Table B.2 is separated into 10 smaller tables in this HTML version for accessibility reasons. It is presented as one table in the PDF version.

Transmission
Disease progression of undiagnosed individuals without treatment
Disease progression of HIV-infected individuals on treatment (detectable viral load)
Disease progression on treatment (undetectable viral load)
Commencement of treatment
Stopping treatment (detectable viral load)
Response to treatment (undetectable viral load)
Response to treatment (detectable viral load)
Mortality rates (detectable viral load)
Mortality rates (undetectable viral load)
References

Transmission

SymbolDescription Values References
HIVTransmission probability of HIV per injection with a contaminated syringe
0.6-0.8%
[155, 156], a

Disease progression of undiagnosed individuals without treatment

SymbolDescription Values References
1/τCD4>500Average time for undiagnosed (without ART) HIV-infected individuals to progress from CD4 count >500 to CD4 count 350-500
4.09 (3.79-4.42) years
[157], b
1/τ350<CD4<500Average time for undiagnosed (without ART) HIV-infected individuals to progress from CD4 count 350-500 to CD4 count 200-350
1.96 (1.81-2.13) years
[157], b
1/τ200<CD4<350Average time for undiagnosed (without ART) HIV-infected individuals to progress from CD4 count 200-350 to CD4 count <200
1.96 (1.81-2.13) years
[157], b

Disease progression of HIV-infected individuals on treatment (detectable viral load)

SymbolDescription Values References
1/ωDCD4>500Average time for HIV-infected individuals on ART with detectable viral load to progress from CD4 count >500 to CD4 count 350-500
10.99 (1.32-12.00) years
[158] , c
1/ωD350<CD4<500Average time for HIV-infected individuals on ART with detectable viral load to progress from CD4 count 350-500 to CD4 count 200-350
6.38 (0.48-8.00) years
[158] , c
1/ωD200<CD4<350Average time for HIV-infected individuals on ART with detectable viral load to progress from CD4 count 200-350 to CD4 count <200
8.88 (0.51-10.00) years
[158] , c
Top of page

Disease progression on treatment (undetectable viral load)

SymbolDescription Values References
1/ωUCD4<200Average time for HIV-infected individuals on ART with undetectable viral load to progress from CD4 count <200 to CD4 count 200-350
2.80 (2.33-3.58) years
[159] , d
1/ωU200<CD4<350Average time for HIV-infected individuals on ART with undetectable viral load to progress from CD4 count 200-350 to CD4 count 350-500
1.42 (0.90-3.42) years
[159] , d
1/ωU350<CD4<500Average time for HIV-infected individuals on ART with undetectable viral load to progress from CD4 count 350-500 to CD4 count >500
2.20 (1.07-7.28) years
[159] , d

Commencement of treatment

SymbolDescription Values References
η D/UCD4>500Proportion of individuals with CD4 count >500 that commence treatment for HIV each year
0.2
Experimental variable
ηD/U350<CD4<500Proportion of individuals with CD4 count 350-500 that commence treatment for HIV each year
0.5
Experimental variable
ηD/U200<CD4<350Proportion of individuals with CD4 count 200-350 that commence treatment for HIV each year
0.75-0.85
Experimental variable
η D/UCD4<200Proportion of individuals with CD4 count <200 that commence treatment for HIV each year
0.85-0.95
Experimental variable

Stopping treatment (detectable viral load)

SymbolDescription Values References
φSPercentage of individuals on ART who cease therapy each year
1-5%

Response to treatment (undetectable viral load)

SymbolDescription Values References
φPercentage of individuals on ART to experience viral rebound per year
3-6%
[160]

Response to treatment (detectable viral load)

SymbolDescription Values References
1/σ200<CD4<350Average time after treatment failure for individuals with CD4 count >200 to go on second line ART
6-18 months
Experimental variable
1/σCD4<200Average time for individuals on ART with CD4 count <200 to go on second-line ART
2-3 months
Experimental variable
Top of page

Mortality rates (detectable viral load)

SymbolDescription Values References
μDCD4>500HIV-related death rate for patients with CD4 count >500 cells per ÁL and detectable viral load
0.051% (0.035-0.068%)
[151]
μD350<CD4<500HIV-related death rate for patients with CD4 count 350-500 cells per ÁL and detectable viral load
0.128% (0.092-0.164%)
[151]
μD200<CD4<350HIV-related death rate per 100 person-years for patients with CD4 count 200-350 cells per ÁL and detectable viral load
1.0% (0.2-2.0)%
[151, 158]
μDCD4<200HIV-related death rate per 100 person-years for patients with CD4 count <200 cells per ÁL and detectable viral load
4.08 (0.30-7.86)%
[151, 158]

Mortality rates (undetectable viral load)

SymbolDescription Values References
μUCD4<200HIV-related death rate for patients with CD4 count <200 cells per ÁL and undetectable viral load
4.08 (0.30-7.86)%

(same as μDCD4<200)

Experimental variable
μU200<CD4<350HIV-related death rate for patients with CD4 count 200-350 cells per ÁL and undetectable viral load
1.0% (0.2-2.0)%

(same as μD200<CD4<350)

Experimental variable
μU350<CD4<500HIV-related death rate for patients with CD4 count 350-500 cells per ÁL and undetectable viral load
0.128% (0.092-0.164%)

(same as μD350<CD4<500)

Experimental variable
μUCD4>500HIV-related death rate for patients with CD4 count >500 cells per ÁL and undetectable viral load
0.051% (0.035-0.068%)

(same as μDCD4>500)

Experimental variable

References

Reference a

Numerous studies have estimated the transmission risk of HIV in an occupational setting due to needlestick injury.161-167 A model-based analysis evaluating population-level data in New Haven estimated the risk as ~0.7%.168 Few studies have directly estimated the probability of HIV transmission per injection by IDUs using a contaminated syringe. In a long-term cohort study among injecting drug users in Bangkok, Thailand, a probability of transmission per exposure with a contaminated syringe was estimated to be 0.6% (0.4-0.9%).156 A review and meta-analysis suggested that the probability of transmission following a needlestick exposure is 0.23% (0-0.46%) and the infectivity per intravenous drug injection had a median of 0.8% (ranging 0.63%-2.4%).155 Estimates from studies based on occupational exposure tend to have lower transmission risk than estimates of risk by intravenous drug injection. Based on the injecting drug studies, we assume that the probability of transmission per drug injection with a contaminated syringe ranges from 0.6 to 0.8%. Top of page

Reference b

A summary of the relation between HIV-1 RNA concentration and decline in CD4+ count from the prospective study by Mellors et al.157 is given below.
Plasma HIV-1 RNA concentration (copies/mL)Mean decrease in CD4 T cell count per year (cells/ÁL)
≤ 500
-36.3 (-30.4,-42.3)
501-3,000
-44.8 (-39.1,-50.5)
3,001-10,000
-55.2 (-50.7,-59.8)
10,001-30,000
-64.8 (-59.6,-70.0)
> 30,000
-76.5 (-70.5,-82.9)
With this data, and assuming that the average viral load is ~104.87 copies per mL for people without treatment, the CD4+ T cell count decreases by an average of 76.5 (70.5, 82.9) every year.

To progress through the >500 CD4 cell category, we assume that the average CD4 count is 800 cells/ÁL after the 2-month acute phase of HIV infection and then declines at the constant rate of 76.5 (70.5, 82.9) cells/ÁL each year. Then the average time to progress through this compartment is 2/12 + 300/(76.5 (70.5, 82.9)) years; that is 4.09 (3.79, 4.42) years.

To progress through the 350-500 and 200-350 CD4 cell categories, we assume an average loss of 150 CD4 cells. Then the average time to progress through this compartment is 150/(76.5 (70.5, 82.9)) years; that is 1.96 (1.81, 2.13) years.

Reference c

The relationship between the CD4+ T cell slope, and the patient HIV-1 RNA concentration, treatment information and demographic characteristics has been estimated by the PLATO Collaboration158 and led to the following regression coefficients.

Change (95% CI) in CD4 count slope (cells per ÁL per year)-multivariate analysis:
  • Current CD4 count per 100 cells per ÁL: 2.2 (-2.3, 6.6)

  • Current viral load, per log10 copies per mL: -25.0 (-29.0, -20.0)

  • Age, per ten years: -6.7 (-12.0, -1.2)

  • Infection via injecting drug use: -2.7 (-21.0, 16.0)

  • Number of drugs, per additional drug: 4.8 (0.22, 9.4)

  • Receiving ART (NNRTI): -23.0 (-35.0, -11.0)
Based on this published data158 the average change in CD4+ T cells for people failing therapy is estimated.

For people on treatment with detectable viral load, we assume their viral load is 103.5 with median age of 35 years.54

To progress through the >500 CD4 count category we assume an average loss of 250 CD4 count in this interval; then the average CD4 cell slope is 123 (82,164) - 6.7 (12.0, 1.2)*3.5 - 2.7 (21.0, -16.0)*1– 25 (29.0, 20.0)*3.5 + 2.2 (-2.3, 6.6)*10 + 4.8 (0.22, 9.4)*3 – 23 (-35.0, 11.0)*1. That is, an average CD4 slope of 22.75 (-69.84, 189.0) cells per ÁL per year. Therefore, the average time to progress through this compartment is 10.99 (1.32, 15#) years.
Top of page
To progress through the 350-500 CD4 count category, we assume an average loss of 75 CD4 count in this interval; then the average CD4 cell slope is 123 (82,164) - 6.7 (12.0, 1.2)*3.5 - 2.7 (21.0, -16.0)*1– 25 (29.0, 20.0)*3.5 + 2.2 (-2.3, 6.6)*5 + 4.8 (0.22, 9.4)*3 – 23 (-35.0, 11.0)*1. That is, an average CD4 slope of 11.75 (-58.34, 156) cells per ÁL per year. Therefore, the average time to progress through this compartment is 6.38 (0.48, 8.00#) years.

To progress through the 200-350 CD4 count category, we assume an average loss of 75 CD4 count in this interval; then the average CD4 cell slope is 123 (82,164) - 6.7 (12.0, 1.2)*3.5 - 2.7 (21.0, -16.0)*1– 25 (29.0, 20.0)*3.5 + 2.2 (-2.3, 6.6)*3.5 + 4.8 (0.22, 9.4)*3 – 23 (-35.0, 11.0)*1. That is, an average CD4 slope of 8.45 (-54.89, 146.1) cells per ÁL per year. Therefore, the average time to progress through this compartment is 8.88 (0.51, 10.00#) years. #: upper bound assumption.

Reference d

Below is a summary of data from159 for changes in CD4 count over time among people who are on effective cART.
CD4 count at initiation of cART (cells per ÁL)Time since starting cART (years)Current CD4 (cells per ÁL) means (95% CI)
≤200
<1
76 (53-99)
≤200
1-3
69 (63-76)
≤200
3-5
50 (36-69)
≤200
>5
32 (18-46)
201-350
<1
129 (91-166)
201-350
1-3
50 (25-74)
201-350
3-5
47 (24-69)
201-350
>5
23 (2-44)
>350
<1
90 (37-144)
>350
1-3
50 (18-82)
>350
3-5
17 (-17-51)
>350
>5
21 (-12-54)
We use this data to estimate the average time to progress through our CD4 categories whilst on effective cART. For people with undetectable viral load:
  • For CD4 count increases from 0 to 200 cells per ÁL, average increases of 76 (53-99) cells per ÁL can be expected during the first year and then 69 (63-76) cells per ÁL during the second and third years. Therefore, it can be expected to take 2.80 (2.33-3.58) years to progress through this category.

  • For CD4 count increases from 200 to 350 cells per ÁL, we have a 150 CD4 count increase. In this interval, the CD4 count increases by 129 (91-166) cells per ÁL during the first year and then 50 (25-74) CD4 count during the second year. Therefore, it can be expected to take 1.42 (0.9-3.42) years to progress through this category.

  • For CD4 count increases from 350 to 500 cells per ÁL, then we have a 150 CD4 count increase. In this interval, the CD4 count increases by 90 (37-144) cells per ÁL during the first year and then 50 (18-82) cells per ÁL during the second year. Therefore, it can be expected to take 2.20 (1.07-7.28) years to progress through this category.
EuroSIDA study169 investigated that the HCV serostatus does not influence CD4 recovery among patients on ART. It was found that there was no difference in CD4 gain among HIV/HCV coinfected and HIV monoinfected patients after starting ART. Therefore we assume the same recovery rate for HIV/HCV coinfected patient as HIV monoinfected patient. Top of page

Reference e

15.4/100 person years is the average rate of stopping one regime due to toxicity but the vast majority usually start another regime.170 Very few people who commence ART stop altogether (expert opinion). Therefore, we take the absolute rate of completely stopping therapy to range from 1-5% per year as an experimental variable.