognitive performance scale against independent clinical diagnosis and the mini-mental state examination in older hospitalized patients. The journal of nutrition, health & aging, 17:435–439, 2013.
  27. Convergent validity, concurrent validity, and diagnostic accuracy of the interrai depression rating scale. Journal of Geriatric Psychiatry and Neurology, 29(6):361–368, 2016.
  28. Development of the interrai pressure ulcer risk scale (purs) for use in long-term care and home care settings. BMC geriatrics, 10:1–10, 2010.
  29. A multivariate technique for multiply imputing missing values using a sequence of regression models. Survey methodology, 27(1):85–96, 2001.
  30. Stef Van Buuren. Multiple imputation of discrete and continuous data by fully conditional specification. Statistical methods in medical research, 16(3):219–242, 2007.
  31. Strategies for imputing missing covariates in accelerated failure time models. Statistics in Medicine, 37(24):3417–3436, 2018. doi:https://doi.org/10.1002/sim.7809. URL https://onlinelibrary.wiley.com/doi/abs/10.1002/sim.7809.
  32. Evaluating the impact of multivariate imputation by mice in feature selection. Plos one, 16(7):e0254720, 2021.
  33. Clinical analytics prediction engine (cape): Development, electronic health record integration and prospective validation of hospital mortality, 180-day mortality and 30-day readmission risk prediction models. PLOS ONE, 15(8):1–15, 08 2020. doi:10.1371/journal.pone.0238065. URL https://doi.org/10.1371/journal.pone.0238065.
  34. Sebastian Pölsterl. scikit-survival: A library for time-to-event analysis built on top of scikit-learn. Journal of Machine Learning Research, 21(212):1–6, 2020. URL http://jmlr.org/papers/v21/20-729.html.
  35. XGBoost Development Team. XGBoost: Extreme gradient boosting. https://pypi.org/project/xgboost/, 2023. Python Package.
  36. Robert Tibshirani. The lasso method for variable selection in the cox model. Statistics in Medicine, 1997.
  37. Regularization and variable selection via the elastic net. Journal of the Royal Statistical Society: Series B, 2005.
  38. Ridge regression: Biased estimation for nonorthogonal problems. Technometrics, 1970.
  39. Robert Tibshirani. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society: Series B, 1996.
  40. Jerome Friedman. Greedy function approximation: A gradient boosting machine. Annals of Statistics, 2001.
  41. Xgboost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016.
  42. Leo Breiman. Random forests. Machine Learning, 2001.
  43. Predicting good probabilities with supervised learning. In Proceedings of the 22nd International Conference on Machine Learning, ICML ’05, page 625–632, New York, NY, USA, 2005. Association for Computing Machinery. ISBN 1595931805. doi:10.1145/1102351.1102430. URL https://doi.org/10.1145/1102351.1102430.
  44. Deep learning model calibration for improving performance in class-imbalanced medical image classification tasks. PLOS ONE, 17(1):1–23, 01 2022. doi:10.1371/journal.pone.0262838. URL https://doi.org/10.1371/journal.pone.0262838.
  45. Evaluating the yield of medical tests. Journal of the American Medical Association, 1982.
  46. Frank E. Harrell. Regression modeling strategies. Springer, 2015.
  47. The meaning and use of the area under a receiver operating characteristic (roc) curve. Radiology, 1982.
  48. Multistate survival models for panel data: The msm package for r. Journal of Statistical Software, 1999.
  49. Shap and lime: An evaluation of discriminative power in credit risk. Frontiers in Artificial Intelligence, page 140, 2021.
  50. A unified approach to interpreting model predictions. In Proceedings of the 31st International Conference on Neural Information Processing Systems, NIPS’17, page 4768–4777. Curran Associates, 2017.
  51. The mds mortality risk index: The evolution of a method for predicting 6-month mortality in nursing home residents. BMC research notes, 3:1–8, 2010.
  52. Mortality-related factors and 1-year survival in nursing home residents. Journal of the American Geriatrics Society, 51(2):213–221, 2003.
  53. Prediction of 6-month survival of nursing home residents with advanced dementia using adept vs hospice eligibility guidelines. Jama, 304(17):1929–1935, 2010.
  54. Estimating prognosis for nursing home residents with advanced dementia. Jama, 291(22):2734–2740, 2004.
  55. Adaptation and initial validation of minimum data set (mds) mortality risk index to mds version 3.0. Journal of the American Geriatrics Society, 66(12):2353–2359, 2018.
  56. A study protocol for the development of a multivariable model predicting 6-and 12-month mortality for people with dementia living in residential aged care facilities (racfs) in australia. Diagnostic and prognostic research, 4:1–8, 2020.
  57. Using mortality risk scores for long-term prognosis of nursing home residents: caution is recommended. Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences, 65(11):1235–1241, 2010.
  58. Prediction of 6-month mortality in nursing home residents with advanced dementia: validity of a risk score. Journal of the American Medical Directors Association, 8(7):464–468, 2007.
  59. Prediction models of all-cause mortality among older adults in nursing home setting: A systematic review and meta-analysis. Health Science Reports, 6(6):e1309, 2023.
  60. interRAI. interRAI Palliative Care (PC) Assessment Form and User’s Manual. Standard English Edition, 9.1.2 edition, 2023. 130-page manual plus 8-page form and 6-page form.
Citations (4)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now