Table 1.
Model Inputs
| Input Parameter | Disease Group | Baseline Assumption | Source |
|---|---|---|---|
| Natural History Component | |||
| Age at death from competing cause | All | N/A | Vital Statistics of the US [16] |
| Incidence and age at clinical diagnosis for cases | Cases Latent | N/A | SEER [15] |
| Incidence and age at clinical diagnosis for benign disease | Benign | N/A | Katsube [17] |
| Tumor characteristics (stage, histology and grade at clinical diagnosis) | Cases Latent | N/A | SEER [15] |
| Malignant disease duration and stage lengths | Cases Latent | See Table 2 | |
| Benign Disease duration | Benign | 9 years | PLCO [26] |
|
| |||
| Screening Component | |||
| Screening frequency | All | Annual (Age 45–85) | N/A |
| CA125 sensitivity | Cases Benign | See Figure 2 | CARET [7] |
| CA125 specificity | Healthy | 95% | Defined by the screening algorithm |
| Hypothetical Marker sensitivity | Cases Benign | 2X sensitivity of CA125 | |
| Hypothetical Marker specificity | Healthy | 95% | |
| TVS sensitivity | Cases | 63% | PLCO [26] |
| TVS specificity | Healthy | 97% | PLCO [26] |
| Hypothetical Imaging sensitivity | Cases | 90% | |
| Hypothetical Imaging specificity | Healthy | 97% | |
|
| |||
| Survival Component | |||
| EOC survival contingent on age and tumor stage, histology and grade at diagnosis | Cases | See Figure S4 | SEER [19] |
|
| |||
| Cost Component* | |||
| CA125 test cost | All | $31 | Havrilesky [12] |
| Hypothetical Marker test cost | All | $210 | |
| TVS test cost | All | $111 | Havrilesky [12] |
| Hypothetical Imaging test cost | All | $750 | |
| Laparoscopy with BSO | Benign Healthy | $4206 | Havrilesky [12] |
| EOC Treatment Costs | Cases | Yabroff [21] | |
| Initial year | Stage I | $36,671.66 | |
| Stage II | $50,718.96 | ||
| Stages III/IV | $70,452.02 | ||
| Continuing Care | All stages | $4,712.30 | |
| Last year of life** | Stage I | $27,523.12 | |
| Stage II | $46,437.70 | ||
| Stages III/IV | $69,313.90 | ||
*
All costs were adjusted to 2010 US dollars.
**
The treatment cost differential for the last year of life between early and late stage diagnoses arises by the way Yabroff et al. allocated treatment costs for patients surviving less than 24 months past diagnosis. Costs for the last 12 months of this period were allocated as last year of life costs, and the remainder considered initial year treatment costs. Our model treats cost allocation for such patients in a similar fashion. Allocation of death related costs may increase initial year treatment costs for women who die within 12 months of diagnosis.