Health Outcomes from Home Hospitalization: Multisource Predictive Modeling

The value of admission avoidance: cost-consequence analysis of one-year activity in a consolidated service

BACKGROUND

Many advantages of hospital at home (HaH), as a modality of acute care, have been highlighted, but controversies exist regarding the cost-benefit trade-offs. The objective is to assess health outcomes and analytical costs of hospital avoidance (HaH-HA) in a consolidated service with over ten years of delivery of HaH in Barcelona (Spain).

METHODS

A retrospective cost-consequence analysis of all first episodes of HaH-HA, directly admitted from the emergency room (ER) in 2017-2018, was carried out with a health system perspective. HaH-HA was compared with a propensity-score-matched group of contemporary patients admitted to conventional hospitalization (Controls). Mortality, re-admissions, ER visits, and direct healthcare costs were evaluated.

RESULTS

HaH-HA and Controls (n = 441 each) were comparable in terms of age (73 [SD16] vs. 74 [SD16]), gender (male, 57% vs. 59%), multimorbidity, healthcare expenditure during the previous year, case mix index of the acute episode, and main diagnosis at discharge. HaH-HA presented lower mortality during the episode (0 vs. 19 (4.3%); p < 0.001). At 30 days post-discharge, HaH-HA and Controls showed similar re-admission rates; however, ER visits were lower in HaH-HA than in Controls (28 (6.3%) vs. 34 (8.1%); p = 0.044). Average costs per patient during the episode were lower in the HaH-HA group (€ 1,078) than in Controls (€ 2,171). Likewise, healthcare costs within the 30 days post-discharge were also lower in HaH-Ha than in Controls (p < 0.001).

CONCLUSIONS

The study showed higher performance and cost reductions of HaH-HA in a real-world setting. The identification of sources of savings facilitates scaling of hospital avoidance.

REGISTRATION

ClinicalTrials.gov (26/04/2017; NCT03130283).

Natural Language Processing Applied to Clinical Documentation in Post-acute Care Settings: A Scoping Review

OBJECTIVES

To determine the scope of the application of natural language processing to free-text clinical notes in post-acute care and provide a foundation for future natural language processing-based research in these settings.

DESIGN

Scoping review; reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines.

SETTING AND PARTICIPANTS

Post-acute care (ie, home health care, long-term care, skilled nursing facilities, and inpatient rehabilitation facilities).

METHODS

PubMed, Cumulative Index of Nursing and Allied Health Literature, and Embase were searched in February 2023. Eligible studies had quantitative designs that used natural language processing applied to clinical documentation in post-acute care settings. The quality of each study was appraised.

RESULTS

Twenty-one studies were included. Almost all studies were conducted in home health care settings. Most studies extracted data from electronic health records to examine the risk for negative outcomes, including acute care utilization, medication errors, and suicide mortality. About half of the studies did not report age, sex, race, or ethnicity data or use standardized terminologies. Only 8 studies included variables from socio-behavioral domains. Most studies fulfilled all quality appraisal indicators.

CONCLUSIONS AND IMPLICATIONS

The application of natural language processing is nascent in post-acute care settings. Future research should apply natural language processing using standardized terminologies to leverage free-text clinical notes in post-acute care to promote timely, comprehensive, and equitable care. Natural language processing could be integrated with predictive models to help identify patients who are at risk of negative outcomes. Future research should incorporate socio-behavioral determinants and diverse samples to improve health equity in informatics tools.

Prevention of Unplanned Hospital Admissions in Multimorbid Patients Using Computational Modeling: Observational Retrospective Cohort Study

BACKGROUND

Enhanced management of multimorbidity constitutes a major clinical challenge. Multimorbidity shows well-established causal relationships with the high use of health care resources and, specifically, with unplanned hospital admissions. Enhanced patient stratification is vital for achieving effectiveness through personalized postdischarge service selection.

OBJECTIVE

The study has a 2-fold aim: (1) generation and assessment of predictive models of mortality and readmission at 90 days after discharge; and (2) characterization of patients' profiles for personalized service selection purposes.

METHODS

Gradient boosting techniques were used to generate predictive models based on multisource data (registries, clinical/functional and social support) from 761 nonsurgical patients admitted in a tertiary hospital over 12 months (October 2017 to November 2018). K-means clustering was used to characterize patient profiles.

RESULTS

Performance (area under the receiver operating characteristic curve, sensitivity, and specificity) of the predictive models was 0.82, 0.78, and 0.70 and 0.72, 0.70, and 0.63 for mortality and readmissions, respectively. A total of 4 patients' profiles were identified. In brief, the reference patients (cluster 1; 281/761, 36.9%), 53.7% (151/281) men and mean age of 71 (SD 16) years, showed 3.6% (10/281) mortality and 15.7% (44/281) readmissions at 90 days following discharge. The unhealthy lifestyle habit profile (cluster 2; 179/761, 23.5%) predominantly comprised males (137/179, 76.5%) with similar age, mean 70 (SD 13) years, but showed slightly higher mortality (10/179, 5.6%) and markedly higher readmission rate (49/179, 27.4%). Patients in the frailty profile (cluster 3; 152/761, 19.9%) were older (mean 81 years, SD 13 years) and predominantly female (63/152, 41.4%, males). They showed medical complexity with a high level of social vulnerability and the highest mortality rate (23/152, 15.1%), but with a similar hospitalization rate (39/152, 25.7%) compared with cluster 2. Finally, the medical complexity profile (cluster 4; 149/761, 19.6%), mean age 83 (SD 9) years, 55.7% (83/149) males, showed the highest clinical complexity resulting in 12.8% (19/149) mortality and the highest readmission rate (56/149, 37.6%).

CONCLUSIONS

The results indicated the potential to predict mortality and morbidity-related adverse events leading to unplanned hospital readmissions. The resulting patient profiles fostered recommendations for personalized service selection with the capacity for value generation.

Machine learning applied to electronic health record data in home healthcare: A scoping review

OBJECTIVE

Despite recent calls for home healthcare (HHC) to integrate informatics, the application of machine learning in HHC is relatively unknown. Thus, this study aimed to synthesize and appraise the literature describing the application of machine learning to predict adverse outcomes (e.g., hospitalization, mortality) using electronic health record (EHR) data in the HHC setting. Our secondary aim was to evaluate the comprehensiveness of predictors used in the machine learning algorithms guided by the Biopsychosocial Model.

METHODS

During March 2022 we conducted a literature search in four databases: PubMed, Embase, CINAHL, and Scopus. Inclusion criteria were 1) describing services provided in the HHC setting, 2) applying machine learning algorithms to predict adverse outcomes, defined as outcomes related to patient deterioration, 3) using EHR data and 4) focusing on the adult population. Predictors were mapped to the Biopsychosocial Model. A risk of bias analysis was conducted using the Prediction Model Risk Of Bias Assessment Tool.

RESULTS

The final sample included 20 studies. Eighteen studies used predictors from standardized assessments integrated in the EHR. The most common outcome of interest was hospitalization (55%), followed by mortality (25%). Psychological predictors were frequently excluded (35%). Tree based algorithms were most frequently applied (75%). Most studies demonstrated high or unclear risk of bias (75%).

CONCLUSION

Future studies in HHC should consider incorporating machine learning algorithms into clinical decision support systems to identify patients at risk. Based on the Biopsychosocial model, psychological and interpersonal characteristics should be used along with biological characteristics to enhance risk prediction. To facilitate the widespread adoption of machine learning, stakeholders should encourage standardization in the HHC setting.

The other side of the crisis: organizational flexibility in balancing Covid-19 and non-Covid-19 health-care services

Background

Many healthcare systems have been unable to deal with Covid-19 without influencing non-Covid-19 patients with pre-existing conditions, risking a paralysis in the medium term. This study explores the effects of organizational flexibility on hospital efficiency in terms of the capacity to deliver healthcare services for both Covid-19 and non-Covid-19 patients.

Method

Focusing on Italian health system, a two-step strategy is adopted. First, Data Envelope Analysis is used to assess the capacity of hospitals to address the needs of Covid-19 and non-Covid-19 patients relying on internal resource flexibility. Second, two panel regressions are performed to assess external organizational flexibility, with the involvement in demand management of external operators in the health-care service, examining the impact on efficiency in hospital capacity management.

Results

The overall response of the hospitals in the study was not fully effective in balancing the needs of the two categories of patients (the efficiency score is 0.87 and 0.58, respectively, for Covid-19 and non-Covid-19 patients), though responses improved over time. Furthermore, among the measures providing complementary services in the community, home hospitalization and territorial medicine were found to be positively associated with hospital efficiency (0.1290, p < 0.05 and 0.2985, p < 0.01, respectively, for non-Covid-19 and Covid-19 patients; 0.0026, p < 0.05 and 0.0069, p < 0.01, respectively, for non-Covid-19 and Covid-19). In contrast, hospital networks are negatively related to efficiency in Covid-19 patients (-0.1037, p < 0.05), while the relationship is not significant in non-Covid-19 patients.

Conclusions

Managing the needs of Covid-19 patients while also caring for other patients requires a response from the entire healthcare system. Our findings could have two important implications for effectively managing health-care demand during and after the Covid-19 pandemic. First, as a result of a naturally progressive learning process, the resource balance between Covid-19 and non-Covid-19 patients improves over time. Second, it appears that demand management to control the flow of patients necessitates targeted interventions that combine agile structures with decentralization. Finally, untested integration models risk slowing down the response, giving rise to significant costs without producing effective results.