Working…
ClinicalTrials.gov
ClinicalTrials.gov Menu

Use of Predictive Modeling to Improve Operating Room Scheduling Efficiency

The safety and scientific validity of this study is the responsibility of the study sponsor and investigators. Listing a study does not mean it has been evaluated by the U.S. Federal Government. Read our disclaimer for details.
 
ClinicalTrials.gov Identifier: NCT01892865
Recruitment Status : Completed
First Posted : July 8, 2013
Results First Posted : January 2, 2018
Last Update Posted : January 2, 2018
Sponsor:
Information provided by (Responsible Party):
VA Office of Research and Development

Study Type Interventional
Study Design Allocation: Randomized;   Intervention Model: Parallel Assignment;   Masking: Quadruple (Participant, Care Provider, Investigator, Outcomes Assessor);   Primary Purpose: Health Services Research
Condition Operating Room Scheduling
Interventions Other: Scheduling using historical means
Other: Scheduling using regression modeling system
Enrollment 735
Recruitment Details Calendar days during which vascular surgery operations were performed were randomly scheduled using either the Historical Means or the Predictive Modeling System methodologies. Please, note that unit of randomization was operative days, not patients
Pre-assignment Details Operative days that were on holidays, or when staff surgeons were out of town were excluded. Similarly, did not schedule any cases during the re-calibration of the predictive models
Arm/Group Title Historical Means Method Predictive Modeling System (PMS)
Hide Arm/Group Description Scheduling using historical means: Scheduling will be performed taking into account historical means only for anesthetic, operative, and turn around time Scheduling using regression modeling system: A regression model that uses predictor of operative length will be used to predict operative, anesthetic, and turn around time length
Period Title: Overall Study
Number of participants Number of units (Operative Days) Number of participants Number of units (Operative Days)
Started 356 [1] 107 379 [2] 100
Completed 356 107 379 100
Not Completed 0 0 0 0
[1]
107 operative days scheduled with the Historical Means method were included
[2]
100 operative days scheduled with the Predictive Modeling System method were included
Arm/Group Title Historical Means Method Predictive Modeling System (PMS) Total
Hide Arm/Group Description Scheduling using historical means: Scheduling will be performed taking into account historical means only for anesthetic, operative, and turn around time Scheduling using regression modeling system: A regression model that uses predictor of operative length will be used to predict operative, anesthetic, and turn around time length Total of all reporting groups
Overall Number of Baseline Participants 356 379 735
Overall Number of Units Analyzed
Type of Units Analyzed: Operative Days
107 100 207
Hide Baseline Analysis Population Description
[Not Specified]
Age, Customized  
Mean (Standard Deviation)
Unit of measure:  Years
Number Analyzed 356 participants 379 participants 735 participants
62  (8) 62.5  (10) 62.5  (10)
Sex/Gender, Customized  
Measure Type: Count of Participants
Unit of measure:  Participants
Sex Number Analyzed 356 participants 379 participants 735 participants
Males
351
  98.6%
373
  98.4%
724
  98.5%
Females
5
   1.4%
6
   1.6%
11
   1.5%
Available operative days for randomization  
Count of Units
Unit of measure:  Operative Days
Number Analyzed 107 Operative Days 100 Operative Days 207 Operative Days
107
 100.0%
100
 100.0%
207
 100.0%
1.Primary Outcome
Title Difference Between the Actual and Predicted Length of Operative Day (in Minutes)
Hide Description The scheduling imprecision between the two scheduling approaches will be compared. Scheduling imprecision is defined as the difference between the actual and predicted length of operative day.
Time Frame Three years
Hide Outcome Measure Data
Hide Analysis Population Description
We analyzed data from 107 operative days in the HM arm, and 100 days in the PMS arm
Arm/Group Title Historical Means Method Predictive Modeling System (PMS)
Hide Arm/Group Description:
Scheduling using historical means: Scheduling will be performed taking into account historical means only for anesthetic, operative, and turn around time
Scheduling using regression modeling system: A regression model that uses predictor of operative length will be used to predict operative, anesthetic, and turn around time length
Overall Number of Participants Analyzed 107 100
Mean (Standard Deviation)
Unit of Measure: Minutes
30.8  (99) 7.2  (67)
Show Statistical Analysis 1 Hide Statistical Analysis 1
Statistical Analysis Overview Comparison Group Selection Historical Means Method, Predictive Modeling System (PMS)
Comments The null hypothesis was that there would be no difference in predictive imprecision for the end of the operative day between the two arms
Type of Statistical Test Superiority or Other
Comments [Not Specified]
Statistical Test of Hypothesis P-Value 0.024
Comments [Not Specified]
Method t-test, 2 sided
Comments [Not Specified]
Method of Estimation Estimation Parameter Mean Difference (Net)
Estimated Value 23.6
Confidence Interval (2-Sided) 95%
3.15 to 44
Estimation Comments [Not Specified]
Show Statistical Analysis 2 Hide Statistical Analysis 2
Statistical Analysis Overview Comparison Group Selection Historical Means Method, Predictive Modeling System (PMS)
Comments Null hypothesis: There would be no difference in throughput between the two arms
Type of Statistical Test Superiority or Other
Comments [Not Specified]
Statistical Test of Hypothesis P-Value 0.04
Comments [Not Specified]
Method Poisson regression
Comments [Not Specified]
Method of Estimation Estimation Parameter Risk Ratio (RR)
Estimated Value 1.16
Confidence Interval (2-Sided) 95%
1.01 to 1.34
Estimation Comments [Not Specified]
2.Secondary Outcome
Title Difference in Throughput
Hide Description Difference in total number of cases scheduled per unit of time analyzed between the two study arms
Time Frame Three years
Hide Outcome Measure Data
Hide Analysis Population Description
Operative days
Arm/Group Title Historical Means Method Predictive Modeling System (PMS)
Hide Arm/Group Description:
Scheduling using historical means: Scheduling will be performed taking into account historical means only for anesthetic, operative, and turn around time
Scheduling using regression modeling system: A regression model that uses predictor of operative length will be used to predict operative, anesthetic, and turn around time length
Overall Number of Participants Analyzed 356 379
Overall Number of Units Analyzed
Type of Units Analyzed: Operative Days
107 100
Measure Type: Number
Unit of Measure: Operations/day analyzed
3.33 3.79
Show Statistical Analysis 1 Hide Statistical Analysis 1
Statistical Analysis Overview Comparison Group Selection Historical Means Method, Predictive Modeling System (PMS)
Comments Null hypothesis was that there was no difference in throughput between the two scheduling methods
Type of Statistical Test Superiority or Other
Comments [Not Specified]
Statistical Test of Hypothesis P-Value 0.04
Comments [Not Specified]
Method Poisson Regression
Comments [Not Specified]
Method of Estimation Estimation Parameter Risk Ratio (RR)
Estimated Value 1.16
Confidence Interval (2-Sided) 95%
1.01 to 1.34
Estimation Comments [Not Specified]
3.Secondary Outcome
Title Operative Suite Personnel Job Satisfaction
Hide Description Comparison of job satisfaction between study arms using three domains of the Maslach Burnout Inventory: Depersonalization (range 0-17, score of 17 indicates worse depersonalization). Emotional Exhaustion (range: 0-36, score of 36 is the worse). Personal accomplishment (range 1-60, score of 60 is best).
Time Frame Three years
Hide Outcome Measure Data
Hide Analysis Population Description
Health care providers
Arm/Group Title Historical Means Method Predictive Modeling System (PMS)
Hide Arm/Group Description:
Scheduling using historical means: Scheduling will be performed taking into account historical means only for anesthetic, operative, and turn around time
Scheduling using regression modeling system: A regression model that uses predictor of operative length will be used to predict operative, anesthetic, and turn around time length
Overall Number of Participants Analyzed 21 21
Overall Number of Units Analyzed
Type of Units Analyzed: Responses
61 53
Mean (Full Range)
Unit of Measure: units on a scale
Depersonalization
3.23
(0 to 17)
2.04
(0 to 10)
Emotional Exhaustion
11.82
(0 to 29)
10.03
(0 to 36)
Personal Accomplishment
37.51
(15 to 48)
40.47
(21 to 48)
Show Statistical Analysis 1 Hide Statistical Analysis 1
Statistical Analysis Overview Comparison Group Selection Historical Means Method, Predictive Modeling System (PMS)
Comments Null hypothesis: There would be no difference in personnel satisfaction between the groups
Type of Statistical Test Superiority or Other
Comments [Not Specified]
Statistical Test of Hypothesis P-Value 0.04
Comments [Not Specified]
Method t-test, 2 sided
Comments [Not Specified]
4.Secondary Outcome
Title Complications: A Composite Endpoint of Death, Myocardial Infarction, Bleeding, Amputation
Hide Description Comparison of the perioperative (30-day postoperative) composite endpoint of death, myocardial infarction, bleeding, amputation between the two study groups
Time Frame Three years
Hide Outcome Measure Data
Hide Analysis Population Description
Patients
Arm/Group Title Historical Means Method Predictive Modeling System (PMS)
Hide Arm/Group Description:
Scheduling using historical means: Scheduling will be performed taking into account historical means only for anesthetic, operative, and turn around time
Scheduling using regression modeling system: A regression model that uses predictor of operative length will be used to predict operative, anesthetic, and turn around time length
Overall Number of Participants Analyzed 356 379
Measure Type: Count of Participants
Unit of Measure: Participants
8 12
Show Statistical Analysis 1 Hide Statistical Analysis 1
Statistical Analysis Overview Comparison Group Selection Historical Means Method, Predictive Modeling System (PMS)
Comments Null hypothesis was that there would be no difference in the adverse event rates between the two scheduling methodologies
Type of Statistical Test Superiority or Other
Comments [Not Specified]
Statistical Test of Hypothesis P-Value 0.44
Comments [Not Specified]
Method Chi-squared
Comments [Not Specified]
Time Frame Randomization unit in this study was operative days; thus, no adverse events were expected.
Adverse Event Reporting Description All-Cause Mortality, Serious, and Other [Not Including Serious] Adverse Events were not monitored/assessed.
 
Arm/Group Title Historical Means Method Predictive Modeling System (PMS)
Hide Arm/Group Description Scheduling using historical means: Scheduling will be performed taking into account historical means only for anesthetic, operative, and turn around time Scheduling using regression modeling system: A regression model that uses predictor of operative length will be used to predict operative, anesthetic, and turn around time length
All-Cause Mortality
Historical Means Method Predictive Modeling System (PMS)
Affected / at Risk (%) Affected / at Risk (%)
Total   --/--   --/-- 
Show Serious Adverse Events Hide Serious Adverse Events
Historical Means Method Predictive Modeling System (PMS)
Affected / at Risk (%) Affected / at Risk (%)
Total   0/0   0/0 
Show Other (Not Including Serious) Adverse Events Hide Other (Not Including Serious) Adverse Events
Frequency Threshold for Reporting Other Adverse Events 0%
Historical Means Method Predictive Modeling System (PMS)
Affected / at Risk (%) Affected / at Risk (%)
Total   0/0   0/0 
Certain Agreements
All Principal Investigators ARE employed by the organization sponsoring the study.
Results Point of Contact
Layout table for Results Point of Contact information
Name/Title: Panos Kougias MD MSc
Organization: Michael E. DeBakey VAMC
Phone: 713.794.7700
EMail: panagiotis.kougias@va.gov
Layout table for additonal information
Responsible Party: VA Office of Research and Development
ClinicalTrials.gov Identifier: NCT01892865     History of Changes
Other Study ID Numbers: IIR 12-113
First Submitted: July 1, 2013
First Posted: July 8, 2013
Results First Submitted: January 30, 2017
Results First Posted: January 2, 2018
Last Update Posted: January 2, 2018