School of Medicine and Health Sciences Poster Presentations

Title

Validation of a Risk Score Calculator for Patient Discharge

Document Type

Poster

Keywords

Anesthesiology; Chronic Pain; Opioid Abuse

Publication Date

Spring 2017

Abstract

Despite public awareness of the increased number of opioid related deaths, opioid use has remained constant over the last decade. Established demographics of populations predisposed to abuse are clear3, however, presently there are no well-defined screening methods. A model predicting risk of discharge from the pain clinic had previously been calculated using data from the GW pain center. Variables examined included smoking, back pain, employment status, type of insurance, drug abuse and referral source. The purpose of our study was to see if this equation did indeed predict risk of discharge in a separate group of patients. With IRB approval, 26 subjects dismissed from clinic over a 13-month period were identified and 25 patients seen during the same time period were randomly selected as control. Subject risk factors were identified based on chart review of risk model characteristics. The risk of involuntary discharge was calculated using the previously developed risk stratification model. Data were divided into ordinal risk quartiles and analyzed using chi-square analysis. Among the 51 patients in the validation sample who had non-missing data for all components of the risk model, 26 (51%) were discharged involuntarily. The rate of involuntary discharge by quartiles of risk score were 8%, 54%, 56%, and 76%, for quartiles 1 to 4, respectively (Figure 1). The association between risk score quartile and being involuntarily discharged was strong (phi=.51) and significant (p=.004). The odds of being involuntarily discharged for those in the highest versus lowest risk quartile were 35.75 (95% confidence interval 3.47-368.83). Risk scores may be helpful in identifying patients at risk for opioid abuse and discharge from a pain clinic. In a previous study based on our patient population, we calculated a risk score based on patient characteristics significantly associated with discharge. These were illicit drug use, unemployment, disability, smoking and low back pain. In this study, we found a significant association between the risk score calculated in subsequent patients discharged and therefore validated the risk score calculation for our patient population.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Open Access

1

Comments

Poster to be presented at GW Annual Research Days 2017.

This document is currently not available here.

Share

COinS
 

Validation of a Risk Score Calculator for Patient Discharge

Despite public awareness of the increased number of opioid related deaths, opioid use has remained constant over the last decade. Established demographics of populations predisposed to abuse are clear3, however, presently there are no well-defined screening methods. A model predicting risk of discharge from the pain clinic had previously been calculated using data from the GW pain center. Variables examined included smoking, back pain, employment status, type of insurance, drug abuse and referral source. The purpose of our study was to see if this equation did indeed predict risk of discharge in a separate group of patients. With IRB approval, 26 subjects dismissed from clinic over a 13-month period were identified and 25 patients seen during the same time period were randomly selected as control. Subject risk factors were identified based on chart review of risk model characteristics. The risk of involuntary discharge was calculated using the previously developed risk stratification model. Data were divided into ordinal risk quartiles and analyzed using chi-square analysis. Among the 51 patients in the validation sample who had non-missing data for all components of the risk model, 26 (51%) were discharged involuntarily. The rate of involuntary discharge by quartiles of risk score were 8%, 54%, 56%, and 76%, for quartiles 1 to 4, respectively (Figure 1). The association between risk score quartile and being involuntarily discharged was strong (phi=.51) and significant (p=.004). The odds of being involuntarily discharged for those in the highest versus lowest risk quartile were 35.75 (95% confidence interval 3.47-368.83). Risk scores may be helpful in identifying patients at risk for opioid abuse and discharge from a pain clinic. In a previous study based on our patient population, we calculated a risk score based on patient characteristics significantly associated with discharge. These were illicit drug use, unemployment, disability, smoking and low back pain. In this study, we found a significant association between the risk score calculated in subsequent patients discharged and therefore validated the risk score calculation for our patient population.