Prediction of Safe Discharge From ICU

NCT05459350 | Completed DMC

• 1 other identifier: SAFEDI
• Study Type: observational
• Target: 24,010
• Locations: 1 country
• Sites: 1

Brief Summary

Patients who have an increased need for monitoring or therapy during their stay in hospital are typically admitted to an intensive care unit. This is characterized by a large number of diagnostic and therapeutic options. If this additional effort is no longer necessary, then typically in most hospitals patients are transferred to wards with a lower presence of nurses and physicians and reduced provision of extensive monitoring and therapeutic procedures such as organ replacement procedures. However, deintensification of medical and nursing care requires that previously monitored and partially supported bodily functions are restored to the point where further monitoring is no longer necessary. For this reason, transfer from an intensive care unit to the normal inpatient area is only possible if the patient in question has neither an increased need for monitoring nor an increased need for therapy. If this is not the case, then there is a risk of life-threatening conditions in the normal ward, which can sometimes occur very quickly. However, the need for further monitoring, or for continued intensive medical therapy, cannot be easily assessed. There is no laboratory value or clinical examination method that can be used to estimate beyond doubt whether a patient's condition could worsen if he or she is transferred to the normal ward. For this reason, the decision to transfer is made on the basis of the individual assessment by the attending physician. Although this is based on the synopsis of a wide variety of examinations and laboratory findings, it is therefore subject to large interindividual variations. Thus, the personal experience of the evaluating physician has a considerable influence on the decision for or against a transfer to the normal inpatient area. In this respect, the decision to deintensify therapy, i.e. to transfer patients from intensive care units to the normal care area, is challenging: The assessing physician has to make a prediction from the combination of the available findings under time pressure whether a transfer to the normal inpatient area is possible without endangering the patient. In this situation, it would be desirable to have an automated warning system that could describe the success of the transfer with sufficient accuracy in the presence of specific laboratory constellations. In the best case, such an approach would prevent dangerous transfers, but at the same time reduce unnecessary lengths of stay in the ICU. Machine learning methods seem particularly suited to support such a decision.

Conditions

Risk

Outcome Measures

Primary Outcomes (1)

• AUROC for Classification of Safe Discharge

  AUROC for Classification of Safe Discharge

  2010-01-01 to 2019-10-31

Secondary Outcomes (2)

• Confusion Matrix Value

  2010-01-01 to 2019-10-31

• Descriptive Statistics

  2010-01-01 to 2019-10-31

Study Arms (2)

Safe Discharge Positive

Safe Discharge Positive

Diagnostic Test: Safe Discharge Classification

Safe Discharge Negative

Safe Discharge Negative

Diagnostic Test: Safe Discharge Classification

Interventions

Safe Discharge Classification DIAGNOSTIC_TEST

Safe Discharge Classification

Safe Discharge Negative; Safe Discharge Positive

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

As described in the inclusion criteria.

You may qualify if:

• All adult patients that were treated in intensive care units at the Kepler University Hospital in Linz, Austria in the period 2010-01-01 to 2019-10-31.

You may not qualify if:

• None.

Sponsors & Collaborators

• Kepler University Hospital: lead

Study Sites (1)

Kepler University Hospital

Linz, Upper Austria, 4021, Austria

Location

Study Officials

• Thomas Tschoellitsch, MD

  Kepler University Hospital and Johannes Kepler University, Linz, Austria

  PRINCIPAL INVESTIGATOR

Study Design

• Study Type: observational
• Observational Model: COHORT
• Time Perspective: RETROSPECTIVE
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Study Record Dates

• First Submitted: July 12, 2022
• First Posted: July 15, 2022
• Study Start: June 1, 2022
• Primary Completion: July 31, 2022
• Study Completion: July 31, 2022
• Last Updated: August 17, 2022

Record last verified: 2022-08

Data Sharing

• IPD Sharing: Will not share

Locations

AU (1)