Dataset Information

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

ABSTRACT:

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.

SUBMITTER: Bishop J

PROVIDER: S-EPMC8610279 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundThere is a growing impetus to reorganize the hospital discharge process to reduce avoidable readmissions and costs. The aim of this study was to provide insight into hospital discharge problems and underlying causes, and to give an overview of solutions that guide providers and policy-makers in improving hospital discharge.MethodsThe Intervention Mapping framework was used. First, a problem analysis studying the scale, causes, and consequences of ineffective hospital discharge was carried out. The analysis was based on primary data from 26 focus group interviews and 321 individual interviews with patients and relatives, and involved hospital and community care providers. Second, improvements in terms of intervention outcomes, performance objectives and change objectives were specified. Third, 220 experts were consulted and a systematic review of effective discharge interventions was carried out to select theory-based methods and practical strategies required to achieve change and better performance.ResultsIneffective discharge is related to factors at the level of the individual care provider, the patient, the relationship between providers, and the organisational and technical support for care providers. Providers can reduce hospital readmission rates and adverse events by focusing on high-quality discharge information, well-coordinated care, and direct and timely communication with their counterpart colleagues. Patients, or their carers, should participate in the discharge process and be well aware of their health status and treatment. Assessment by hospital care providers whether discharge information is accurate and understood by patients and their community counterparts, are important examples of overcoming identified barriers to effective discharge. Discharge templates, medication reconciliation, a liaison nurse or pharmacist, regular site visits and teach-back are identified as effective and promising strategies to achieve the desired behavioural and environmental change.ConclusionsThis study provides a comprehensive guiding framework for providers and policy-makers to improve patient handover from hospital to primary care.

Project description:IntroductionIntrahospital transitions (IHTs) represent movements of patients during hospitalisation. While transitions are often clinically necessary, such as a transfer from the emergency department to an intensive care unit, transitions may disrupt care coordination, such as discharge planning. Family carers often serve as liaisons between the patient and healthcare professionals. However, carers frequently experience exclusion from care planning during IHTs, potentially decreasing their awareness of patients' clinical status, postdischarge needs and carer preparation. The purpose of this study was to explore family carers' perceptions about IHTs, patient and carer ratings of patient discharge readiness and carer self-perception of preparation to engage in at home care.MethodsSequential, explanatory mixed-methods study involving retrospective analysis of hospital inpatients from a parent study (1R01HS026248; PI Wallace) for whom patient and family carer Readiness for Hospital Discharge Scale (RHDS) score frequency of IHTs and patient and caregiver characteristics were available. Maximum variation sampling was used to recruit a subsample of carers with diverse backgrounds and experiences for the participation in semistructured interviews to understand their views of how IHTs influenced preparation for discharge.ResultsOf discharged patients from July 2020 to April 2021, a total of 268 had completed the RHDS and 23 completed the semistructured interviews. Most patients experienced 0-2 IHTs and reported high levels of discharge readiness. During quantitative analysis, no association was found between IHTs and patients' RHDS scores. However, carers' perceptions of patient discharge readiness were negatively associated with increased IHTs. Moreover, non-spouse carers reported lower RHDS scores than spousal carers. During interviews, carers shared barriers experienced during IHTs and discussed the importance of inclusion during discharge care planning.ConclusionsIHTs often represent disruptive events that may influence carers' understanding of patient readiness for discharge to home and, thus, their own preparation for discharge. Further consideration is needed regarding how to support carers during IHT to facilitate high-quality discharge planning.

Project description:Patient-reported outcomes (PRO) have been increasingly emphasized, however, determining clinically valuable PRO has been problematic and investigation limited. This study examines the association of readiness for discharge, which has been described previously, with patient satisfaction and readmission.Data from adult patients admitted to our institution from 2009 to 2012 who completed both the Hospital Consumer Assessment of Healthcare Providers and Systems and the Press Ganey surveys post discharge were extracted from an existing database of patients (composed of 220 patients admitted for small bowel obstruction and 98 patients with hospital stays ? 21 days). Using the survey question, "Did you feel ready for discharge?" (RFD), 2 groups were constructed, those RFD and those with lesser degrees of readiness (ie, less ready for discharge [LRFD]) using topbox methodology. Outcomes, readmission rates, and satisfaction were compared between RFD and LRFD groups.Three hundred and eighteen patients met the inclusion criteria; 45% were female and 94% were Caucasian. Median age was 62.3 years (interquartile range 52.5 to 70.8 year). Median length of stay was 10 days (interquartile range 6.0 to 24.0 days) and 69.2% were admitted with small bowel obstruction. The 30-day readmission rate was 14.3% and 55% indicated they were RFD. Those RFD and LRFD had similar demographics, comorbidity scores, and rates of surgery. Those RFD had higher overall hospital satisfaction (87.3% RFD vs 62.4% LRFD; p < 0.001), higher physician communication scores (median 3.0 RFD vs 2.0 LRFD; p < 0.001), and higher nursing communication scores (median 3.0 RFD vs 2.0 LRFD, p < 0.001). Readmission rates were similar between the groups (11.4% RFD vs 18.2% LRFD; p = 0.09).Readiness for discharge appears to be a clinically useful patient-reported metric, as those RFD have higher satisfaction with the hospital and physicians. Prospective investigation into variables affecting patient satisfaction in those LRFD is needed.

Project description:The discharge summary (DS) is a document that contains the diagnosis, comorbidities, procedures, complications, and future treatment plan for a particular patient after an inpatient hospital stay. The DS is completed by junior medical staff and is delivered to the general practitioner (GP). DS completion is time consuming and tedious, and DSs are usually not completed within the recommended time frame after a patient is discharged. Time spent completing DSs correlate to junior doctor overtime, which costs the hospital money in overtime pay. Information that is required in the DS is generally already entered into numerous electronic information systems in the hospital, including the "electronic patient journey board" which lists all the patients in a given ward with their clinical information. This information is constantly updated by all staff in the hospital. A program was developed that transferred this information directly into the patient DS. Ten junior doctors in two departments kept daily records for one week of the time spent compiling DSs, the time at work and the actual overtime claimed, before and after the introduction of the intervention. The mean (± SD) time for DS compilation per week reduced by 2.8 (± 2.4) hours from 10.0 (±3.5) hours (p<0.01) and the mean overtime worked per week reduced by 2.8 (± 3.1) hours from 8.5 (± 4.4) hours (p<0.05). The mean overtime claimed reduced by 1.8 (± 2.8) hours from 5.3 (± 5.4) hours per week (p<0.05), resulting in reduction in mean overtime payment of $114.95 from $290.57 per doctor, per week. Extrapolating to the 60 ward based junior doctors, the potential annual savings for the hospital budget are over $350,000. Additionally, the number of DSs completed within 48 hours increased from 45% to 58%. In summary, the transfer of electronic data from the electronic patient journey board to the discharge summary program has yielded improvements in DS completion rates and overtime worked by medical staff, resulting in significant reduction in overtime costs.