Table 1.
· Description, point values, range and sources for all variables used in a cost-effectiveness model of different algorithms of HIV and syphilis testing in pregnancy·
| Variable | Base-case value | Range | Sources |
|---|---|---|---|
| Maternal characteristics | |||
| Median age at first birth | 20 | w18 | |
|
| |||
| Disease prevalence | |||
| HIV Prevalence among pregnant women | 10·6% | 5·3%–15·9%† | 28 |
| AIDS prevalence among HIV infected pregnant women | 24·82% | Assumption | |
| Syphilis prevalence in HIV uninfected pregnant women | 1·09% | 0·54%–1·63%† | 28 |
| Syphilis prevalence in HIV infected pregnant women | 2·17% | 1·09%–3·26%† | 28 |
| History of syphilis infection (Adequately treated previous infections) in HIV uninfected women | 5% | 2·5%–7·5%† | Assumption |
| History of syphilis infection (Adequately treated previous infections) in HIV infected women | 10% | 5%–15%† | Assumption |
|
| |||
| Disease progression | |||
| Average progression time HIV to AIDS in treated child | 10 years | Assumption | |
| Average progression time AIDS to death in treated child | 5 years | Assumption | |
| Average progression time HIV to AIDS in untreated child | 1 year | Assumption | |
| Average progression time AIDS to death in treated child | 1 year | Assumption | |
|
| |||
| Test performance | |||
| Syphilis test sensitivity | |||
| Dual HIV and syphilis test | 0·89 | 0·84–0·94 | 20 |
| Syphilis rapid | 0·82 | 0·63–0·97 | 30 w1 w2 |
| Syphilis laboratory-based | 1 | 24 | |
| Syphilis test specificity | |||
| Dual HIV and syphilis test | 0·99 | 0·97–1.00 | 20 |
| Syphilis rapid | 0·96 | 0·92–0·99 | 30 w1 w2 |
| Syphilis Laboratory-based | 1 | w16 | |
| Syphilis test specificity in those with previously treated syphilis infection | |||
| Dual HIV and syphilis test | 0·91 | Assumption | |
| Syphilis rapid | 0·91 | w1 | |
| Syphilis laboratory-based | 1 | ||
| HIV Test sensitivity | |||
| Dual HIV and syphilis test | 0·99 | 0·95–1·00 | 20 |
| HIV rapid | 1 | w10 | |
| HIV Test specificity | |||
| Dual HIV and Syphilis test | 0·99 | 0·997–·999 | 20 |
| HIV rapid | 0·96 | 0·85–1 | w10 |
|
| |||
| Loss to Follow-up | |||
| Syphilis loss to follow up (for laboratory tests) | 20% | w19 | |
| HIV loss to follow-up | 23·9% | w20 | |
|
| |||
| HIV and Syphilis Treatment | |||
| Infants born to women who test positive for HIV and receive treatment who receive Nevirapine | 77·19% | Assumption | |
| Proportion of children known to be HIV-exposed who were enrolled in ART | 67% | 28 | |
| Probability of syphilis treatment success for the mother | 98% | w16 | |
|
| |||
| Pregnancy outcomes | |||
| Syphilis Uninfected mothers | |||
| Stillbirth/early fetal death | 4·6% | 3·0%–7·1% | 3 |
| Neonatal death | 3% | 2·1%–4·3% | 3 |
| Prematurity or low birth weight | 6·3% | 3·5%–11·0% | 3 |
| MTCT of HIV in HIV treated mothers (in utero + intrapartum/postnatal, 12 month) | 4·99% | w21 | |
| MTCT of HIV in HIV untreated mothers (in utero + intrapartum/postnatal, 12 month) | 24.17% | w22 | |
| Syphilis infected mothers (syphilis untreated) | |||
| Congenital syphilis | 15·5% | 7·5%–29·0% | 3 |
| Prematurity or low birth weight | 12·1% | 3·9%–31·8% | 3 |
| Neonatal death | 12·3% | 9·3%–16·2% | 3 |
| Stillbirth/early fetal death | 25·6% | 18·5%–34·2% | 3 |
| Additional effects in HIV co-infected mothers (HIV treated) | |||
| MTCT of HIV in HIV treated mothers (in utero + intrapartum/postnatal) | 9·05% | 29 | |
| Prematurity or low birth weight | 2·74% | 3 | |
| Additional effects in HIV co-infected mothers (HIV untreated) | |||
| MTCT of HIV | 34·5% | 29 | |
| Prematurity or low birth weight & MTCT of HIV | 9·36% | 3,29 | |
|
| |||
| Costs (2012 USD) | |||
| Labor Costs | |||
| Pre-test counseling (both HIV & syphilis) | $0·44 | w5 Φ | |
| Sample collection (single test) | $0·27 | w5 Φ | |
| Preparing and inoculating test (single test) | $0·37 | 24 | |
| Reading and recording results (single test) | $0·63 | 24 | |
| Post-test counseling, syphilis positive | $0·62 | w5 Φ | |
| Post-test counseling, syphilis negative | $0·18 | w5 Φ | |
| Post-test counseling, HIV positive | $1·24 | w5 Φ | |
| Post-test counseling, HIV negative | $0·18 | w5 Φ | |
| Patient cost | |||
| Travel cost | $1·49 | Assumption | |
| Testing time cost (dual test) | $0·48 | 24 | |
| Testing time cost (single tests) | $0·71 | Assumption | |
| Test Cost | |||
| Single syphilis rapid test | $0·55 | WHO Catalog | |
| Single HIV rapid test | $0·81 | UNICEF agreement | |
| Laboratory-based syphilis tests | $2·53 | w13 | |
| Dual HIV and syphilis test | $1·30 | $1·20–$2·60 | Assumption |
| Early infant diagnosis | $32·50 | w23 | |
| Treatment for syphilis (2·4 MU benzathine penicillin) | $2·38 | w16 | |
| Pregnancy outcome cost | |||
| Healthy | $72·59 | 36·30–108·89† | w16 |
| Congenital syphilis | $804·54 | 402·27–1206·81† | w16 |
| Premature | $1508·84 | 754·42–2263·26† | w16 |
| Neonatal death | $3577·58 | 1788·79–5366·37† | w16 |
| Stillbirth | $72·59 | 36·30–108·89† | w16 |
| MTCT HIV | $1358·18 | 679·09–2037·27† | w24 |
| Nevirapine for 12 months for infant | $16·60 | 8·30–24·90† | |
|
| |||
| Disability weights | |||
| AIDS (no treatment) | 0·545 | 25 | |
| HIV | 0·053 | 25 | |
| Death | 1 | - | |
| Congenital syphilis (3 years) | 0·315 | 26 | |
| Low birth weight (1 year) | 0·106 | 26 | |
| Neonatal death | 1 | - | |
| Stillbirth | 0 | - | |
| Miscarriage | 0 | - | |
|
| |||
| Test Uptake | |||
| Syphilis rapid test | 0·08 | 0·04–0·12† | w9 |
| Syphilis laboratory test | 0·08 | 0·04–0·12† | w9 |
| HIV rapid test | 0·83 | 0·415–0·95** | w9 |
| Dual HIV and syphilis test | 0·83 | 0·415–0·95** | Assumption |
| Early infant diagnosis | 0·9 | w23 | |
|
| |||
| Other | |||
| Life expectancy of newborn | 50 years | w25 | |
| Discount rate | 3% | w26 | |
NOTE: Women are assumed to have unknown HIV and syphilis status at time of testing·
**
ranges are calculated using a 50% reduction for the lower bound and an assumption for the top
†
ranges calculated using a 50% spread
¥
lower end of range is another common way to determine the disability weight of co-infection: highest disability weight of the two infections
Φ
Labor costs were calculated using salaries published by the WHO (w5) and assumed times for length of each procedure