1. National STD Conference 2008
Association Between Enhanced STD Screening and Declines in ED and Hospital Diagnoses of Chlamydia trachomatis and Neisseria gonorrhoeae: Philadelphia,
Caroline Johnson, MD, Greta Anschuetz, MPH, Lenore Asbel, MD, Melinda Salmon, Martin Goldberg, C. Victor Spain, DVM, PhD
National STD Conference 2008
Chicago, IL

2. Background
Availability of non-invasive STD testing permits Health Departments to conduct high volume screening programs for Neisseria gonorrhoeae (GC) and Chlamydia trachomatis (CT)
Aggressive screening programs have not necessarily been accompanied by reductions in case rates
Because screening for GC and CT facilitates recognition of asymptomatic and/or early infections, disease complications may be reduced

3. Objectives
To determine if an association exists between expanded STD screening and preventing complications of CT and GC infections
Complications of CT/GC infections were determined using two data sources
Women reported with positive CT or GC lab test from a hospital or Emergency Department (ED cases)
Women hospitalized for Pelvic Inflammatory Disease (PID)
Examine the cost effectiveness of screening programs for preventing complications

4. Methods Ecologic study from 1996-2006
Population: Philadelphia, 1.5 million
Data sources:
Lab records of STD screening tests supported by public health funding
Citywide STD morbidity database
Pennsylvania Health Care Cost Containment Council (PHC4) database
mandatory hospital reporting on uniform claims and billing for all Philadelphia hospitalizations
contains case-based demographic data, ICD9 discharge code data, and cost data

5. Methods: Case Definition
ED Cases
Women, ages yrs, residents of Philadelphia
Positive CT or GC lab test reported to DPH
Report originated from hospital or ED (excluded cases reported for STD screening purposes)
PID Cases
Case data reported to PHC4
ICD9 code for discharge diagnosis consistent with PID ( )
(excluded cases with ICD9 codes related to neoplasm)

6. Methods: Variables Outcome variables: Predictor variables:
# of ED cases
# of PID cases
Predictor variables:
# of screening tests performed
Represented as total tests and stratified by gender
High prevalence population
PID costs obtained from PHC4 data
Median cost for women with PID as sole discharge diagnosis
Screening program costs obtained from public health program expenses

7. Screening Tests in Philadelphia, by gender
Female Tests
Male Tests
1996
56,291
1997
81,216
1998
80,433
1999
79,801
2000
83,057
4,741
2001
83,723
6,647
2002
80,472
28,359
2003
99,679
53,644
2004
101,530
57,564
2005
101,884
54,639
2006
103,946
57,785

8. Methods: Analyses
Examined the association between screening in high prevalence populations and ED cases
Linear regression analyses examined the association between screening and PID cases
Simple linear regression looked at the association between total screening and PID cases
Multivariable linear regression used to determine the effect of screening females and males on the trend of PID cases
Cost analysis performed to determine if screening is cost effective

9. Temporal Association Between Increased CT/GC Screening and ED Cases
Year
# Screening Tests Performed
# ED Cases

10. Temporal Association Between Increased CT/GC Screening and PID Cases
Year
# Screening Tests Performed
# PID Cases

11. Simple Linear Regression Analysis
For every 10,000 screening tests performed,
28 cases of PID were prevented
# PID prevented = (# screening tests/10,000)
Number Hospitalized PID cases
R2=0.93
p<0.001
Number of screening tests

12. Multivariable Regression
For each 10,000 females screened, the reduction in PID cases observed is:
35.8 cases (95% CI: 6.0 to 65.6)
By screening 10,000 males, in addition to 10,000 females, the additional reduction in PID cases observed is:
23.5 cases (95% CI: 7.6 to 39.5)
Output from multivariable regression:
Expected Hospitalized PID cases = 970.9
- 23.5(Males tested/10,000)
- 35.8(Females Tested/10,000)

13. Cost Analysis Cost of hospitalization for PID cases in 2006
Range from $4,713 - $84,985/per case
Median = $18,881/case
Cost of screening program
GC/CT prevalence in Philadelphia screening program = 6%
Screen and treat a person with a positive test result= $52
Screen a person with a negative result = $13

14. Cost Analysis
For every additional 10,000 screening tests performed in a high prevalence population:
Cost savings for 28 cases of PID averted = $528, (28 x $18,881/case)
Costs of 600 positive tests plus 9,400 negative tests =$153, (600 x $52) + (9,400 x $13)
Total cost savings of $375, ($528,668 - $153,400)
If the cost to treat one case of hospitalized PID exceeds $5,479, screening is cost saving
($153,400 / 28)

15. Conclusions
Screening for GC/CT is associated with declines in complications of infection, both for ED cases and hospitalized PID cases
Both female and male screening contribute to the decline in complications of infection
Screening activity in Philadelphia provides cost savings to the healthcare system

16. Thanks! Questions? Lenore Asbel, MD - lenore.asbell@phila.gov
Greta Anschuetz, MPH -

17. Limitations of Analyses
Strength of Analyses
Two unrelated data sources demonstrate same association between screening and outcome
Limitations of Analyses
PHC4 data only give information on hospitalized cases
No information on cases treated on an outpatient basis
Underestimating the cost savings without outpatient information
PHC4 is a de-identified data source
Not able to link to reported STD cases to assess how many cases actually have a CT or GC diagnosis.

18. 3-D representation of the Multivariable Regression Analysis Expected PID cases = – 23.5 (#Males Screened/10,000) – 35.8(#Females Screened/10,000)
PID cases
Males Screened
Females
Screened