Abstract
Implemented in all fifty states, HIV/AIDS case surveillance
includes required reporting of patient demographics, laboratory
data, treatment and a brief behavioral survey. An algorithm
developed in the 1980's by the Centers for Disease Control and
Prevention (CDC) assigns each case a transmission category based on
this behavioral survey. The three categories which comprise the
vast majority of cases are male sexual contact with another male
(MSM); heterosexual contact with a person known to have HIV
infection or at least with a person at increased risk of HIV
infection; and "injection drug use" (IDU). Unfortunately, many
cases are reported with little or no risk history information and
thus do not meet the criteria for any of the CDC-defined
transmission categories. According to the Georgia Department of
Public Health (DPH), each year a significant percentage of HIV/AIDS
cases diagnosed in Georgia lack complete patient risk history
information. For example, in 2011 70% of HIV cases diagnosed among
males (n=2,002) and 92% of HIV cases diagnosed among females
(n=787) were in the category of "no identified risk" (NIR). This is
a growing problem of public health significance as health
departments depend on the completeness of surveillance data to
monitor changes in HIV/AIDS incidence, track the burden of disease,
plan programs and services, allocate limited resources for HIV
care, and develop strategies for targeting prevention interventions
to populations most at risk of infection. A literature review was
conducted to identify factors that have been shown to be associated
with incomplete reporting of patient information as well as
strategies with demonstrated effectiveness for addressing this
challenge. Also, a logistic regression analysis was conducted of
25,022 adult AIDS cases in the CDC AIDS Public Information Dataset
(APIDS). This analysis identified the following variables as
significantly associated with the binary outcome of having or not
having sufficient risk information for transmission category
classification: patient age at diagnosis, race/ethnicity, residence
in a large metropolitan area, birth in the United States and sex at
birth.
Table of Contents
TABLE OF CONTENTS
Section 1: Background (p. 1)
--1.1: Introduction (p. 1)
--1.2: Problem statement and context (p. 2)
--1.3: Purpose statement (p. 5)
--1.4: Significance statement (p. 6)
--1.5: Definition of terms (p. 6)
Section 2: Review of the literature (p. 12)
Section 3: Key informant interviews (p. 32)
Section 4: Methodology (p. 38)
--4.1: Introduction (p. 38)
--4.2: Population and sample (p. 38)
--4.3: Data analysis procedures (p. 39)
--4.4: Limitations and delimitations (p. 41)
Section 5: Results (p. 42)
--5.1: Findings (p. 42)
--5.2: Discussion (p. 51)
Section 6: Conclusions and recommendations (p. 54)
References (p. 56)
LIST OF FIGURES
Figure 1: Number of AIDS cases diagnosed from 1981 to 2002
among adults and adolescents and percentage classified as "No
Identified Risk" (NIR), adjusted for reporting delays (APIDS
dataset)
Figure 2: Number of AIDS cases diagnosed from 1981 to 2002
among adults and adolescents in the Atlanta MSA and percentage
classified as "No Identified Risk" (NIR), adjusted for reporting
delays (APIDS dataset)
Figure 3: List of recommendations to improve reliability and
validity of HIV-related sexual behaviors reported by patients.
(Weinhardt 1998)
Figure 4: Simplified illustration of CDC algorithm for
determining "transmission category" for adult HIV/AIDS cases
Figure 5: Proposed alternative layout for "Patient Risk
History" section of the Georgia HIV/AIDS case report form.
Figure 6: Among adult and adolescent AIDS cases diagnosed in
2002: Percentage classified as "No Identified Risk" (NIR), by age
group (APIDS dataset)
LIST OF TABLES
Table 1: Number of HIV cases diagnosed and reported in 2010 in
Georgia by exposure category: Comparison of two data presentation
options
Table 2: Number of HIV cases diagnosed and reported in 2011 in
Georgia by exposure category: Comparison of two data presentation
options
Table 3: Number of AIDS cases diagnosed in 2002 among adults
and adolescents and percentage classified as "No Identified Risk"
(NIR), by selected characteristics (APIDS dataset)
Table 4: Logistic regression analysis of the association
between various sociodemographic characteristics and classification
as "No Identified Risk" (NIR): AIDS cases reported in 2002 among
adults and adolescents (APIDS dataset)
Table 5: Evaluation of logistic regression model for the
association between various sociodemographic characteristics and
classification as "No Identified Risk" (NIR): AIDS cases reported
in 2002 among adults and adolescents (APIDS dataset)
Table 6: Observed and predicted frequencies for classification
as "No Identified Risk" (NIR) in a logistic regression model with a
0.50 prediction cuttoff: AIDS cases reported in 2002 among adults
and adolescents (APIDS dataset)
Table 7: Cumulative predicted probability of being classified
as "No Identified Risk" (NIR) by logistic regression for
hypothetical case examples: AIDS cases reported in 2002 among
adults and adolescents (APIDS dataset)
LIST OF ATTACHMENTS
Attachment #1: Key informant interview guide
Attachment #2: Georgia confidential HIV/AIDS case report form
(2 pages)
Attachment #3: CDC 2011 Adult HIV Confidential Case Report
Form (4 pages)
Attachment #4: Codebook for the APIDS public access
database
Attachment #5: Exempt study approval letter (Emory IRB)
Attachment #6: Code for logistic regression model in SAS
9.2
Attachment #7: SAS output
About this Master's Thesis
Rights statement
- Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
| School |
|
|---|
| Department |
|
|---|
| Degree |
|
|---|
| Submission |
|
|---|
| Language |
|
|---|
| Research Field |
|
|---|
| 关键词 |
|
|---|
| Committee Chair / Thesis Advisor |
|
|---|
| Committee Members |
|
|---|
| Partnering Agencies |
|
|---|