Emergencies
Conditions
Keywords
Emergency Department visits,, machine learning, smartphone
Brief summary
Development, validation and impact of an alert management system using social workers' observations and machine learning algorithms to predict 7-to-14-day alerts for the risk of Emergency Department (ED) Visit and unplanned hospitalization. Multi-center trial implementation of electronic Home Care Aides-reported outcomes measure system among patients, frail adults \>= 65 years living at home and receiving assistance from home care aides (HCA).
Detailed description
On a weekly basis, after home visit, HCAs reported on participants' functional status using a smartphone application that recorded 23 functional items about each participant (e.g., ability to stand, move, eat, mood, loneliness). Predictive system using Machine learning techniques (i.e., leveraging random forest predictors) was developed and generated 7 to 14-day predictive alerts for the risk of ED visit to nurses. This questionnaire focused on functional and clinical autonomy (ie, activities of daily life), possible medical symptoms (eg, fatigue, falls, and pain), changes in behavior (eg, recognition and aggressiveness), and communication with the HA or their surroundings. This questionnaire is composed of very simple and easy-to-understand questions, giving a global view of the person's condition. For each of the 23 questions, a yes/no answer was requested. Data recorded by HAs were sent in real time to a secure server to be analyzed by our machine learning algorithm, which predicted the risk level and displayed it on a web-based secure medical device called PRESAGE CARE, which is CE marked. Particularly, when the algorithm predicted a high-risk level, an alert was displayed in the form of a notification on the screen to the coordinating nurse of the health care network center of the district. This risk notification was accompanied by information about recent changes in the patients' functional status, identified from the HAs' records, to assist the coordinating nurse in interacting with family caregiver and other health professionals. In the event of an alert, the coordinating nurse called the family caregiver to inquire about recent changes in the patient's health condition and for doubt removal and could then decide to ask for a health intervention according to a health intervention model developed before the start of the study. In brief, this alert-triggered health intervention (ATHI) consisted of calling the patient's nurse (if the patient had regular home visits of a nurse) or the patient's general practitioner and informing them of a worsening of the patient's functional status and a potential risk of an ED visit or unplanned hospitalization in the next few days according to the eHealth system algorithm. This model of ATHI had been presented and approved by the Agences Régionales de Santé of the regions involved in our study
Interventions
DEVICEPRESAGE CARE
Participants in this arm will be followed by HCA and might benefit from Nurse health interventions
Sponsors
Assistance Publique Hopitaux De Marseille
Assistance Publique - Hôpitaux de Paris
University Hospital, Lille
Presage
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
PREVENTION
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
75 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* age of 75 yo mini * receiving the help of a social worker * patient should give their consent * patient should had seen their primary care professional within the past 12 months
Exclusion criteria
* People with severe dependence (French national instrument, which stratifies dependency level from group iso-resources (GIR) : 1 (very severe dependency) and 2 (severe dependency)
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Unplanned Hospitalization rate | through study completion, an average of 1 year | Comparison between unplanned hospitalization ratio from 2 randomized groups (intervention and control arms). P values \<.05 will be considered statistically significant. |
| Event-free survival (EFS) | through study completion, an average of 1 year | Comparison average Time for first adverse event between intervention and control groups. P values \<.05 will be considered statistically significant. |
| Impact on older adults and relatives' quality of life (European Quality of Life 5 Dimensions and 3 Lines scale) | through study completion, an average of 1 year | Comparison of the average score of EQ5D-3L quality of life scale (European Quality of Life 5 Dimensions and 3 Lines) between intervention and control groups. P values \<.05 will be considered statistically significant. |
| Cost-effectiveness | through study completion, an average of 1 year | Incremental cost-effectiveness ratio (ICER), QALY. Willingness-to-pay thresholds of €30,000 per quality-adjusted life year (QALY) and €90,000 per QALY were used to define a very cost-effective and cost-effective strategy, respectively |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Impact on quality of care | through study completion, an average of 1 year | Positive or very positive impact on quality of care : rate of over 80% is considered acceptable. |
| Impact on users : time needed to complete questionnaire | through study completion, an average of 1 year | Time needed to complete questionnaire (minutes) : a time of less than 2 minutes will be considered acceptable |
| Impact on Professional' Relationship and coordination | through study completion, an average of 1 year | Positive or very positive impact on professionnal relationship and coordination :rate of over 80% is considered acceptable. |
| Intervention rate | through study completion, an average of 1 year | Part of alert which leads to interventions and intervention time (%). Rate of over 70% is considered acceptable. |
| Intervention time | through study completion, an average of 1 year | Mean of the duration between day of alert and day of intervention (in days). A delay of less than 4 days is considered acceptable. |
| Time needed to analysis patient statut | through study completion, an average of 1 year | Time needed to analysis patient statut (hours and minutes) : a time of less than 15 minutes by patient will be considered acceptable |
Countries
France
Outcome results
None listed