2.50
Hdl Handle:
http://hdl.handle.net/10755/157882
Type:
Presentation
Title:
Risk Adjustment in Outcome Research
Abstract:
Risk Adjustment in Outcome Research
Conference Sponsor:Western Institute of Nursing
Conference Year:2006
Author:Tourangeau, Ann, RN, PhD
P.I. Institution Name:University of Toronto
Title:Assistant Professor
Contact Address:Faculty of Nursing, 215-155 College Street, Toronto, ON, M5T-1P8, Canada
Contact Telephone:416-978-6919
Purpose: This paper describes an indirect standardization risk adjustment approach useful for minimizing threats to internal validity that arise from the impact of patients' own characteristics and their associated risks on outcomes under investigation. Risk adjustment procedures for the outcome of mortality in acute medical hospitalized patients are illustrated using this approach. Background: In outcome research, effective risk adjustment is necessary to make meaningful comparisons among patient outcomes. There are at least three major sources of variation in patient outcomes including: individual patient characteristics, quality of patient care (e.g., structures and processes of health care), and random sources. For many outcomes, the largest source of variation are patients' own characteristics such as age and preexisting health conditions. When researchers do not effectively control for patient characteristics, threats to internal validity that arise from patients' characteristics may weaken the credibility of results found. Methods: A two-step approach to outcomes research is suggested. Patient data from large administrative and clinical databases are used to illustrate development of risk-adjusted outcomes. The first step includes risk-adjusting outcomes for patient characteristics by developing standard outcome rates. The suggested general formula for developing a standard rate is the ratio: number of the observed outcomes of interest (e.g., deaths) divided by the expected number of the outcome of interest (e.g., deaths). The numerator for this rate is calculated directly from gathered study data. The denominator is calculated through logistic regression modeling that estimates the expected value or probability of the outcome for each study patient. In the second step, risk-adjusted standard rates can then be used as dependent variables in further outcome analytic models. This paper describes the first step suggested for outcome research; that of risk-adjusting outcomes to account for the impact of patients' own characteristics on the outcome. Results: C-statistics for logistic regression models used to calculate the denominator of the standard outcome rate indicate effective risk-adjusted for the impact of patients' own characteristics on 30-day mortality (ranged from .78 to .82). Risk adjustment results in changes in both absolute values and rank ordering of hospital mortality rates compared to crude rates. Implications: This suggested approach is useful for developing risk-adjusted outcomes. However, risk-adjustment success is dependent on having access to adequate data that indicates patient characteristics hypothesized to affect the outcome of interest. The author gratefully acknowledges grants received from the both the Canadian Institute of Health Research and the Canadian Health Services Research Foundation that funded this research.
Repository Posting Date:
26-Oct-2011
Date of Publication:
17-Oct-2011
Sponsors:
Western Institute of Nursing

Full metadata record

DC FieldValue Language
dc.typePresentationen_GB
dc.titleRisk Adjustment in Outcome Researchen_GB
dc.identifier.urihttp://hdl.handle.net/10755/157882-
dc.description.abstract<table><tr><td colspan="2" class="item-title">Risk Adjustment in Outcome Research</td></tr><tr class="item-sponsor"><td class="label">Conference Sponsor:</td><td class="value">Western Institute of Nursing</td></tr><tr class="item-year"><td class="label">Conference Year:</td><td class="value">2006</td></tr><tr class="item-author"><td class="label">Author:</td><td class="value">Tourangeau, Ann, RN, PhD</td></tr><tr class="item-institute"><td class="label">P.I. Institution Name:</td><td class="value">University of Toronto</td></tr><tr class="item-author-title"><td class="label">Title:</td><td class="value">Assistant Professor</td></tr><tr class="item-address"><td class="label">Contact Address:</td><td class="value">Faculty of Nursing, 215-155 College Street, Toronto, ON, M5T-1P8, Canada</td></tr><tr class="item-phone"><td class="label">Contact Telephone:</td><td class="value">416-978-6919</td></tr><tr class="item-email"><td class="label">Email:</td><td class="value">ann.tourangeau@utoronto.ca</td></tr><tr><td colspan="2" class="item-abstract">Purpose: This paper describes an indirect standardization risk adjustment approach useful for minimizing threats to internal validity that arise from the impact of patients' own characteristics and their associated risks on outcomes under investigation. Risk adjustment procedures for the outcome of mortality in acute medical hospitalized patients are illustrated using this approach. Background: In outcome research, effective risk adjustment is necessary to make meaningful comparisons among patient outcomes. There are at least three major sources of variation in patient outcomes including: individual patient characteristics, quality of patient care (e.g., structures and processes of health care), and random sources. For many outcomes, the largest source of variation are patients' own characteristics such as age and preexisting health conditions. When researchers do not effectively control for patient characteristics, threats to internal validity that arise from patients' characteristics may weaken the credibility of results found. Methods: A two-step approach to outcomes research is suggested. Patient data from large administrative and clinical databases are used to illustrate development of risk-adjusted outcomes. The first step includes risk-adjusting outcomes for patient characteristics by developing standard outcome rates. The suggested general formula for developing a standard rate is the ratio: number of the observed outcomes of interest (e.g., deaths) divided by the expected number of the outcome of interest (e.g., deaths). The numerator for this rate is calculated directly from gathered study data. The denominator is calculated through logistic regression modeling that estimates the expected value or probability of the outcome for each study patient. In the second step, risk-adjusted standard rates can then be used as dependent variables in further outcome analytic models. This paper describes the first step suggested for outcome research; that of risk-adjusting outcomes to account for the impact of patients' own characteristics on the outcome. Results: C-statistics for logistic regression models used to calculate the denominator of the standard outcome rate indicate effective risk-adjusted for the impact of patients' own characteristics on 30-day mortality (ranged from .78 to .82). Risk adjustment results in changes in both absolute values and rank ordering of hospital mortality rates compared to crude rates. Implications: This suggested approach is useful for developing risk-adjusted outcomes. However, risk-adjustment success is dependent on having access to adequate data that indicates patient characteristics hypothesized to affect the outcome of interest. The author gratefully acknowledges grants received from the both the Canadian Institute of Health Research and the Canadian Health Services Research Foundation that funded this research.</td></tr></table>en_GB
dc.date.available2011-10-26T20:17:45Z-
dc.date.issued2011-10-17en_GB
dc.date.accessioned2011-10-26T20:17:45Z-
dc.description.sponsorshipWestern Institute of Nursingen_GB
All Items in this repository are protected by copyright, with all rights reserved, unless otherwise indicated.