Sensitivity analysis—effect of varying (selected) assumptions and values, on national prevalence estimates for adults aged 15–49 years at years 2000 and 2016: (a) syphilis; (b) gonorrhoea
| Parameter | Default value | Alternative assumption | Prevalence Zimbabwe, women, at 2000→2016 | Prevalence Morocco, women,at 2000→2016 | |
|---|---|---|---|---|---|
| Lower | Upper | ||||
| (a) Syphilis | |||||
| Default best syphilis prevalence estimates | 1.94→1.54% | 1.48%→0.55% | |||
| Diagnostic test performance, relative to RPR-positive plus TPHA-positive (dual positivity) as ‘gold standard’ | |||||
| TPHA in ANC or FP population | 0.8 | 0.6 | 1.0 | Lower value: 1.73→1.58% Upper value: 2.11→1.49% | Lower value: 1.45→0.36% Upper value: 1.57→0.74% |
| RPR | 0.6 | 0.45 | 1.0 | Lower value: 1.99→1.38% Upper value: 1.84→1.99% | Lower value: 1.22→0.58% Upper value: 2.27→0.46% |
| Rapid syphilis test (TPHA-based) | 0.7 | 0.55 | 1.0 | Lower value: 1.90→1.34% Upper value: 1.96→1.96% | Not used |
| (b) Gonorrhoea | |||||
| Default best gonorrhoea prevalence estimates | 2.5→3.8% | 0.60→0.36% | |||
| Diagnostic test performance—gonorrhoea | |||||
| Specificity PCR and LCR for gonorrhoea, women, genital fluid | 99.7% | 99.0% | 100% | Lower value: 2.3→3.6% Upper value: 2.7→4.0% | Lower value: 0.28→0.11% Upper value: 0.67→0.55% |
| Specificity PCR, LCR and SDA for gonorrhoea, women, urine | 99.7% | 99.0% | 100% | Lower value: 2.3→3.7% Upper value: 2.7→4.0% | Lower value:0.16→0.09% Upper value:0.82→0.44% |
| Specificity PCR, LCR and SDA for gonorrhoea, women, urine and genital fluid | 99.7 | 99.0% | 100% | Lower value: 2.1→3.6% Upper value: 2.8→4.1% | Lower value: 0.11→0.05% Upper value:0.88→0.64% |
| Sensitivity of PCR/LCR and SDA for gonorrhoea, women, urine and genital fluid | 93.3%; 91.6%; 91.6% | 80% | 100% | Lower value: 2.8→4.3% Upper value: 2.4→3.6% | Lower value:0.67→0.41% Upper value:0.54→0.32% |
| Sensitivity of culture for gonorrhoea, women | 75.7% | 50% | 90% | Lower value: 2.7→4.1% Upper value: 2.6→3.8% | Not used |
| Other parameters in the gonorrhoea prevalence estimation | |||||
| Weight of UD case report data points, relative to national prevalence studies (the latter at 100%) | 25% | 1% | 50% | Not used | Lower value: 0.70→0.26% Upper value: 0.54→0.37% |
| Weight of neighbouring country prevalence surveys, relative to national prevalence studies | 10% | 0% | 30% | Lower value: 2.2→2.5% Upper value: 3.1→4.5% | Lower value: 0.43→0.31% Upper value:0.82→0.43% |
| Annual dilution factor in moving average estimation for gonorrhoea* | 20% | 0% | 40% | Lower value: 3.0% constant Upper value: 2.4→6.0% | Lower value: 0.49% constant Upper value: 0.67→0.37% |
*The annual dilution factor weights the contribution of each data point to the estimation for other years than the study year (earlier and later, within the period from first to last data point) downward, by a fixed proportion (default value: 20%) for each additional year away from the estimated year.
ANC, antenatal care; FP, family planning; LCR, ligase chain reaction; RPR, rapid plasma reagin; SDA, strand displacement amplification; TPHA, Treponema pallidum haemagglutination assay; UD, urethral discharge.