Improving palliative and end-of-life care with machine learning and routine data: a rapid review

Predicting Patient Mortality for Earlier Palliative Care Identification in Medicare Advantage Plans: Features of a Machine Learning Model

BACKGROUND

Machine learning (ML) can offer greater precision and sensitivity in predicting Medicare patient end of life and potential need for palliative services compared to provider recommendations alone. However, earlier ML research on older community dwelling Medicare beneficiaries has provided insufficient exploration of key model feature impacts and the role of the social determinants of health.

OBJECTIVE

This study describes the development of a binary classification ML model predicting 1-year mortality among Medicare Advantage plan members aged ≥65 years (N=318,774) and further examines the top features of the predictive model.

METHODS

A light gradient-boosted trees model configuration was selected based on 5-fold cross-validation. The model was trained with 80% of cases (n=255,020) using randomized feature generation periods, with 20% (n=63,754) reserved as a holdout for validation. The final algorithm used 907 feature inputs extracted primarily from claims and administrative data capturing patient diagnoses, service utilization, demographics, and census tract-based social determinants index measures.

RESULTS

The total sample had an actual mortality prevalence of 3.9% in the 2018 outcome period. The final model correctly predicted 44.2% of patient expirations among the top 1% of highest risk members (AUC=0.84; 95% CI 0.83-0.85) versus 24.0% predicted by the model iteration using only age, gender, and select high-risk utilization features (AUC=0.74; 95% CI 0.73-0.74). The most important algorithm features included patient demographics, diagnoses, pharmacy utilization, mean costs, and certain social determinants of health.

CONCLUSIONS

The final ML model better predicts Medicare Advantage member end of life using a variety of routinely collected data and supports earlier patient identification for palliative care.

Predicting mortality in the very old: a machine learning analysis on claims data

Machine learning (ML) may be used to predict mortality. We used claims data from one large German insurer to develop and test differently complex ML prediction models, comparing them for their (balanced) accuracy, but also the importance of different predictors, the relevance of the follow-up period before death (i.e. the amount of accumulated data) and the time distance of the data used for prediction and death. A sample of 373,077 insured very old, aged 75 years or above, living in the Northeast of Germany in 2012 was drawn and followed over 6 years. Our outcome was whether an individual died in one of the years of interest (2013-2017) or not; the primary metric was (balanced) accuracy in a hold-out test dataset. From the 86,326 potential variables, we used the 30 most important ones for modeling. We trained a total of 45 model combinations: (1) Three different ML models were used; logistic regression (LR), random forest (RF), extreme gradient boosting (XGB); (2) Different periods of follow-up were employed for training; 1-5 years; (3) Different time distances between data used for prediction and the time of the event (death/survival) were set; 0-4 years. The mortality rate was 9.15% in mean per year. The models showed (balanced) accuracy between 65 and 93%. A longer follow-up period showed limited to no advantage, but models with short time distance from the event were more accurate than models trained on more distant data. RF and XGB were more accurate than LR. For RF and XGB sensitivity and specificity were similar, while for LR sensitivity was significantly lower than specificity. For all three models, the positive-predictive-value was below 62% (and even dropped to below 20% for longer time distances from death), while the negative-predictive-value significantly exceeded 90% for all analyses. The utilization of and costs for emergency transport as well as emergency and any hospital visits as well as the utilization of conventional outpatient care and laboratory services were consistently found most relevant for predicting mortality. All models showed useful accuracies, and more complex models showed advantages. The variables employed for prediction were consistent across models and with medical reasoning. Identifying individuals at risk could assist tailored decision-making and interventions.

Contribution of Synthetic Data Generation towards an Improved Patient Stratification in Palliative Care

AI model development for synthetic data generation to improve Machine Learning (ML) methodologies is an integral part of research in Computer Science and is currently being transferred to related medical fields, such as Systems Medicine and Medical Informatics. In general, the idea of personalized decision-making support based on patient data has driven the motivation of researchers in the medical domain for more than a decade, but the overall sparsity and scarcity of data are still major limitations. This is in contrast to currently applied technology that allows us to generate and analyze patient data in diverse forms, such as tabular data on health records, medical images, genomics data, or even audio and video. One solution arising to overcome these data limitations in relation to medical records is the synthetic generation of tabular data based on real world data. Consequently, ML-assisted decision-support can be interpreted more conveniently, using more relevant patient data at hand. At a methodological level, several state-of-the-art ML algorithms generate and derive decisions from such data. However, there remain key issues that hinder a broad practical implementation in real-life clinical settings. In this review, we will give for the first time insights towards current perspectives and potential impacts of using synthetic data generation in palliative care screening because it is a challenging prime example of highly individualized, sparsely available patient information. Taken together, the reader will obtain initial starting points and suitable solutions relevant for generating and using synthetic data for ML-based screenings in palliative care and beyond.

Using predicted length of stay to define treatment and model costs in hospitalized adults with serious illness: an evaluation of palliative care

BACKGROUND

Economic research on hospital palliative care faces major challenges. Observational studies using routine data encounter difficulties because treatment timing is not under investigator control and unobserved patient complexity is endemic. An individual's predicted LOS at admission offers potential advantages in this context.

METHODS

We conducted a retrospective cohort study on adults admitted to a large cancer center in the United States between 2009 and 2015. We defined a derivation sample to estimate predicted LOS using baseline factors (N = 16,425) and an analytic sample for our primary analyses (N = 2674) based on diagnosis of a terminal illness and high risk of hospital mortality. We modelled our treatment variable according to the timing of first palliative care interaction as a function of predicted LOS, and we employed predicted LOS as an additional covariate in regression as a proxy for complexity alongside diagnosis and comorbidity index. We evaluated models based on predictive accuracy in and out of sample, on Akaike and Bayesian Information Criteria, and precision of treatment effect estimate.

RESULTS

Our approach using an additional covariate yielded major improvement in model accuracy: R² increased from 0.14 to 0.23, and model performance also improved on predictive accuracy and information criteria. Treatment effect estimates and conclusions were unaffected. Our approach with respect to treatment variable yielded no substantial improvements in model performance, but post hoc analyses show an association between treatment effect estimate and estimated LOS at baseline.

CONCLUSION

Allocation of scarce palliative care capacity and value-based reimbursement models should take into consideration when and for whom the intervention has the largest impact on treatment choices. An individual's predicted LOS at baseline is useful in this context for accurately predicting costs, and potentially has further benefits in modelling treatment effects.

Machine learning in oncology—Perspectives in patient-reported outcome research

Background

Increasing data volumes in oncology pose new challenges for data analysis. Machine learning, a branch of artificial intelligence, can identify patterns even in very large and less structured datasets.

Objective

This article provides an overview of the possible applications for machine learning in oncology. Furthermore, the potential of machine learning in patient-reported outcome (PRO) research is discussed.

Materials and methods

We conducted a selective literature search (PubMed, MEDLINE, IEEE Xplore) and discuss current research.

Results

There are three primary applications for machine learning in oncology: (1) cancer detection or classification; (2) overall survival prediction or risk assessment; and (3) supporting therapy decision-making and prediction of treatment response. Generally, machine learning approaches in oncology PRO research are scarce and few studies integrate PRO data into machine learning models.

Discussion

Machine learning is a promising area of oncology, but few models have been transferred into clinical practice. The promise of personalized cancer therapy and shared decision-making through machine learning has yet to be realized. As an equally important emerging research area in oncology, PROs should also be incorporated into machine learning approaches. To gather the data necessary for this, broad implementation of PRO assessments in clinical practice, as well as the harmonization of existing datasets, is suggested.

Leveraging Advances in Artificial Intelligence to Improve the Quality and Timing of Palliative Care

In recent years, research on artificial intelligence (AI) in medicine has seen great advances, especially with regards to the detection of diseases [...].

Prediction model for short-term mortality after palliative radiotherapy for patients having advanced cancer: a cohort study from routine electronic medical data

We developed a predictive score system for 30-day mortality after palliative radiotherapy by using predictors from routine electronic medical record. Patients with metastatic cancer receiving first course palliative radiotherapy from 1 July, 2007 to 31 December, 2017 were identified. 30-day mortality odds ratios and probabilities of the death predictive score were obtained using multivariable logistic regression model. Overall, 5,795 patients participated. Median follow-up was 39.6 months (range, 24.5-69.3) for all surviving patients. 5,290 patients died over a median 110 days, of whom 995 (17.2%) died within 30 days of radiotherapy commencement. The most important mortality predictors were primary lung cancer (odds ratio: 1.73, 95% confidence interval: 1.47-2.04) and log peripheral blood neutrophil lymphocyte ratio (odds ratio: 1.71, 95% confidence interval: 1.52-1.92). The developed predictive scoring system had 10 predictor variables and 20 points. The cross-validated area under curve was 0.81 (95% confidence interval: 0.79-0.82). The calibration suggested a reasonably good fit for the model (likelihood-ratio statistic: 2.81, P = 0.094), providing an accurate prediction for almost all 30-day mortality probabilities. The predictive scoring system accurately predicted 30-day mortality among patients with stage IV cancer. Oncologists may use this to tailor palliative therapy for patients.

Population-based palliative care planning in Ireland: how many people will live and die with serious illness to 2046?

Background: All countries face growing demand for palliative care services.  Projections of need are essential to plan care in an era of demographic change.   We aim to estimate palliative care needs in the Republic of Ireland from 2016 to 2046. Methods: Static modelling of secondary data.  First, we estimate the numbers of people who will die from a disease associated with palliative care need.  We combine government statistics on cause of death (2007-2015) and projected mortality (2016-2046).  Second, we combine these statistics with survey data to estimate numbers of people aged 50+ living and dying with diseases associated with palliative care need.  Third, we use these projections and survey data to estimate disability burden, pain prevalence and health care utilisation among people aged 50+ living and dying with serious medical illness. Results: In 2016, the number of people dying annually from a disease indicating palliative care need was estimated as 22,806, and the number of people not in the last year of life aged 50+ with a relevant diagnosis was estimated as 290,185.  Equivalent estimates for 2046 are up to 40,355 and 548,105, increases of 84% and 89% respectively.  These groups account disproportionately for disability burden, pain prevalence and health care use among older people, meaning that population health burdens and health care use will increase significantly in the next three decades. Conclusion: The global population is ageing, although significant differences in intensity of ageing can be seen between countries. Prevalence of palliative care need will nearly double over 30 years, reflecting Ireland’s relatively young population.  Older people living with a serious disease outnumber those in the last year of life by approximately 12:1, necessitating implementation of integrated palliative care across the disease trajectory. Urgent steps on funding, workforce development and service provision are required to address these challenges.

Population-based palliative care planning in Ireland: how many people will live and die with serious illness to 2046?

Background: All countries face growing demand for palliative care services.  Projections of need are essential to plan care in an era of demographic change.   We aim to estimate palliative care needs in Ireland from 2016 to 2046. Methods: Static modelling of secondary data.  First, we estimate the numbers of people in Ireland who will die from a disease associated with palliative care need.  We combine government statistics on cause of death (2007-2015) and projected mortality (2016-2046).  Second, we combine these statistics with survey data to estimate numbers of people aged 50+ living and dying with diseases associated with palliative care need.  Third, we use these projections and survey data to estimate disability burden, pain prevalence and health care utilisation among people aged 50+ living and dying with serious medical illness. Results: In 2016, the number of people dying annually from a disease indicating palliative care need was estimated as 22,806, and the number of people not in the last year of life aged 50+ with a relevant diagnosis was estimated as 290,185.  Equivalent estimates for 2046 are 40,355 and 548,105, increases of 84% and 89% respectively.  These groups account disproportionately for disability burden, pain prevalence and health care use among older people, meaning that population health burdens and health care use will increase significantly in the next three decades. Conclusion: The global population is ageing, although significant differences in intensity of ageing can be seen between countries. Prevalence of palliative care need in Ireland will nearly double over 30 years, reflecting Ireland's relatively young population.  People living with a serious disease outnumber those in the last year of life by approximately 12:1, necessitating implementation of integrated palliative care across the disease trajectory. Urgent steps on funding, workforce development and service provision are required to address these challenges.

Improving palliative and end-of-life care with machine learning and routine data: a rapid review

Introduction: Improving end-of-life (EOL) care is a priority worldwide as this population experiences poor outcomes and accounts disproportionately for costs. In clinical practice, physician judgement is the core method of identifying EOL care needs but has important limitations. Machine learning (ML) is a subset of artificial intelligence advancing capacity to identify patterns and make predictions using large datasets.  ML approaches have the potential to improve clinical decision-making and policy design, but there has been no systematic assembly of current evidence. Methods: We conducted a rapid review, searching systematically seven databases from inception to December 31st, 2018: EMBASE, MEDLINE, Cochrane Library, PsycINFO, WOS, SCOPUS and ECONLIT.  We included peer-reviewed studies that used ML approaches on routine data to improve palliative and EOL care for adults.  Our specified outcomes were survival, quality of life (QoL), place of death, costs, and receipt of high-intensity treatment near end of life.  We did not search grey literature and excluded material that was not a peer-reviewed article. Results: The database search identified 426 citations. We discarded 162 duplicates and screened 264 unique title/abstracts, of which 22 were forwarded for full text review.  Three papers were included, 18 papers were excluded and one full text was sought but unobtainable.  One paper predicted six-month mortality, one paper predicted 12-month mortality and one paper cross-referenced predicted 12-month mortality with healthcare spending.  ML-informed models outperformed logistic regression in predicting mortality but poor prognosis is a weak driver of costs.  Models using only routine administrative data had limited benefit from ML methods. Conclusion: While ML can in principle help to identify those at risk of adverse outcomes and inappropriate treatment near EOL, applications to policy and practice are formative.  Future research must not only expand scope to other outcomes and longer timeframes, but also engage with individual preferences and ethical challenges.