This study examined the clinical and economic impact of a single pill, versus two or three pills, as first-line antiretroviral therapy (ART), for people living with HIV. The authors concluded that the single pill was as effective as the two- and three-pill regimens of the same drugs, and if started early, before AIDS onset, it could save costs. The methods were valid and transparent, making the authors’ conclusions robust.

Cost-effectiveness analysis

This study examined the clinical and economic impact of a single pill, versus two or three pills, as first-line antiretroviral therapy (ART), for people living with HIV.

Four combination therapies were considered; two consisted of three pills, one was two pills, and one was one pill. The two three-pill regimens were tenofovir, emtricitabine, and efavirenz; and tenofovir, lamivudine, and efavirenz. The two-pill regimen was Truvada (tenofovir and emtricitabine) and efavirenz. The single pill was Atripla (tenofovir, emtricitabine, and efavirenz).

UK/secondary care.

Analytical approach:

The analysis was based on one study, with a one-year time horizon. The authors stated that they took the perspective of the public service.

Effectiveness data:

The clinical data on the four ART options were from a prospective database, held by the National Prospective Monitoring System (NPMS), which collected data from UK hospitals, from 1996 onwards. A total of 1,448 patients were included in the analysis; 1,122 were patients without AIDS, with 681 receiving one pill, 248 receiving two pills, 129 receiving three pills including lamivudine, and 64 receiving three pills including emtricitabine; and 326 were patients with AIDS, with 168 receiving one pill, 78 receiving two pills, 62 receiving lamivudine three pills, and 18 receiving emtricitabine three pills. The statistical approaches used to extract and extrapolate the data were reported. Missing data at each follow-up assessment (six months and one year) were imputed randomly. The rates of treatment failure, at six months and one year, were the key endpoints.

Monetary benefit and utility valuations:

Not considered.

Measure of benefit:

No summary benefit measure was used. The likelihood of first-line treatment failure was the main clinical outcome.

Cost data:

The economic analysis included the costs of the ART, other drugs, in-patient days, out-patient services, and other procedures. The quantities of resources were from the database of nine UK hospitals participating in the analysis. The method used to calculate the mean service use was developed within a European Union Action to analyse HIV health care resource use across Europe. The unit costs of health services were from the NPMS HIV Health-economics Collaboration (HHC) database. The costs for drugs were prices negotiated with pharmaceutical companies by the London HIV Consortium in 2008. All costs were in UK £ and the price year was 2008.

Analysis of uncertainty:

Conventional tests for parametric and nonparametric data were used to assess the significance of the comparison between groups.

The likelihood of first-line failure at one year was 19.5% with three pills (emtricitabine), 27.8% with three pills (lamivudine), 17.2% with two pills, and 16.0% with one pill. The hazard ratios, adjusted for gender, age, ethnic group, baseline CD4 cell count, baseline viral load, stage of HIV at the start of therapy, and year of starting first-line ART, were 0.98 with three pills (emtricitabine), 1.41 with three pills (lamivudine), 0.94 with two pills, and 1.00 with one pill. The differences between these hazard ratios were not statistically significant.

For patients without AIDS, the one-year total costs were £9,821 with one pill, £11,775 with two pills, £12,643 with three pills (emtricitabine), and £13,467 with three pills (lamivudine). For patients with AIDS, the costs were £19,933 with one pill, £20,261 with two pills, £17,710 with three pills (emtricitabine), and £25,811 with three pills (lamivudine). The total costs were significantly lower for one pill, than for two or three pills, for patients without AIDS.

The authors concluded that the single pill was as effective as the two- and three-pill regimens of the same drugs, and if started early, before AIDS onset, it could save costs.

Interventions:

The rationale for the selection of the comparators was clear, as one-, two-, and three-pill regimens of the same drugs, and a three-pill alternative regimen, were considered.

Effectiveness/benefits:

The clinical analysis relied on data from a prospective database, which should have been representative of the UK HIV epidemiology. The efficacy of the treatments was defined by their failure rates, and the calculation for these was clearly reported. Other potentially critical issues related to missing data were analysed and discussed. The authors used appropriate statistical methods to deal with differences in the baseline characteristics of the patients, but they acknowledged that some confounding might remain and could not be ruled out. A large sample of patients was used, but there were substantial differences in the number of patients in each group, with a relatively small number in the three-pill tenofovir, emtricitabine, and efavirenz group.

Costs:

The cost categories reflected the perspective of the third-party payer as stated by the authors. The quantities of resources were from a large database of nine hospitals, which were representative of the UK. The unit costs were from standard UK sources. The costs were not varied in the sensitivity analysis, but analyses were conducted to control for differences in the baseline characteristics of the patents. The price year was reported, allowing reflection exercises.

Analysis and results:

The total and incremental costs and benefits were clearly reported, but a ratio of the costs and benefits was not calculated. Appropriate sensitivity analyses were carried out to investigate uncertainty, mainly using statistical tests to refine the administrative data extracted from the HIV database. The authors acknowledged some limitations to their analysis, mainly relating to the design of the clinical study. These findings should be considered to be specific to the UK and might be difficult to transfer to other countries with different costs or clinical practices.

Concluding remarks:

The methods were valid and transparent, making the authors’ conclusions robust.

Supported by a grant from Gilead, manufacturer of the single pill, Atripla.