Improving patient flow during infectious disease outbreaks using machine learning for real-time prediction of patient readiness for discharge.

BACKGROUND: Delays in patient flow and a shortage of hospital beds are commonplace in hospitals during periods of increased infection incidence, such as seasonal influenza and the COVID-19 pandemic. The objective of this study was to develop and evaluate the efficacy of machine learning methods at identifying and ranking the real-time readiness of individual patients for discharge, with the goal of improving patient flow within hospitals during periods of crisis. METHODS AND PERFORMANCE: Electronic Health Record data from Oxford University Hospitals was used to train independent models to classify and rank patients' real-time readiness for discharge within 24 hours, for patient subsets according to the nature of their admission (planned or emergency) and the number of days elapsed since their admission. A strategy for the use of the models' inference is proposed, by which the model makes predictions for all patients in hospital and ranks them in order of likelihood of discharge within the following 24 hours. The 20% of patients with the highest ranking are considered as candidates for discharge and would therefore expect to have a further screening by a clinician to confirm whether they are ready for discharge or not. Performance was evaluated in terms of positive predictive value (PPV), i.e., the proportion of these patients who would have been correctly deemed as 'ready for discharge' after having the second screening by a clinician. Performance was high for patients on their first day of admission (PPV = 0.96/0.94 for planned/emergency patients respectively) but dropped for patients further into a longer admission (PPV = 0.66/0.71 for planned/emergency patients still in hospital after 7 days). CONCLUSION: We demonstrate the efficacy of machine learning methods at making operationally focused, next-day discharge readiness predictions for all individual patients in hospital at any given moment and propose a strategy for their use within a decision-support tool during crisis periods.

Sharing a household with children and risk of COVID-19: a study of over 300 000 adults living in healthcare worker households in Scotland.

OBJECTIVE: Children are relatively protected from COVID-19, due to a range of potential mechanisms. We investigated if contact with children also affords adults a degree of protection from COVID-19. DESIGN: Cohort study based on linked administrative data. SETTING: Scotland. STUDY POPULATION: All National Health Service Scotland healthcare workers and their household contacts as of March 2020. MAIN EXPOSURE: Number of young children (0-11 years) living in the participant's household. MAIN OUTCOMES: COVID-19 requiring hospitalisation, and any COVID-19 (any positive test for SARS-CoV-2) in adults aged ≥18 years between 1 March and 12 October 2020. RESULTS: 241 266, 41 198, 23 783 and 3850 adults shared a household with 0, 1, 2 and 3 or more young children, respectively. Over the study period, the risk of COVID-19 requiring hospitalisation was reduced progressively with increasing numbers of household children-fully adjusted HR (aHR) 0.93 per child (95% CI 0.79 to 1.10). The risk of any COVID-19 was similarly reduced, with the association being statistically significant (aHR per child 0.93; 95% CI 0.88 to 0.98). After schools reopened to all children in August 2020, no association was seen between exposure to young children and risk of any COVID-19 (aHR per child 1.03; 95% CI 0.92 to 1.14). CONCLUSION: Between March and October 2020, living with young children was associated with an attenuated risk of any COVID-19 and COVID-19 requiring hospitalisation among adults living in healthcare worker households. There was no evidence that living with young children increased adults' risk of COVID-19, including during the period after schools reopened.

Risk of hospital admission with coronavirus disease 2019 in healthcare workers and their households: nationwide linkage cohort study.

OBJECTIVE: To assess the risk of hospital admission for coronavirus disease 2019 (covid-19) among patient facing and non-patient facing healthcare workers and their household members. DESIGN: Nationwide linkage cohort study. SETTING: Scotland, UK, 1 March to 6 June 2020. PARTICIPANTS: Healthcare workers aged 18-65 years, their households, and other members of the general population. MAIN OUTCOME MEASURE: Admission to hospital with covid-19. RESULTS: The cohort comprised 158 445 healthcare workers, most of them (90 733; 57.3%) being patient facing, and 229 905 household members. Of all hospital admissions for covid-19 in the working age population (18-65 year olds), 17.2% (360/2097) were in healthcare workers or their households. After adjustment for age, sex, ethnicity, socioeconomic deprivation, and comorbidity, the risk of admission due to covid-19 in non-patient facing healthcare workers and their households was similar to the risk in the general population (hazard ratio 0.81 (95% confidence interval 0.52 to 1.26) and 0.86 (0.49 to 1.51), respectively). In models adjusting for the same covariates, however, patient facing healthcare workers, compared with non-patient facing healthcare workers, were at higher risk (hazard ratio 3.30, 2.13 to 5.13), as were household members of patient facing healthcare workers (1.79, 1.10 to 2.91). After sub-division of patient facing healthcare workers into those who worked in "front door," intensive care, and non-intensive care aerosol generating settings and other, those in front door roles were at higher risk (hazard ratio 2.09, 1.49 to 2.94). For most patient facing healthcare workers and their households, the estimated absolute risk of hospital admission with covid-19 was less than 0.5%, but it was 1% and above in older men with comorbidity. CONCLUSIONS: Healthcare workers and their households contributed a sixth of covid-19 cases admitted to hospital. Although the absolute risk of admission was low overall, patient facing healthcare workers and their household members had threefold and twofold increased risks of admission with covid-19.