Return on investment in needle and syringe programs in Australia: report

5.3 Methodology for estimating the QALYs gained from the prevention of HIV and HCV infections

Page last updated: 2002

In order to calculate the QALYs gained it is necessary to estimate the duration of time individuals will experience particular health states and the 'utility' values for those states. The durations of time spent in particular health states for someone with HIV or HCV has been estimated from the epidemiological literature and Australian data. These have been incorporated into the calculation of the number of cases prevented by stage of disease and the projections of survivors beyond 2000, as outlined in Section 3, and presented in Table 3.4.5 and Table 3.5.5 in Appendix C.

Health state preference values for the disease stages are not readily available. One approach would be to undertake a free-standing study to estimate the values by standard gamble or time trade-off (Drummond, et al, 1997). However, this approach was considered beyond the scope of the current project.

Rather, it was decided to use values for health states existing in the literature. For example, Tengs et al (2000) report more than 1,000 health state values, many of which relate to HIV or HCV. In addition, there have been some papers published specifically relating to HIV (Holtgrave and Pinkerton, 1997) and HCV (Bennett, et al, 1997)

5.3.1 QALY values for HIV
5.3.2 QALY values for HCV
5.3.3 Additional elements of the QALY calculations

5.3.1 QALY values for HIV

The quality of life estimates for HIV have been based on a review updating costs of HIV illness and quality of life estimates since the introduction of combination antiretroviral therapy (Holtgrave & Pinkerton 1997). Their estimates were based on an extensive review of studies, in which HIV-infected patients formed the study population of almost all identified studies. There are some differences in the disease phases they used and those we have employed. We have divided the undiagnosed HIV phase into two, one with "early HIV disease" and one with "progressive HIV disease". In Holtgrave & Pinkerton (1997) all people with undiagnosed HIV were given a quality of life rating of 0.94. We have assumed that people with "progressive HIV disease" who remain undiagnosed would in general have a slightly lower quality of life rating than those with earlier HIV disease. We have assumed that all people with AIDS are diagnosed.

Definitions for disease states and associated quality of life adjustments for HIV are presented in Table 5.3.1:
Top of page

Table 5.3.1 Quality of life values by disease stage of HIV

Disease StageDescriptionQALY Value
Early HIV Disease – undiagnosedHIV infection with CD4 count above 500/mm3, unaware of HIV serostatus
0.94
Early HIV Disease – diagnosedHIV infection with CD4 count above 500/mm3, aware of HIV serostatus and no antiretroviral therapy
0.87
Progressive HIV Disease – undiagnosedHIV infection with CD4 count below 500/mm3, unaware of HIV serostatus
0.90
Progressive HIV disease – diagnosedHIV infection with CD4 count nadir below 500/mm3 and commenced on antiretroviral therapy
0.76
AIDSAIDS as defined by clinical condition
0.62

5.3.2 QALY values for HCV

Quality of life adjustments for HCV were partly based on previous published estimates from a panel of hepatologists (Bennett et al. 1997). However, recent evidence from quality of life assessments among patient assessments was used to adjust the ratings provided by Bennett et al (1997). For example, studies indicate no significant difference in quality of life based on either degree of hepatic inflammation (as measured by ALT/AST) or extent of hepatic fibrosis (Bonkovsky et al. 1999). Therefore, we have used the same quality of life adjustment for diagnosed mild and moderate chronic hepatitis. Undiagnosed categories have higher quality of life estimates for two reasons. Firstly, development of symptomatic disease may often be a reason for HCV testing. Secondly, recent evidence suggests that quality of life impairment increases following diagnosis of hepatitis C (Rodger et al. 1999). We have combined the quality of life adjustments from Bennett et al (1997) for ascities (0.35), variceal haemorrhage (0.28), and hepatic encephalopathy (0.30), to produce a category for liver failure (0.32). We have assumed that all people with liver failure and HCC are aware of their HCV status. For the 25% of HCV infections that do not progress to chronic hepatitis, we have allocated a quality of life value of 1.0. This is a conservative estimate, as many people with hepatitis C who have not progressed to chronic infection are unaware of their non-viraemic status and may suffer significant impairment in quality of life related to psychosocial mechanisms.

Definitions for disease states and associated quality of life adjustments for HCV are presented in Table 5.3.2:

Table 5.3.2 Quality of life values by disease stage of HCV

Disease StageDescriptionQoL Value
HCV antibody positive – non-chronic hepatitis CHCV infected but does not progress to chronic hepatitis
1.00
Mild chronic hepatitis – undiagnosedChronic hepatitis C, unaware of HCV status, with stage 0-1 (nominimal) hepatic fibrosis
0.94
Mild chronic hepatitis – diagnosedChronic hepatitis C, aware of HCV status, with stage 0-1 (nominimal) hepatic fibrosis
0.82
Moderate chronic hepatitis – undiagnosedChronic hepatitis C, unaware of HCV status, with stage 2-3 (moderate-severe) hepatic fibrosis
0.94
Moderate chronic hepatitis – diagnosedChronic hepatitis C, aware of HCV status, with stage 2-3 (moderate-severe) hepatic fibrosis
0.82
Compensated cirrhosis – undiagnosedChronic hepatitis C, unaware of HCV status, with associated cirrhosis but no evidence of liver failure or hepatocellular carcinoma (HCC)
0.84
Compensated cirrhosis – diagnosedChronic hepatitis C, aware of HCV status, with associated cirrhosis but no evidence of liver failure or hepatocellular carcinoma (HCC)
0.74
Liver failureChronic hepatitis C associated cirrhosis that has progressed to decompensation
0.32
Hepatocellular Carcinoma (HCC)Chronic hepatitis C associated cirrhosis that has progressed to HCC
0.10
Top of page

5.3.3 Additional elements of the QALY calculations

Although somewhat more controversial than discounting costs, it is conventional to discount QALYs by the same rate. This has the effect of slightly reducing the estimate of the total QALYs gained from NSPs, as many of the QALYs are gained in the future.