The estimated cost of switching from d4T to TDF in South Africa

Polly Clayden, HIV i-Base

The majority (70%) of patients in resource limited settings starting antiretroviral therapy (ART) do so with a d4T containing regimen. d4T is associated with high rates of toxicities including lactic acidosis, lipodystrophy and peripheral neuropathy and is responsible for the majority of drug switches.

Many countries are considering replacing d4T with tenofovir (TDF), which, although currently more expensive than d4T, is associated with fewer adverse events.

A poster from Ian Sanne and coworkers from the University of the Witwatersrand, Johannesburg, South Africa and Boston University, MA, US, reported findings from an analysis using a Markov model to estimate the cost of switching from d4T to TDF in South Africa’s first line regimen.

The model used existing prices of d4T and TDF and also estimated the incremental cost per quality-adjusted life year (QALY) gained from the switch. This analysis then determined the prices of TDF at which the switch would become highly cost-effective and budget-neutral.

The investigators modelled both the ARV and toxicity management costs of treating a hypothetical cohort of 1000 adult patients for two years with the current first-line regimen of d4T/3TC/EFV (d4T scenario) and compared them with corresponding cost estimates for a regimen of TDF/3TC/EFV (TDF scenario).

They used data from the patient database at a public hospital in Johannesburg to estimate rates of toxicities, associated drug switches and management for the d4T scenario. They estimated corresponding rates and resource use parameters for the TDF scenario from the literature.

In the model, events defined for the analysis attributed to d4T were: peripheral neuropathy, hyperlactatemia/lactic acidosis, lipodistrophy, and pancreatitis. Renal failure was the only toxicity attributed to TDF.

Resources used for estimates for toxicity management included hospital admissions, outpatient visits, laboratory tests, and drugs.

They used the current public sector price for d4T ($3.15/month) and the Clinton Foundation HIV/AIDS Initiative (CHAI) price for TDF ($12.42/month).

The investigators reported, in the d4T scenario, 50.5% of patients experienced a confirmed or suspected d4T-related toxicity and 16.2% were switched to a different drug (mainly AZT) by the end of two years. Hyperlactatemia (64%) and peripheral neuropathy (26%) accounted for the majority of d4T toxicities. In the TDF scenario, only 2.5% of patients experienced a TDF-related toxicity and were switched to AZT. The average toxicity management costs ranged from $45 for mild hyperlactatemia to $5506 for lactic acidosis.

They found, after two years, the total cost of the TDF scenario was 15% greater than that of the d4T scenario. The average cost increase per patient treated was $89/year. Savings on toxicity management offset 41% of the higher price of TDF compared to d4T.

The investigators also looked at the question: “What if d4T is responsible for a proportion of observed loss to follow up?”

They found, if 10-20% of observed loss to follow is a result of d4T-related toxicities, then TDF would be “very cost effective” at only slightly less than the modelled price of $17/patient/month.

The investigators noted that their analysis was limited by data quality, model constraints, and time frame. Also the assumptions used “were conservative and likely underestimated the costs of d4T-related toxicities”.

But from this model they concluded:

  • If TDF was priced at $17/patient/month, the reduced costs for managing d4T-related toxicities would offset about 20% of the higher price of TDF.
  • For the switch to TDF in first-line regimens to be cost-neutral for the South African government, TDF will need to be priced at approximately $6/patient/month.
  • For patient welfare (QALYs,), the switch will be highly cost-effective at a TDF price of $13/patient/month.
  • If 10-20% of observed loss to follow up is attributable to d4T-related toxicities, then switching to TDF is very cost effective even at currently available prices.


Sanne I, Long L, Fox M et al. The cost of switching from stavudine to tenofovir in first-line ARV regimens in South Africa. 15th CROI. February 2008. Boston, USA. Poster abstract 989.

Links to other websites are current at date of posting but not maintained.