Improving palliative care with deep learning

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 [...].

Generalized Sparse Convolutional Neural Networks for Semantic Segmentation of Point Clouds Derived from Tri-Stereo Satellite Imagery

We studied the applicability of point clouds derived from tri-stereo satellite imagery for semantic segmentation for generalized sparse convolutional neural networks by the example of an Austrian study area. We examined, in particular, if the distorted geometric information, in addition to color, influences the performance of segmenting clutter, roads, buildings, trees, and vehicles. In this regard, we trained a fully convolutional neural network that uses generalized sparse convolution one time solely on 3D geometric information (i.e., 3D point cloud derived by dense image matching), and twice on 3D geometric as well as color information. In the first experiment, we did not use class weights, whereas in the second we did. We compared the results with a fully convolutional neural network that was trained on a 2D orthophoto, and a decision tree that was once trained on hand-crafted 3D geometric features, and once trained on hand-crafted 3D geometric as well as color features. The decision tree using hand-crafted features has been successfully applied to aerial laser scanning data in the literature. Hence, we compared our main interest of study, a representation learning technique, with another representation learning technique, and a non-representation learning technique. Our study area is located in Waldviertel, a region in Lower Austria. The territory is a hilly region covered mainly by forests, agriculture, and grasslands. Our classes of interest are heavily unbalanced. However, we did not use any data augmentation techniques to counter overfitting. For our study area, we reported that geometric and color information only improves the performance of the Generalized Sparse Convolutional Neural Network (GSCNN) on the dominant class, which leads to a higher overall performance in our case. We also found that training the network with median class weighting partially reverts the effects of adding color. The network also started to learn the classes with lower occurrences. The fully convolutional neural network that was trained on the 2D orthophoto generally outperforms the other two with a kappa score of over 90% and an average per class accuracy of 61%. However, the decision tree trained on colors and hand-crafted geometric features has a 2% higher accuracy for roads.

The Application of Deep Learning in Cancer Prognosis Prediction

Deep learning has been applied to many areas in health care, including imaging diagnosis, digital pathology, prediction of hospital admission, drug design, classification of cancer and stromal cells, doctor assistance, etc. Cancer prognosis is to estimate the fate of cancer, probabilities of cancer recurrence and progression, and to provide survival estimation to the patients. The accuracy of cancer prognosis prediction will greatly benefit clinical management of cancer patients. The improvement of biomedical translational research and the application of advanced statistical analysis and machine learning methods are the driving forces to improve cancer prognosis prediction. Recent years, there is a significant increase of computational power and rapid advancement in the technology of artificial intelligence, particularly in deep learning. In addition, the cost reduction in large scale next-generation sequencing, and the availability of such data through open source databases (e.g., TCGA and GEO databases) offer us opportunities to possibly build more powerful and accurate models to predict cancer prognosis more accurately. In this review, we reviewed the most recent published works that used deep learning to build models for cancer prognosis prediction. Deep learning has been suggested to be a more generic model, requires less data engineering, and achieves more accurate prediction when working with large amounts of data. The application of deep learning in cancer prognosis has been shown to be equivalent or better than current approaches, such as Cox-PH. With the burst of multi-omics data, including genomics data, transcriptomics data and clinical information in cancer studies, we believe that deep learning would potentially improve cancer prognosis.

Machine Learning Based Opioid Overdose Prediction Using Electronic Health Records

Opioid addiction in the United States has come to national attention as opioid overdose (OD) related deaths have risen at alarming rates. Combating opioid epidemic becomes a high priority for not only governments but also healthcare providers. This depends on critical knowledge to understand the risk of opioid overdose of patients. In this paper, we present our work on building machine learning based prediction models to predict opioid overdose of patients based on the history of patients' electronic health records (EHR). We performed two studies using New York State claims data (SPARCS) with 440,000 patients and Cerner's Health Facts database with 110,000 patients. Our experiments demonstrated that EHR based prediction can achieve best recall with random forest method (precision: 95.3%, recall: 85.7%, F1 score: 90.3%), best precision with deep learning (precision: 99.2%, recall: 77.8%, F1 score: 87.2%). We also discovered that clinical events are among critical features for the predictions.

Developing Predictive Models to Determine Patients in End-of-Life Care in Administrative Datasets

Introduction

In observational studies with mortality endpoints, one needs to consider how to account for subjects whose interventions appear to be part of ‘end-of-life’ care.

Objective

The objective of this study was to develop a diagnostic predictive model to identify those in end-of-life care at the time of a drug exposure.

Methods

We used data from four administrative claims datasets from 2000 to 2017. The index date was the date of the first prescription for the last new drug subjects received during their observation period. The outcome of end-of-life care was determined by the presence of one or more codes indicating terminal or hospice care. Models were developed using regularized logistic regression. Internal validation was through examination of the area under the receiver operating characteristic curve (AUC) and through model calibration in a 25% subset of the data held back from model training. External validation was through examination of the AUC after applying the model learned on one dataset to the three other datasets.

Results

The models showed excellent performance characteristics. Internal validation resulted in AUCs ranging from 0.918 (95% confidence interval [CI] 0.905–0.930) to 0.983 (95% CI 0.978–0.987) for the four different datasets. Calibration results were also very good, with slopes near unity. External validation also produced very good to excellent performance metrics, with AUCs ranging from 0.840 (95% CI 0.834–0.846) to 0.956 (95% CI 0.952–0.960).

Conclusion

These results show that developing diagnostic predictive models for determining subjects in end-of-life care at the time of a drug treatment is possible and may improve the validity of the risk profile for those treatments.

Electronic supplementary material

The online version of this article (10.1007/s40264-020-00906-7) contains supplementary material, which is available to authorized users.

Prediction of gestational diabetes based on nationwide electronic health records

Gestational diabetes mellitus (GDM) poses increased risk of short- and long-term complications for mother and offspring^1-4. GDM is typically diagnosed at 24-28 weeks of gestation, but earlier detection is desirable as this may prevent or considerably reduce the risk of adverse pregnancy outcomes^5,6. Here we used a machine-learning approach to predict GDM on retrospective data of 588,622 pregnancies in Israel for which comprehensive electronic health records were available. Our models predict GDM with high accuracy even at pregnancy initiation (area under the receiver operating curve (auROC) = 0.85), substantially outperforming a baseline risk score (auROC = 0.68). We validated our results on both a future validation set and a geographical validation set from the most populated city in Israel, Jerusalem, thereby emulating real-world performance. Interrogating our model, we uncovered previously unreported risk factors, including results of previous pregnancy glucose challenge tests. Finally, we devised a simpler model based on just nine questions that a patient could answer, with only a modest reduction in accuracy (auROC = 0.80). Overall, our models may allow early-stage intervention in high-risk women, as well as a cost-effective screening approach that could avoid the need for glucose tolerance tests by identifying low-risk women. Future prospective studies and studies on additional populations are needed to assess the real-world clinical utility of the model.

Axes of a revolution: challenges and promises of big data in healthcare

Health data are increasingly being generated at a massive scale, at various levels of phenotyping and from different types of resources. Concurrent with recent technological advances in both data-generation infrastructure and data-analysis methodologies, there have been many claims that these events will revolutionize healthcare, but such claims are still a matter of debate. Addressing the potential and challenges of big data in healthcare requires an understanding of the characteristics of the data. Here we characterize various properties of medical data, which we refer to as 'axes' of data, describe the considerations and tradeoffs taken when such data are generated, and the types of analyses that may achieve the tasks at hand. We then broadly describe the potential and challenges of using big data in healthcare resources, aiming to contribute to the ongoing discussion of the potential of big data resources to advance the understanding of health and disease.

A fitting machine learning prediction model for short-term mortality following percutaneous catheterization intervention: a nationwide population-based study

Background

A suitable multivariate predictor for predicting mortality following percutaneous coronary intervention (PCI) remains undetermined. We used a nationwide database to construct mortality prediction models to find the appropriate model.

Methods

Data were analyzed from the Taiwan National Health Insurance Research Database (NHIRD) covering the period from 2004 to 2013. The study cohort was composed of 3,421 patients with acute myocardial infarction (AMI) diagnosis undergoing PCI. The dataset of enrolled patients was used to construct multivariate prediction models. Of these, 3,079 and 342 patients were included in the training and test groups, respectively. Each patient had 22 input features and 2 output features that represented mortality. This study implemented an artificial neural network model (ANN), a decision tree (DT), a linear discriminant analysis classifier (LDA), a logistic regression model (LR), a naïve Bayes classifier (NB), and a support vector machine (SVM) to predict post-PCI patient mortality.

Results

The DT model was found to be the most suitable in terms of performance and real-world applicability. The DT model achieved an area under receiving operating characteristic of 0.895 (95% confidence interval: 0.865-0.925), F1 of 0.969, precision of 0.971, and recall of 0.974.

Conclusions

The DT model constructed using data from the NHIRD exhibited effective 30-day mortality prediction for patients with AMI following PCI.

Quality Measurement of Serious Illness Communication: Recommendations for Health Systems Based on Findings from a Symposium of National Experts

Background: Communication between clinicians and patients fundamentally shapes the experience of serious illness. There is increasing recognition that health systems should routinely implement structures and processes to assure high-quality serious illness communication (SIC) and measure the effectiveness of their efforts on key outcomes. The absence, underdevelopment, or limited applicability of quality measures related specifically to SIC, and their limited application only to those seen by specialist palliative and hospice care teams, hinder efforts to improve care planning, service delivery, and health outcomes for all seriously ill patients. Objective: We convened an expert stakeholder symposium and subsequently surveyed participants to consider challenges, opportunities, priorities, and strategies to improve quality measurement specific to SIC. Results: We identified several barriers and opportunities to improving quality measurement of SIC. These include issues related to the definition of SIC, methodological challenges related to measuring SIC and related outcomes, underutilization of technologies that can facilitate measurement, and measurement development, and dissemination. Conclusions: Patients, clinicians, and health systems increasingly align around the importance of high-quality communication in serious illness. We offer recommendations for various stakeholder groups to advance SIC quality measurement. Enthusiasm and a sense of urgency among health systems to drive and measure communication improvements inform our proposal for a set of example measures for implementation now.

Human factors challenges for the safe use of artificial intelligence in patient care

The use of artificial intelligence (AI) in patient care can offer significant benefits. However, there is a lack of independent evaluation considering AI in use. The paper argues that consideration should be given to how AI will be incorporated into clinical processes and services. Human factors challenges that are likely to arise at this level include cognitive aspects (automation bias and human performance), handover and communication between clinicians and AI systems, situation awareness and the impact on the interaction with patients. Human factors research should accompany the development of AI from the outset.

Assessment of Deep Learning Using Nonimaging Information and Sequential Medical Records to Develop a Prediction Model for Nonmelanoma Skin Cancer

Importance

A prediction model for new-onset nonmelanoma skin cancer could enhance prevention measures, but few patient data-driven tools exist for more accurate prediction.

Objective

To use machine learning to develop a prediction model for incident nonmelanoma skin cancer based on large-scale, multidimensional, nonimaging medical information.

Design, Setting, and Participants

This study used a database comprising 2 million randomly sampled patients from the Taiwan National Health Insurance Research Database from January 1, 1999, to December 31, 2013. A total of 1829 patients with nonmelanoma skin cancer as their first diagnosed cancer and 7665 random controls without cancer were included in the analysis. A convolutional neural network, a deep learning approach, was used to develop a risk prediction model. This risk prediction model used 3-year clinical diagnostic information, medical records, and temporal-sequential information to predict the skin cancer risk of a given patient within the next year. Stepwise feature selection was also performed to investigate important and determining factors of the model. Statistical analysis was performed from November 1, 2016, to October 31, 2018.

Main Outcomes and Measures

Sensitivity, specificity, and area under the receiver operating characteristic (AUROC) curve were used to evaluate the performance of the models.

Results

A total of 1829 patients (923 women [50.5%] and 906 men [49.5%]; mean [SD] age, 65.3 [15.7] years) with nonmelanoma skin cancer and 7665 random controls without cancer (3951 women [51.5%] and 3714 men [48.4%]; mean [SD] age, 47.5 [17.3] years) were included in the analysis. The 1-year incident nonmelanoma skin cancer risk prediction model using sequential diagnostic information and drug prescription information as a time-incorporated feature matrix could attain an AUROC of 0.89 (95% CI, 0.87-0.91), with a mean (SD) sensitivity of 83.1% (3.5%) and mean (SD) specificity of 82.3% (4.1%). Carcinoma in situ of skin (AUROC, 0.867; -2.80% loss) and other chronic comorbidities (eg, degenerative osteopathy [AUROC, 0.872; -2.32% loss], hypertension [AUROC, 0.879; -1.53% loss], and chronic kidney insufficiency [AUROC, 0.879; -1.52% loss]) served as more discriminative factors for the prediction. Medications such as trazodone, acarbose, systemic antifungal agents, statins, nonsteroidal anti-inflammatory drugs, and thiazide diuretics were the top-ranking discriminative features in the model; each led to more than a 1% decrease of the AUROC when eliminated individually (eg, trazodone AUROC, 0.868; -2.67% reduction; acarbose AUROC, 0.870; -2.50 reduction; and systemic antifungal agents AUROC, 0.875; -1.99 reduction).

Conclusions and Relevance

The findings of this study suggest that a risk prediction model may have potential predictive factors for nonmelanoma skin cancer. This model may help health care professionals target high-risk populations for more intensive skin cancer preventive methods.

Economic Analysis of Hospital Palliative Care: Investigating Heterogeneity by Noncancer Diagnoses

Background. Single-disease-focused treatment and hospital-centric care are poorly suited to meet complex needs in an era of multimorbidity. Understanding variation in palliative care’s association with treatment choices is essential to optimizing interdisciplinary decision making in care of complex patients. Aim. To estimate the association between palliative care and hospital costs by primary diagnosis and multimorbidity for adults with one of six life-limiting conditions: heart failure, chronic obstructive pulmonary disease (COPD), liver failure, kidney failure, neurodegenerative conditions including dementia, and HIV/AIDS. Methods. Data from four studies (2002–2015) were pooled to provide an analytic dataset of 73,304 participants with mean costs $10,483, of whom 5,348 (7%) received palliative care. We estimated average effect of palliative care on direct hospital costs among the treated, using propensity scores to control for observed confounding. Results. Palliative care was associated with a statistically significant reduction in total direct costs for heart failure (estimated treatment effect: −$2666; 95% confidence interval [CI]: −$3440 to −$1892), neurodegenerative conditions (−$3523; −$4394 to −$2651), COPD (−$1613; −$2217 to −$1009), kidney failure (−$3589; −$5132 to −$2045), and liver failure (−$7574; −$9232 to −$5916). The association for liver failure patients was statistically significantly larger than for any other disease group. Cost-saving associations were also statistically larger for patients with multimorbidity than single disease for two of the six groups: neurodegenerative and liver failure. Conclusions. Heterogeneity in treatment effect estimates was observable in assessing association between palliative care and hospital costs for adults with serious life-limiting illnesses other than cancer. The results illustrate the importance of careful definition of palliative care populations in research and practice, and raise further questions about the role of interdisciplinary decision making in treatment of complex medical illness.

Electronic Health Record Mortality Prediction Model for Targeted Palliative Care Among Hospitalized Medical Patients: a Pilot Quasi-experimental Study

BACKGROUND

Development of electronic health record (EHR) prediction models to improve palliative care delivery is on the rise, yet the clinical impact of such models has not been evaluated.

OBJECTIVE

To assess the clinical impact of triggering palliative care using an EHR prediction model.

DESIGN

Pilot prospective before-after study on the general medical wards at an urban academic medical center.

PARTICIPANTS

Adults with a predicted probability of 6-month mortality of ≥ 0.3.

INTERVENTION

Triggered (with opt-out) palliative care consult on hospital day 2.

MAIN MEASURES

Frequencies of consults, advance care planning (ACP) documentation, home palliative care and hospice referrals, code status changes, and pre-consult length of stay (LOS).

KEY RESULTS

The control and intervention periods included 8 weeks each and 138 admissions and 134 admissions, respectively. Characteristics between the groups were similar, with a mean (standard deviation) risk of 6-month mortality of 0.5 (0.2). Seventy-seven (57%) triggered consults were accepted by the primary team and 8 consults were requested per usual care during the intervention period. Compared to historical controls, consultation increased by 74% (22 [16%] vs 85 [63%], P < .001), median (interquartile range) pre-consult LOS decreased by 1.4 days (2.6 [1.1, 6.2] vs 1.2 [0.8, 2.7], P = .02), ACP documentation increased by 38% (23 [17%] vs 37 [28%], P = .03), and home palliative care referrals increased by 61% (9 [7%] vs 23 [17%], P = .01). There were no differences between the control and intervention groups in hospice referrals (14 [10] vs 22 [16], P = .13), code status changes (42 [30] vs 39 [29]; P = .81), or consult requests for lower risk (< 0.3) patients (48/1004 [5] vs 33/798 [4]; P = .48).

CONCLUSIONS

Targeting hospital-based palliative care using an EHR mortality prediction model is a clinically promising approach to improve the quality of care among seriously ill medical patients. More evidence is needed to determine the generalizability of this approach and its impact on patient- and caregiver-reported outcomes.

Providing Data-Driven Equitable Palliative and End-of-Life Care for Structurally Vulnerable Populations: A Pilot Survey of Information Management Strategies

BACKGROUND

Considerable gains are being made in data-driven efforts to advance quality improvement in health care. However, organizations providing hospice-oriented palliative care for structurally vulnerable persons with terminal illnesses may not have the enabling data infrastructure or framework to derive such benefits.

METHODS

We conducted a pilot cross-sectional qualitative study involving a convenience sample of hospice organizations across North America providing palliative care services for structurally vulnerable patients. Through semistructured interviews, we surveyed organizations on the types of data collected, the information systems used, and the challenges they faced.

RESULTS

We contacted 13 organizations across North America and interviewed 9. All organizations served structurally vulnerable populations, including the homeless and vulnerably housed, socially isolated, and HIV-positive patients. Common examples of collected data included the number of referrals, the number of admissions, length of stay, and diagnosis. More than half of the organizations (n = 5) used an electronic medical record, although none of the record systems were specifically designed for palliative care. All (n = 9) the organizations used the built-in reporting capacity of their information management systems and more than half (n = 6) augmented this capacity with chart reviews.

DISCUSSION

A number of themes emerged from our discussions. Present data collection is heterogeneous, and storage of these data is highly fragmented within and across organizations. Funding appeared to be a key enabler of more robust data collection and use. Future work should address these gaps and examine opportunities for innovative ways of analysis and reporting to improve care for structurally vulnerable populations.

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.

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.

Development and Validation of a Deep Learning Algorithm for Mortality Prediction in Selecting Patients With Dementia for Earlier Palliative Care Interventions

Importance

Early palliative care interventions drive high-value care but currently are underused. Health care professionals face challenges in identifying patients who may benefit from palliative care.

Objective

To develop a deep learning algorithm using longitudinal electronic health records to predict mortality risk as a proxy indicator for identifying patients with dementia who may benefit from palliative care.

Design, Setting, and Participants

In this retrospective cohort study, 6-month, 1-year, and 2-year mortality prediction models with recurrent neural networks used patient demographic information and topics generated from clinical notes within Partners HealthCare System, an integrated health care delivery system in Boston, Massachusetts. This study included 26 921 adult patients with dementia who visited the health care system from January 1, 2011, through December 31, 2017. The models were trained using a data set of 24 229 patients and validated using another data set of 2692 patients. Data were analyzed from September 18, 2018, to May 15, 2019.

Main Outcomes and Measures

The area under the receiver operating characteristic curve (AUC) for 6-month and 1- and 2-year mortality prediction models and the factors contributing to the predictions.

Results

The study cohort included 26 921 patients (16 263 women [60.4%]; mean [SD] age, 74.6 [13.5] years). For the 24 229 patients in the training data set, mean (SD) age was 74.8 (13.2) years and 14 632 (60.4%) were women. For the 2692 patients in the validation data set, mean (SD) age was 75.0 (12.6) years and 1631 (60.6%) were women. The 6-month model reached an AUC of 0.978 (95% CI, 0.977-0.978); the 1-year model, 0.956 (95% CI, 0.955-0.956); and the 2-year model, 0.943 (95% CI, 0.942-0.944). The top-ranked latent topics associated with 6-month and 1- and 2-year mortality in patients with dementia include palliative and end-of-life care, cognitive function, delirium, testing of cholesterol levels, cancer, pain, use of health care services, arthritis, nutritional status, skin care, family meeting, shock, respiratory failure, and swallowing function.

Conclusions and Relevance

A deep learning algorithm based on patient demographic information and longitudinal clinical notes appeared to show promising results in predicting mortality among patients with dementia in different time frames. Further research is necessary to determine the feasibility of applying this algorithm in clinical settings for identifying unmet palliative care needs earlier.

Timing of Palliative Care in Colorectal Cancer Patients: Does It Matter?

BACKGROUND

Palliative care can improve end-of-life care and reduce health care expenditures, but the optimal timing for initiation remains unclear. We sought to characterize the association between timing of palliative care, in-hospital deaths, and health care costs.

METHODS

This is a retrospective cohort study including all patients who were diagnosed and died of colorectal cancer between 2004 and 2012 in Manitoba, Canada. The primary exposure was timing of palliative care, defined as no involvement, late involvement (less than 14 d before death), early involvement (14 to 60 d before death), and very early involvement (>60 d before death). The primary outcome was in-hospital deaths and end-of-life health care costs.

RESULTS

A total of 1607 patients were included; 315 (20%) received palliative care and 162 (10%) died in hospital. Compared to those who did not receive palliative care, patients with early and very early involvement experienced significantly decreased odds of dying in hospital (OR 0.21 95% CI 0.06-0.69 P = 0.01 and OR 0.11 95% CI 0.01-0.78 P = 0.03, respectively) and significantly lower health care costs. There were no significant differences in in-hospital deaths and health care costs between patients without palliative care and those who received late palliative care.

CONCLUSIONS

Early palliative care involvement is associated with decreased odds of dying in hospital and lower health care utilization and costs in patients with colorectal cancer. These findings provide real-world evidence supporting early integration of palliative care, although the optimal timing (early versus very early) remains a matter of debate.

Dynamic-DeepHit: A Deep Learning Approach for Dynamic Survival Analysis With Competing Risks Based on Longitudinal Data

Currently available risk prediction methods are limited in their ability to deal with complex, heterogeneous, and longitudinal data such as that available in primary care records, or in their ability to deal with multiple competing risks. This paper develops a novel deep learning approach that is able to successfully address current limitations of standard statistical approaches such as landmarking and joint modeling. Our approach, which we call Dynamic-DeepHit, flexibly incorporates the available longitudinal data comprising various repeated measurements (rather than only the last available measurements) in order to issue dynamically updated survival predictions for one or multiple competing risk(s). Dynamic-DeepHit learns the time-to-event distributions without the need to make any assumptions about the underlying stochastic models for the longitudinal and the time-to-event processes. Thus, unlike existing works in statistics, our method is able to learn data-driven associations between the longitudinal data and the various associated risks without underlying model specifications. We demonstrate the power of our approach by applying it to a real-world longitudinal dataset from the U.K. Cystic Fibrosis Registry, which includes a heterogeneous cohort of 5883 adult patients with annual follow-ups between 2009 to 2015. The results show that Dynamic-DeepHit provides a drastic improvement in discriminating individual risks of different forms of failures due to cystic fibrosis. Furthermore, our analysis utilizes post-processing statistics that provide clinical insight by measuring the influence of each covariate on risk predictions and the temporal importance of longitudinal measurements, thereby enabling us to identify covariates that are influential for different competing risks.

Predicting life expectancy with a long short-term memory recurrent neural network using electronic medical records

BACKGROUND

Life expectancy is one of the most important factors in end-of-life decision making. Good prognostication for example helps to determine the course of treatment and helps to anticipate the procurement of health care services and facilities, or more broadly: facilitates Advance Care Planning. Advance Care Planning improves the quality of the final phase of life by stimulating doctors to explore the preferences for end-of-life care with their patients, and people close to the patients. Physicians, however, tend to overestimate life expectancy, and miss the window of opportunity to initiate Advance Care Planning. This research tests the potential of using machine learning and natural language processing techniques for predicting life expectancy from electronic medical records.

METHODS

We approached the task of predicting life expectancy as a supervised machine learning task. We trained and tested a long short-term memory recurrent neural network on the medical records of deceased patients. We developed the model with a ten-fold cross-validation procedure, and evaluated its performance on a held-out set of test data. We compared the performance of a model which does not use text features (baseline model) to the performance of a model which uses features extracted from the free texts of the medical records (keyword model), and to doctors' performance on a similar task as described in scientific literature.

RESULTS

Both doctors and the baseline model were correct in 20% of the cases, taking a margin of 33% around the actual life expectancy as the target. The keyword model, in comparison, attained an accuracy of 29% with its prognoses. While doctors overestimated life expectancy in 63% of the incorrect prognoses, which harms anticipation to appropriate end-of-life care, the keyword model overestimated life expectancy in only 31% of the incorrect prognoses.

CONCLUSIONS

Prognostication of life expectancy is difficult for humans. Our research shows that machine learning and natural language processing techniques offer a feasible and promising approach to predicting life expectancy. The research has potential for real-life applications, such as supporting timely recognition of the right moment to start Advance Care Planning.

Privacy in the age of medical big data

Big data has become the ubiquitous watch word of medical innovation. The rapid development of machine-learning techniques and artificial intelligence in particular has promised to revolutionize medical practice from the allocation of resources to the diagnosis of complex diseases. But with big data comes big risks and challenges, among them significant questions about patient privacy. Here, we outline the legal and ethical challenges big data brings to patient privacy. We discuss, among other topics, how best to conceive of health privacy; the importance of equity, consent, and patient governance in data collection; discrimination in data uses; and how to handle data breaches. We close by sketching possible ways forward for the regulatory system.

Palliative Care: Taking the Long View

Our medicalised modern cultures render reason and mystery mutually exclusive, define death by disease as failure, and dying as disgraceful. Providers and policymakers alike marginalize aging and dying individuals, formulating largely ineffective strategies without palliative care and pain relief budgets. The aim of palliative care is to support the person with incurable illness to live their remaining life as well and as meaningfully as possible and to support them as they eventually die from their illness and reaching the natural end of their lives It acknowledges that each life is morally significant, restoring patients' and families' quality of life where possible, and attending meticulously to the dying period as necessary (Saunders, 1965). Hospices are far more than mere buildings; they house an ethos of care. The field is currently challenged by its variable situation over the world and the pressing need to incorporate new technology to its practice. This article provides a review of some important milestones in the history and development of Palliative Care and evolution of Palliative Medicine in some countries, some current issues concerning consistency in its implementation, and some likely prospects for its future advance and expected expansion, from the perspective of one central question: "What constitutes the ethos of Palliative Care replicating its foundational philosophy and principles?" which helps to set the scene for possible future advances to integrate ethical, legal, and social implications. Technology will help expansion by facilitating communication and predicting needs.

#Deathbedlive: the end-of-life trajectory, reflected in a cancer patient's tweets

Background

Understanding physical and psycho-social illness trajectories towards the end of life can help in the planning of palliative and supportive care. With terminal patients increasingly seeking and sharing health information and support via social media, it is timely to examine whether these trajectories are reflected in their digital narratives. In this exploratory study, we analysed the Twitter feed of prominent cancer sufferer and physician, Kate Granger, over the final 6 months of her life.

Methods

With the consent of Kate’s widower, Chris Pointon, 1628 Twitter posts from @GrangerKate were manually screened. The 550 tweets judged relevant to her disease were qualitatively content analysed with reference to the six modifiable dimensions of the patient experience in Emanuel and Emanuel’s ‘framework for a good death’. The frequency of each tweet category was charted over time and textual content was examined and cross-referenced with key events, to obtain a deeper understanding of its nature and significance.

Results

Tweets were associated with physical symptoms (N = 270), psychological and cognitive symptoms (N = 213), social relationships and support (N = 85), economic demands and care giving needs (N = 85), hopes and expectations (N = 51) and spiritual beliefs (N = 7). While medical treatments and procedures were discussed in detail, medical information-seeking was largely absent, likely reflecting Kate clinical expertise. Spirituality was expressed more as hope in treatments or “someone out there listening”, than in religious terms. The high value of Kate’s palliative care team was a dominant theme in the support category, alongside the support she received from her online community of fellow sufferers, friends, family and colleagues. Significant events, such as medical procedures and hospital stays generated the densest Twitter engagement. Transitions between trajectory phases were marked by changes in the relative frequency of tweet-types.

Conclusions

In Kate’s words, “the power of patient narrative cannot be underestimated”. While this analysis spanned only 6 months, it yielded rich insights. The results reflect theorised end-of-life dimensions and reveal the potential of social media data and digital bio-ethnography to shine a light on terminal patients’ lived experiences, coping strategies and support needs, suggesting new opportunities for enhancing personalised palliative care and avenues for further research.