Development of a predictive model for integrated medical and long-term care resource consumption based on health behaviour

Development of a predictive model for integrated medical and long-term care resource consumption based on health behaviour: application of healthcare big data of patients with circulatory diseases.
BMC Med 19, 15 (2021). https://doi.org/10.1186/s12916-020-01874-6

• A predictive model of integrated healthcare resource consumption (Adherence Score for Healthcare Resource Outcome, ASHRO) that incorporates patient health behaviours, and examined its association with clinical outcomes.

COVID-19 epidemic prediction and the impact of public health interventions: A review of COVID-19 epidemic models

COVID-19 epidemic prediction and the impact of public health interventions: A review of COVID-19 epidemic models.
Infect Dis Model. 2021;6:324-342. doi: 10.1016/j.idm.2021.01.001. Epub 2021 Jan 7. 

• A systematic review of epidemic prediction models of COVID-19 and the public health intervention strategies identified 55 studies. The input epidemiological parameters of the prediction models had significant differences in the prediction of the severity of the epidemic spread.

Cancer diagnostic tools to aid decision-making in primary care:

Cancer diagnostic tools to aid decision-making in primary care: mixed-methods systematic reviews and cost-effectiveness analysis
Health Technol Assess 2020;24(66) November 2020

• Systematic reviews to examine the clinical effectiveness and the development, validation and accuracy of diagnostic prediction models for aiding general practitioners in cancer diagnosis identified little good-quality evidence. Many diagnostic prediction models are limited by a lack of external validation. 
• A survey of general practitioners/ practice level staff found that cancer decision support tools were available in 83 out of 227 practices, and were likely to be used in 38 out of 227 practices.

Breast cancer risk models and tools

Breast cancer risk models and tools
PHG Foundation 2020

• An “explainer” based on work for the report Personalising breast cancer prevention - bridging the gap between research policy.

Citizen generated data and health Predictive prevention of disease

Citizen generated data and health: Predictive prevention of disease
PHG Foundation November 2020

• Citizen generated data (CGD) is a form of data generated by citizens outside formal health systems that can nevertheless provide insights into health and wellbeing, and that could potentially be harnessed for disease monitoring and treatment. The depth and variety of CGD covers the full spectrum of the health sector, from public health and social care to primary and hospital-based care. We outline essential policy considerations in order to support the development of cooperative strategies for optimising the future use of CGD for predictive prevention.

Consistency of variety of machine learning and statistical models in predicting clinical risks of individual patients

Consistency of variety of machine learning and statistical models in predicting clinical risks of individual patients: longitudinal cohort study using cardiovascular disease as exemplar
BMJ 2020; 371 :m3919, 4 November 2020

• Research using data from the UK Clinical Practice Research Datalink has tested the consistency of 19 machine learning and statistical techniques in predicting individual level and population level risks of cardiovascular disease. The predictions varied widely between and within different types of machine learning and statistical models, especially in patients with higher risks.

Early prognostication of COVID-19 to guide hospitalisation versus outpatient monitoring using a point-of-test risk prediction score

Early prognostication of COVID-19 to guide hospitalisation versus outpatient monitoring using a point-of-test risk prediction score
medRxiv 2020.10.19.20215426; doi: https://doi.org/10.1101/2020.10.19.20215426 [Preprint, not peer reviewed] 

• A five-predictor score termed SOARS (SpO2, Obesity, Age, Respiratory rate, Stroke history) has been developed through sequential modelling to correlate COVID-19 severity across low, moderate and high strata of mortality risk. The tool was based on a cohort of COVID-19 patients presenting at Watford Hospital, West Herts NHS Hospitals Trust March – May 2020.

Development of the COVID-AID risk tool

Development and external validation of a prediction risk model for short-term mortality among hospitalized U.S. COVID-19 patients: A proposal for the COVID-AID risk tool.
PLoS One. 2020 Sep 30;15(9):e0239536. doi: 10.1371/journal.pone.0239536.

• Development of the COVID-AID risk tool that demonstrates accuracy in the prediction of both 7-day and 14-day mortality risk among patients hospitalized with COVID-19. This prediction score could assist with resource utilization, patient and caregiver education, and provide a risk stratification instrument for future research trials.

A clinician's guide for developing a prediction model

A clinician's guide for developing a prediction model: a case study using real-world data of patients with castration-resistant prostate cancer
J Cancer Res Clin Oncol . 2020 Aug;146(8):2067-2075. doi: 10.1007/s00432-020-03286-8.

• A retrospective registry of 20 Dutch hospitals with data on patients treated for castration-resistant prostate cancer was used to develop a comprehensive guide for clinicians through the steps of developing a risk prediction model. 

Abstract

Social Care Data Collection for Pandemic Planning and Researc

Social Care Data Collection for Pandemic Planning and Research [Press release]
NHS Digital 15 July 2020

• The Social Care Data Collection for Pandemic Planning and Research will collect existing, anonymised data on the COVID-19 status of care givers and care receivers directly from IT systems. The collection will show trends within care settings at a local and national level which will support forecasting of future waves of the pandemic. It will be used to inform research, to plan better services, to deliver better value for money and to improve the quality of individual care. 

Coronavirus (COVID-19) Infection Survey pilot: Englan

Coronavirus (COVID-19) Infection Survey pilot: England
ONS First released 14 May 2020 (weekly update)

• Estimates of the number of coronavirus (COVID-19) infections within the community population; where community refers to private residential households, and it excludes those in hospitals, care homes or other institutional settings.
• This survey is being delivered in partnership with IQVIA, Oxford University and UK Biocentre.

Remote Patient Monitoring Technologies for Predicting COPD Exacerbations: Review and Comparison.

Remote Patient Monitoring Technologies for Predicting Chronic Obstructive Pulmonary Disease Exacerbations: Review and Comparison.
JMIR Mhealth Uhealth 2020;8(5):e16147

• A list of handheld and hands-free commercially available remote patient monitoring (RPM) tools which focused on predicting COPD exacerbations were identified. They were assessed based on forecasting ability, cost, ease of use, and appearance. Devices that ranked higher on all criteria tended to have a high or unlisted price. Commonly mass-marketed devices like the pulse oximeter and spirometer fulfilled the least criteria.

Abstract

Predicting Total Knee Replacement from Symptomology and Radiographic Structural Change Using Artificial Neural Networks-Data from the Osteoarthritis Initiative (OAI)

Predicting Total Knee Replacement from Symptomology and Radiographic Structural Change Using Artificial Neural Networks-Data from the Osteoarthritis Initiative (OAI)
J Clin Med. 2020;9(5):1298. Published 2020 May 1. doi:10.3390/jcm9051298

• A prediction model based on easily available patient data has been developed which give the opportunity to assess a patient’s need for total knee replacement surgery two years in advance and which could be used in a primary care setting.

Abstract

Development and validation of the Cambridge Multimorbidity Score

Development and validation of the Cambridge Multimorbidity Score
CMAJ February 03, 2020 192 (5) E107-E114; DOI: https://doi.org/10.1503/cmaj.190757

• Researchers at the NIHR School for Primary Care Research have developed a new score for measuring multiple long-term health conditions in patients in primary care.
• The new Cambridge Multimorbidity Score is a transparent, simple measure of multimorbidity that can predict different outcomes in people with multiple conditions. The score is based on  data obtained from the UK Clinical Practice Research Datalink (CPRD)
• See NIHR news

Challenges to the Reproducibility of Machine Learning Models in Health Care

Challenges to the Reproducibility of Machine Learning Models in Health Care
JAMA. 2020;323(4):305-306. doi:10.1001/jama.2019.20866

• High-capacity machine learning models are beginning to demonstrate early successes in clinical applications, and some have received approval from the US FDA. This new class of clinical prediction tools presents unique challenges and obstacles to reproducibility, which must be carefully considered to ensure that these techniques are valid and deployed safely and effectively.

Abstract

Trust and telecoms giant may use patient phone data to predict crises

Trust and telecoms giant may use patient phone data to predict crises
HSJ 22 November 2019 [Subscription required]

• Telefonica, which owns mobile firm O2, is working with Birmingham and Solihull Mental Health Trust to see whether using call records, location information and other forms of “mobile network data” can be used to flag people about to go into crisis.
• In a pilot project the project team used anonymised historical health records to build an algorithm that aims to flag patients at risk of crisis. This “phase one” algorithm did not use mobile network data.

Is it right to use AI to identify children at risk of harm? (Lynn Eaton)

Is it right to use AI to identify children at risk of harm? (Lynn Eaton)
Guardian 18 November 2019

• A discussion of the predictive analytics technologies in use to help social workers assess the probability of a child coming on to the at-risk register. Including Xantura (Thurrock and Dagenham councils), Essex Centre for Data Analytics, Centre for Artificial Intelligence and Public Policy (New Zealand), Office of Data Analytics (NESTA), Project ExplAIn (Alan Turing Institute and ICO).

Health Secretary announces £250 million investment in artificial intelligence

Health Secretary announces £250 million investment in artificial intelligence
DHSC News 8 August 2019

• A new National Artificial Intelligence Lab will use the power of artificial intelligence (AI) to improve the health and lives of patients.
• The AI Lab will bring together the industry’s best academics, specialists and technology companies to work on some of the biggest challenges in health and care, including earlier cancer detection, new dementia treatments and more personalised care.
• The lab will sit within NHSX, the new organisation that will oversee the digitisation of the health and care system, in partnership with the Accelerated Access Collaborative.
• Among other things the AI Lab’s work could:
• use predictive models to better estimate future needs of beds, drugs, devices or surgeries
• identify which patients could be more easily treated in the community, reducing the pressure on the NHS and helping patients receive treatment closer to home
• identify patients most at risk of diseases such as heart disease or dementia, allowing for earlier diagnosis and cheaper, more focused, personalised prevention
• build systems to detect people at risk of post-operative complications, infections or requiring follow-up from clinicians, improving patient safety and reducing readmission rates

Predictive Modelling for Unplanned Care in the North East and North Cumbria

Predictive Modelling for Unplanned Care in the North East and North Cumbria (research project)
Connected Health Cities

• A collaboration to produce statistical models that can be routinely used by appropriate health/local authority/other analytics teams to produce daily forecasts up to six months in advance with the pertinent associated uncertainties and variations in Urgent and Emergency Care.
• One of the Connected Health Cities research projects using local health data and advanced technology to improve health services for patients across the North of England.

A lifecourse map of 308 physical and mental health conditions

A chronological map of 308 physical and mental health conditions from 4 million individuals in the English National Health Service
Lancet Digital Health 2019; 1: e63–77, 1 June 2019

• This research aims to understand which sections of the population are susceptible to which health conditions and at which ages. It presents a lifecourse map of human health, charting the 50 most common conditions in each decade of life, and the median age at diagnosis for 308 conditions by sex and ethnicity. The results illustrate the varying dominance of different conditions through the passage of life.
•  Supplementary files include data tables of prevalence by age band, sex, and ethnicity across primary-care and secondary-care records in England.