Validation of EPIC's Readmission Risk Model, the LACE+ Index and SQLape as Predictors of Unplanned Hospital Readmissions

NCT04306172 | Completed

• 1 other identifier: LUKS_RRM_2019
• Study Type: observational
• Target: 23,116
• Locations: 1 country
• Sites: 1

Brief Summary

The primary objective of this study is to externally validate the EPIC's Readmission Risk model and to compare it with the LACE+ index and the SQLape Readmission model. As secondary objective, the EPIC's Readmission Risk model will be adjusted based on the validation sample, and finally, it´s performance will be compared with machine learning algorithms.

Conditions

Hospital Readmission

Keywords

Prediction model; hospital readmission; external validation

Outcome Measures

Primary Outcomes (10)

• Discrimination at 18 days

  For discrimination of the scores under investigation, the area under the receiver operating characteristics curves (AUC) will be calculated.

  18 days after index discharge date

• Discrimination at 30 days

  For discrimination of the scores under investigation, the area under the receiver operating characteristics curves (AUC) will be calculated.

  30 days after index discharge date

• Calibration at 18 days

  For calibration, the Hosmer-Lemeshow goodness-of-fit test will be graphically illustrated by plotting the predicted outcomes by decile against the observations.

  18 days after index discharge date

• Calibration at 30 days

  For calibration, the Hosmer-Lemeshow goodness-of-fit test will be graphically illustrated by plotting the predicted outcomes by decile against the observations.

  30 days after index discharge date

• Overall Performance at 18 days

  Brier Score (The Brier score is a quadratic scoring rule, where the squared difference between actual binary outcomes Y and predictions p are calculated. The Brier score can range from 0 for a perfect model to 0.25 for a non-informative model with a 50% incidence of the outcome.)

  18 days after index discharge date

• Overall Performance at 30 days

  Brier Score (The Brier score is a quadratic scoring rule, where the squared difference between actual binary outcomes Y and predictions p are calculated. The Brier score can range from 0 for a perfect model to 0.25 for a non-informative model with a 50% incidence of the outcome.)

  30 days after index discharge date

• Clinical usefulness (NRI) at 18 days

  Net Reclassification Improvement (NRI): In the calculation of the NRI, the improvement in sensitivity and the improvement in specificity are summed. The NRI ranges from 0 for no improvement and 1 for perfect improvement.

  18 days after index discharge date

• Clinical usefulness (NRI) at 30 days

  Net Reclassification Improvement (NRI): In the calculation of the NRI, the improvement in sensitivity and the improvement in specificity are summed. The NRI ranges from 0 for no improvement and 1 for perfect improvement.

  30 days after index discharge date

• Clinical usefulness (NB) at 18 days

  Net Benefit (NB): NB = (TP - w FP) / N, where TP is the number of true positive decisions, FP the number of false positive decisions, N is the total number of patients and w is a weight equal to the odds of the cut-off (pt/(1-pt), or the ratio of harm to benefit

  18 days after index discharge date

• Clinical usefulness (NB) at 30 days

  Net Benefit (NB): NB = (TP - w FP) / N, where TP is the number of true positive decisions, FP the number of false positive decisions, N is the total number of patients and w is a weight equal to the odds of the cut-off (pt/(1-pt), or the ratio of harm to benefit

  30 days after index discharge date

Study Arms (2)

Readmitted inpatients/Cases

Outcome 1: Patients who were readmitted within 18 days of index hospitalization discharge date to the same hospital, with a diagnosis leading to the same Major Diagnostic Group as the index stay (definition according to Swiss Diagnosis Related Groups system, case merger) Outcome 2: Patients with an unplanned readmission within 30 days of index hospitalization discharge date to the same hospital. An unplanned readmission was defined as a readmission through the emergency department.

Other: An US Readmission Risk Prediction Model; Other: LACE+ score; Other: SQLAPE model

Non-Readmitted inpatients/Controls

Outcome 1 \& 2: Patients who were not readmitted within 30 days of index hospitalization discharge date.

Other: An US Readmission Risk Prediction Model; Other: LACE+ score; Other: SQLAPE model

Interventions

An US Readmission Risk Prediction Model OTHER

Logistic regression model that predicts the risk of all-cause unplanned readmissions developed by the privately held healthcare software company EPIC.

Non-Readmitted inpatients/Controls; Readmitted inpatients/Cases

LACE+ score OTHER

The LACE+ score is a point score that can be used to predict the risk of post-discharge death or urgent readmission. It was developed based on administrative data in Ontario, Canada.

Non-Readmitted inpatients/Controls; Readmitted inpatients/Cases

SQLAPE model OTHER

The readmission risk model (Striving for Quality Level and analyzing of patient expenditures), is a computerized validated algorithm and was developed in 2002 to identify potentially avoidable readmissions.

Non-Readmitted inpatients/Controls; Readmitted inpatients/Cases

Eligibility Criteria

• Age: 1 Year - 100 Years
• Sex: all
• Healthy Volunteers: No
• Age Groups: Child (0-17), Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

Inpatients from an acute care hospital in Central-Switzerland

You may qualify if:

• \- All inpatients, aged one year or older (max. 100 years), who were hospitalized either between the 1st of January 2018 and the 31st of December 2018, or between the 23rd of September and the 31st of December 2019 will be included.

You may not qualify if:

• admission/transfer from another psychiatric, rehabilitative or acute care ward from the same institution,
• discharge destination other than the patient's home or
• transfer to another acute care hospital, both being considered as treatment continuation;
• foreign residence,
• deceased before discharge,
• discharged on admission day,
• refusal of general consent, and
• unknown patient residence or discharge destination.

Sponsors & Collaborators

• Luzerner Kantonsspital: lead
• Universität Luzern: collaborator

Study Sites (1)

Cantonal Hospital of Lucerne

Lucerne, Canton Lucerne, 6000, Switzerland

Location

Related Publications (2)

• van Walraven C, Wong J, Forster AJ. LACE+ index: extension of a validated index to predict early death or urgent readmission after hospital discharge using administrative data. Open Med. 2012 Jul 19;6(3):e80-90. Print 2012.

  PMID: 23696773 BACKGROUND

• Halfon P, Eggli Y, Pretre-Rohrbach I, Meylan D, Marazzi A, Burnand B. Validation of the potentially avoidable hospital readmission rate as a routine indicator of the quality of hospital care. Med Care. 2006 Nov;44(11):972-81. doi: 10.1097/01.mlr.0000228002.43688.c2.

  PMID: 17063128 BACKGROUND

Study Officials

• Aljoscha B. Hwang

  University Lucerne (Switzerland)

  PRINCIPAL INVESTIGATOR

• Stefan Boes

  University Lucerne (Switzerland)

  PRINCIPAL INVESTIGATOR

Study Design

• Study Type: observational
• Observational Model: COHORT
• Time Perspective: RETROSPECTIVE
• Sponsor Type: OTHER
• Responsible Party: PRINCIPAL INVESTIGATOR
• PI Title: Research Project Manager & Advanced Analytics Analyst

Study Record Dates

• First Submitted: March 6, 2020
• First Posted: March 12, 2020
• Study Start: March 10, 2020
• Primary Completion: April 10, 2020
• Study Completion: October 1, 2020
• Last Updated: October 20, 2020

Record last verified: 2020-10

Data Sharing

• IPD Sharing: Will not share

Locations

CH (1)