A Unified Optimal Resource Allocation Model for Screening and Treating Asymptomatic Women for Chlamydia Trachomatis and Neisseria Gonorrhoeae

1
A Unified Optimal Resource Allocation Model for
Screening and Treating Asymptomatic Women for
Chlamydia Trachomatis and Neisseria Gonorrhoeae

• Abban B, Tao G, Gift T, Irwin K
• Centers for Disease Control and Prevention (CDC)

2
Background

• Up to 70 CT and up to 50 GC infections are
  asymptomatic
• CT infection among GC infected populations can be
  as high 50
• Different segments of the population have
  different prevalences of CT, GC, and
  co-infection range of disparities is wide
• Availability of different testing technologies at
  varying cost and performance
• Many clinics operate under fixed budgets and
  cannot accommodate universal screening

3
Study Objective

• Determines the optimal combination of screening
  coverage, test selection and treatment for CT and
  GC in asymptomatic women specifically

• At what prevalence is it cost-saving to screen a
  population for CT or GC?
• Is it more beneficial to screen with more
  sensitive but more expensive tests?
• Is presumptive treatment cost-saving?

4
Clinical Management Decision
5
Clinical Alternatives Considered
For each risk-group the following strategies are
possible

• Screen and treat for CT only
• Screen and treat for GC only
• Screen and treat for both CT and GC
• Screen and treat for CT only and presumptively
  treat for GC
• Screen for and treat for GC only and
  presumptively treat for CT

6
Methods

• The optimal strategy was defined as one that
  maximized
• the number of women cured or
• the cost-saving value (cost of averted PID minus
  screening and treatment costs for CT and/or GC)
• Selective screening based on readily ascertained
  risk-factor Age
• 4 tests each for CT and GC, including dual
  test(s)
• 2 treatment regimens for CT and 3 for GC
• A mixed integer optimization model for a
  hypothetical cohort of 1000 asymptomatic women

7
Model Assumptions

• All women who visited the clinic lacked symptoms
  of CT and GC infections
• A strategy could allow the screening of selected
  age groups or all patients
• Return rate for treatment was assumed to be the
  same for all age groups
• Test and treatment for each infection were the
  same all age groups

8
Variables

• CT and GC positivity by age group
• Co-infection rates by age group
• Tests sensitivity, specificity and cost
• Treatments effectiveness and cost
• All parameter values were from published
  literature

9
Test Positivity Ratesby Clinic Type

• Family planning clinic

• STD clinic

10
Variables - Test
BDPT Becton Dickinson Probe Tec
11
Variables - Treatment
12
Clinical Costs and Outcomes
13
Results
14
Test Positivity at which Screening is Cost-saving

• Sensitive to PID cost

15
Results FP Clinic
CT (2.3 - 10.6), GC (0.4 - 1.2), GC with CT
(30.0 - 46.0)
All costs in US dollars (2003) BDPT Becton
Dickinson Probe Tec Optimal cost-saving
strategy pres. presumptively treat
16
Results FP Clinic
CT (2.3 - 10.6), GC (0.4 - 1.2), GC with CT
(30.0 - 46.0)
All costs in US dollars (2003) BDPT Becton
Dickinson Probe Tec Optimal cost-saving
strategy pres. presumptively treat
17
Results FP Clinic
CT (2.3 - 10.6), GC (0.4 - 1.2), GC with CT
(30.0 - 46.0)
All costs in US dollars (2003) BDPT Becton
Dickinson Probe Tec Optimal cost-saving
strategy pres. presumptively treat
18
Results STD Clinic
CT (3.0 12.5), GC (2.0 8.1), GC with CT
(20.0 45.5)
All costs in US dollars (2003) BDPT Becton
Dickinson Probe Tec Optimal cost-saving
strategy pres. presumptively treat
19
Results STD Clinic
CT (3.0 12.5), GC (2.0 8.1), GC with CT
(20.0 45.5)
All costs in US dollars (2003) BDPT Becton
Dickinson Probe Tec Optimal cost-saving
strategy pres. presumptively treat
20
Results STD Clinic
CT (3.0 12.5), GC (2.0 8.1), GC with CT
(20.0 45.5)
All costs in US dollars (2003) BDPT Becton
Dickinson Probe Tec Optimal cost-saving
strategy pres. presumptively treat
21
Limitations

• The alternative of screening and treating for CT
  and screening CT-positives for GC was not
  considered
• Published range of values for direct cost
  attributable to PID is wide (1,433 5,000)
• Repeat infections were not considered
• CT and GC positivity in asymptomatic STD clinic
  patients may be less than the reported
  population-wide rates

22
Conclusions

• Optimal control strategy varies with CT and GC
  positivity, CT-GC co-infection rates, total
  program budget, test costs and PID cost
• Influence of treatment cost on overall program
  cost is minimal
• A switch from one test to another may not yield
  significant change in the number of women cured
• The optimal strategy from a cost-saving
  perspective and from a number-of-cures
  perspective may vary
• The model provides a flexible tool to analyze
  different scenarios when identifying a control
  strategy for CT, GC, or both