Selected univariate sensitivity analyses of the cost effectiveness of ELISA→WB relative to no screening
| Projected number of new infections | |||
|---|---|---|---|
| Variable | No screening | ELISA→WB | Incremental cost effectiveness of ELISA→WB relative to no screening ($/infection averted) |
| HSV-2 = herpes simplex virus type 2; ELISA = enzyme linked immunosorbent assay; WB = western blot. All cost effectiveness ratios rounded to the nearest $100. | |||
| *Lower bound values for prevalence (6% in men and women) were based on prevalence in individuals younger than 20 years of age in NHANES-III, while upper bound values (24% in men, 43% in women) were based on prevalence in African-American individuals aged 20–29 in NHANES-III.3 | |||
| †Although aciclovir markedly reduces viral shedding in individuals with asymptomatic HSV-2 infection,17 data pertaining to the drug’s effectiveness in preventing HSV-2 transmission are currently lacking. In these analyses, aciclovir was estimated to reduce the monthly probability of HSV-2 infection by 95%, at a cost of $40 per month. | |||
| ‡When specificity of ELISA was greater than or equal to 99%, ELISA alone ceased to be dominated by the strategy of ELISA→WB, so that the incremental cost effectiveness ratio of $7800 per infection averted with ELISA specificity of 100% applies to the use of ELISA alone. | |||
| §In the base case, it was assumed that neonatal HSV-2 infection occurred exclusively in the context of incident third trimester infection in the mother. When transmission by women with chronic infection was added to the model, neonates were also at risk of infection if born to women who had prevalent unrecognised disease prior to initiation of a monogamous relationship, and to women who acquired HSV-2 within the context of a monogamous partnership, but outside the third trimester of pregnancy. The incremental cost per neonatal infection averted through the use of ELISA→WB decreased to $2.90 million with the addition of a 1/10 000 risk of transmission by chronically infected women, and to $1.86 million when the risk was increased to 1/2000. | |||
| || As the probability of maternal-fetal transmission following third trimester infection increased from 5% to 25%, the incremental cost per neonatal infection averted decreased from $6.3 million to $1.2 million. | |||
| ¶The impact of decreasing infectiousness, frequency of intercourse and condom compliance were simulated using a discount function. In the case of decreasing infectiousness, the risk of infection was discounted at an annual rate of 10–30%, while decreasing frequency of intercourse was simulated by discounting both the risk of infection and condom costs. Decreasing condom compliance was simulated by discounting the cost and efficacy of condoms in a similar manner. | |||
| **In this analysis, baseline infection in the female partner was not assumed to be independent of infection in the male partner. In other words, choice of sexual partner was modelled as assortative, so that an infected man was more likely to choose an infected female partner than would have been expected by chance alone. | |||
| Calculated as a weighted average of costs of primary HSV-2 without hospitalisation (two clinic visits, 7 day course of acyclovir, and 2 days off work), and with hospitalization (costs of hospitalisation and medications with HSV-2 as a primary diagnosis and cost of 1 week off work). | |||
| ††We assumed that that the prevalence of HSV-2 in concurrent partners was equivalent to that seen in the general population, and that the rate of transmission per unit time of HSV-2 within a concurrent partnership was the same as that seen in an ostensibly monogamous long term relationship. | |||
| Baseline values | 170 | 132 | 8200 |
| Initial population prevalence* | |||
| 0.06 | 61 | 50 | 24 600 |
| 0.34 | 303 | 233 | 5300 |
| Annual probability of infection by partner | |||
| Highest | 362 | 312 | 12 600 |
| Lowest | 139 | 91 | 7300 |
| Probability of symptomatic disease after infection | |||
| 0.09 | 230 | 177 | 7700 |
| 0.50 | 148 | 114 | 8900 |
| Probability of primary genital herpes syndrome among symptomatic individuals | |||
| 0.35 | 170 | 132 | 8200 |
| 0.80 | 170 | 132 | 8200 |
| Relative risk of transmission with regular condom use | |||
| 0.25 | 170 | 91 | 3600 |
| 0.75 | 170 | 150 | 22 400 |
| Aciclovir used to prevent transmission† | 170 | 30 | 12 700 |
| ELISA sensitivity | |||
| 80% | 170 | 135 | 9500 |
| 100% | 170 | 128 | 8000 |
| ELISA specificity‡ | |||
| 80% | 170 | 132 | 8700 |
| 100% | 170 | 132 | 7800 |
| Probability of maternal-fetal transmission with chronic infection§ | |||
| 1/10000 | 170 | 132 | 8200 |
| 1/2000 | 170 | 132 | 8100 |
| Probability of maternal-fetal transmission following third trimester infection || | |||
| 5% | 170 | 132 | 8400 |
| 25% | 170 | 132 | 7900 |
| Annual decrease in frequency of intercourse¶ | |||
| 10% | 113 | 75 | 6800 |
| 30% | 62 | 36 | 9800 |
| Annual decrease in infectiousness¶ | |||
| 10% | 113 | 75 | 8500 |
| 30% | 62 | 36 | 13 500 |
| Annual decrease in condom compliance¶ | |||
| 10% | 170 | 159 | 25 800 |
| 30% | 170 | 166 | 65 800 |
| Conditional probability of baseline infection in female partner of infected male** | |||
| 30% | 150 | 115 | 8900 |
| 50% | 122 | 92 | 10 200 |
| Average annual number of concurrent partnerships†† | |||
| 2 | 339 | 312 | 12 200 |
| 6 | 624 | 611 | 25 100 |
| Expected duration of relationship (years) | |||
| 1 | 21 | 12 | 25 700 |
| 5 | 76 | 47 | 8300 |
| 10 | 179 | 145 | 7000 |
| Discount rate | |||
| 0% | – | – | 9200 |
| 5% | – | – | 7800 |
| 10% | – | – | 7400 |
| Monthly cost of condoms | |||
| $1 | – | – | 6400 |
| $10 | – | – | 14 900 |
| Cost of western blot | |||
| $45 | – | – | 8100 |
| $90 | – | – | 8500 |
| Cost of ELISA | |||
| $3 | – | – | 8100 |
| $35 | – | – | 9800 |