Transcript Master slide

How To Collect and Evaluate
Surveillance and Epidemiologic Data
For Hepatitis C
Miriam J. Alter, Ph.D.
Division of Viral Hepatitis
Centers for Disease Control and Prevention
Atlanta, GA USA
Surveillance for Hepatitis C Virus Infection
Acute
disease
All infected
persons
Detect outbreaks
Assess disease/infection burden
Monitor trends
Identification & follow-up of
infected persons
Develop, implement, evaluate
prevent programs
Guide allocation of resources
Surveillance Components

Case reporting

Prevalence assessment and monitoring

Special studies
Establishing Surveillance



Standardized case definitions
Reliable laboratory reporting
Infrastructure for identifying cases
– Identify biases that affect interpretation of data
• Generalizability of cases identified
• Testing and reporting practices

Determine types of information to be collected
Anti-HCV EIA RR Results by RIBA™ 3.0
and Population Tested
RIBA 3.0
Positive
Indeterminate
Negative
100
Percent
80
60
40
20
0
Students
HCWs
NHANES
STD
Dialysis
Source: LY Hwang, Houston; R. Gunn, San Diego; S. Harris, Austin;
I. Weisfuse, NYC; CDC, Atlanta
Hi-Risk
Proportion of Anti-HCV RR EIA Results Testing
RIBA™ Positive by S/CO Ratio
Students
HCWs
NHANES IV
STD
Dialysis
Hi-risk
Percent RIBA Positive
100
80
60
40
20
0
1.0-<3.0
(N=231)
3.0-<3.5
3.5-<3.8
(N=18)
(N=21)
EIA S/CO Ratio
Source: LY Hwang, Houston; R. Gunn, San Diego; S. Harris, Austin;
I. Weisfuse, NYC; CDC, Atlanta
3.8+
(N=765)
ALT levels in HCV-infected persons
Acute Hepatitis C ( n=267)
80
7 x ULN
70
60
%
50
40
15%
30
85%
20
10
0
Chronic HCV infection (n=4702)
60
50
30
3%
97%
20
10
ALT x ULN
15+
14-14.9
13-13.9
12-12.9
11-11.9
10-10.9
9-9.9
8-8.9
7-7.9
6-6.9
5-5.9
4-4.9
3-3.9
2-2.9
1-1.9
0
0-0.9
%
40
Epidemiologic Studies

Identify persons at risk for infection

Determine amount of disease/infection
attributable to each risk factor

Provide guidance for surveillance and
prevention programs
Types of Epidemiological Studies

Cohort (prospective) - direct estimate of risk
– Presence of exposure determined in sample of population
– Entire sample followed and incidence of disease compared for
those with and without the exposure

Case control (retrospective) - indirect estimate of risk
– Sample selected based on presence or absence of disease
– Proportion of cases with history of exposure before onset of
disease compared with controls

Cross-sectional or prevalence - associations
– Presence of disease determined in sample of population
• Proportion of cases with history of exposure compared with non-cases
• Prevalence of disease compared for those with and without the exposure
– Temporal sequence of exposure relative to disease unknown
Cohort Studies



Directly measure relative risk and population
attributable risk
Require large sample sizes, long follow-up,
expensive
Only evaluate a single exposure
Case Control Studies

Sample size, logistics, and expense reasonable

Odds ratio good estimate of risk if certain
assumptions met
– Frequency of disease in population small
• <2% incidence/year
– Cases and controls representative -- CRITICAL

Will not detect rare events
Prevalence Studies



Logistics less complex, less expensive
Determining specific exposures preceding infection
problematic when onset unknown or many years ago
Substantial differences in methodology
– Population-based
– Highly selected groups
• Blood donors
• Clinic patients

Inconsistent results among studies
– Under-ascertain some risk factors
– Cannot generalize to the rest of the population
Sources of Study Populations That Affect
Reliable Interpretation of Results
Controls
Cases


Disease vs asymptomatic
Single source
– Referral (e.g., GI clinic)
– Clinics serving
disadvantaged population
– Highly specialized setting
for specific condition

Case reports




Blood donors
Family member
Cases of other types of
viral hepatitis
Single disease group
Risk Factors Associated With
Acquiring HCV Infection, United States
Cohort and Case Control Studies






Transfusion, transplant
Injecting drug use
Occupational blood exposure (needle sticks)
Birth to an infected mother
Infected sex partner
Multiple heterosexual partners
Exposures Not Associated With Acquiring HCV
Case Control Studies of Acute Hepatitis C, U.S., 1979-1985
Exposure (prior 6 months)
Cases
n=148
Controls
n=200
Medical care procedures
Dental work
Health care work (no blood contact)
Ear piercing
Tattooing
Acupuncture
Incarceration
Foreign travel
Military service
30.4%
24.3%
4.1%
2.7%
0.7%
0
4.1%
4.1%
1.3%
29.5%
23.5%
5.0%
3.0%
0.5%
1.0%
1.0%
2.5%
4.9%
Source: JID 1982;145:886-93; JAMA 1989;262:1201-5.
HCV Related to Health-Care Procedures
United States


Not associated with sporadic or background
Recognized primarily in context of outbreaks
– Contaminated equipment
• Hemodialysis
– Unsafe injection practices
• Plasmapheresis
• Multiple dose medication vials
– Hospitalized patients
– Private practice
• Home infusion therapy
Cross-sectional/Prevalence Studies of HCV
Variation of Results – Low Prevalence Countries
Exposure
Injecting drug use
Transfusion
Tattooing
Nasal cocaine use
Ear/body piercing
Acupuncture
Incarceration
Donors
US US AU UK
+
+
+
+/-
+
+
+
+
+
+
+
ND
ND
ND
-
+
+
ND
ND
Patients
College
GI Spinal VA Students
ND
ND
+
ND
ND
+
+
ND
ND
ND
+
+
+
+
+
+
+
ND
+
History of Tattooing and Acute Hepatitis C
1982-2000, United States
Time period of
reported case
History of tattooing prior 6 mo
All patients
No IDU/BT
N
Total (95% CI)
1856
3.2% (2.5-3.8) 1.5% (0.9-2.1)
1982-1986
839
2.7%
1.8%
1987-1990
625
2.7%
1.1%
1991-2000
392
4.3%
1.5%
Source: CDC Sentinel Counties Study
History of Body Piercing
Acute Hepatitis B and Acute Hepatitis C
1996-2000, United States
Type
History of piercing* prior 6 mo
N
All patients
No IDU
Acute hepatitis B
603
2.3%
1.5%
Acute hepatitis C
134
3.7%
1.0%
* Other than ears
Source: CDC Sentinel Counties Study
Tattoos and HCV Infection
Cross-sectional, GI Clinic, Albuquerque, 95-96

40% Hispanic, 40% indigent
– Cases – referred for positive HCV test
– Controls – gastroesophageal reflux disease
• HCV status not ascertained
Exposure
Positive (total tested)
Cases
Controls
IDU or BT
87% (477)
Tattooed
Subset (no IDU/BT) 43% (58)
Of total cases
5% (477)*
Adj. (95%
OR CI)
NA
16% (58)
1% (58)
5.9 (1.1-30.7)
*Attributable fraction 0.8% (estimated from data)
Balasekaran et al. Am J Gastro 1999;94:1341-6
Limitations
Balasekaran et al. Am J Gastro 1999;94:1341-6

Representativeness
– Cases not representative of all persons with HCV
– Controls not representative of nondiseased persons
• Is prevalence of characteristic under study same in control group as
in the general population?
• Possible selection bias from using a single disease group
• Not tested for HCV


History of incarceration not ascertained
Even if tattooing associated with HCV in this group,
accounts for <1% of infections
Tattoos and HCV Infection
Prevalence, Orthopedic Spinal Clinic, Dallas, TX 91-92


Over represented blacks, hisp, men, middle/low income
43/626 HCV positive (6.9% sample; 2.8% standardized)
% HCV positive
Exposure
Yes
No
Tattoo
22%
3.5%
Commercial parlor
33%
3.5%
Beer drinker
12%
5%
Injection drug use
37.5% 5%
>1 yr
58%
5%
Male Ancillary HCW 32%
6%
Transfusion
4%
7.5%
Adj. (95%
OR CI)
NA
6.5 (2.9-14.8)
4.0 (1.8-8.7)
NA
23.0 (7.5-70.6)
9.6 (3.8-24.3)
NS
AR*
41%
(30%)
23%
17%
(14%)
8%
--
*Attributable risk % adjusted for other risk factors and standardized to population
Source: Haley et al. Medicine 2001;80:134-51.
Limitations
Haley et al. Medicine 2001;80:134-151




Population not representative
Inconsistent with virtually all other studies
Dose response relationships inconsistent for tattooing,
but not for IDU
IDU likely under-reported
– >50% of HCV-positives admit to IDU when re-interviewed
after receiving results

Some factors likely surrogates for known risks
– Male ancillary HCW (why not females?)
– Beer drinking (why not other forms of alcohol?)
HCV and HBV Among College Students 18-35 yrs old,
U.S., 2000-2001
Characteristic
Transfusion
Yes
No
IDU
Yes
No
Tattoo
Yes
No
Body piercing Yes
No
Snorted drugs Yes
No
* p<.001
† excluding IDU and transfusion
Hwang et al., unpublished data
Total Tested (%)
337 (4.5)
7236 (95.5)
116 (1.5)
7718 (98.5)
1430 (20.5)
5533 (79.5)
1202 (17.4)
5701 (82.6)
617 (9.1)
6179 (90.9)
% Positive
HCV HBV
6.2* 11.8*
0.7
5.6
22.4* 17.1*
0.6
5.7
0.3† 5.3
0.5
6.2
0.4† 3.7
0.4
6.5
0.6† 6.8
0.4
5.9
Geographic Differences in
HCV Transmission Patterns
Exposures among
prevalent infections
Importance of Exposures by
HCV Endemicity
Low Moderate High
++++
Transfusions (unscreened) +++
Injecting drug use
Health-care related
Contaminated equipment +/Unsafe injections
+/Folk medicine
-
++
+++
+
+++
++++
++++
++
++++
++++
No data
HCV Related to Therapeutic and Cosmetic
Procedures in Moderate/High Endemic Countries

Associated with “background” infections in
some studies
– unsafe therapeutic injections
– hospitalization, surgery, dental work
• Control populations may not have been representative
– Acupuncture (one village in Japan)

Geographic clustering by age, town, region
– considerable variation within and between
countries
Health-Care Procedures and HCV Infection
Moderate Endemic Countries
Country
Case-Control
Italy
Cross-Sectional
Italy
Taiwan
Pakistan
Japan
Surgery
HCV Pos HCV Neg
17%*
2%
56%*
36%
77%
57%
13%
3%
No data
32%*
10%
* P<.05, independent of other risk factors
Dental
HCV Pos HCV Neg
22%*
11%
91%*
80%
90%
90%
24%
28%
33%
39%
No data
Cosmetic Procedures and HCV Infection
Moderate Endemic Countries
Country (author)
Case control
Taiwan (Chen)
Cross sectional
Taiwan (Sun)
(Ho)
Japan (Kiyosawa)
Pakistan (Luby)
Korea (Kim)
* Ear piercing women only
Tattooing
HCV Pos HCV Neg
Body Piercing
HCV Pos HCV Neg
0%
0%
0%
1%
3%
21%
1%
3%
7%
11%
1%
34%
0%
0%
0%
7%
-78%
--7%
14%
-83%*
--0%
20%