The study assessed the cost-effectiveness of long-term double versus single praziquantel treatment in schistosomiasis/bilharzia-endemic areas. The authors concluded that treating children with two annual doses of praziquantel could be cost-effective in areas with high initial prevalence of Schistosoma mansoni or S. haematobium parasites. The quality of the study methods was adequate. Although a very limited costing analysis was performed, the authors’ conclusions appear to be valid given the scope of the study.

Cost-effectiveness analysis, cost-utility analysis

The objective of the study was to assess the cost-effectiveness of long-term double versus single praziquantel treatment interventions in a Schistosoma-endemic setting.

Three treatment strategies for schistosomiasis (Schistosoma mansoni or S. haematobium) were assessed in high-risk communities in Africa. The strategies were no treatment, single annual dosing of praziquantel, and double annual dosing (separated by two to eight weeks) of praziquantel.

Africa/community care.

Analytical approach:

A decision-tree Markov model was used to estimate the transition of patients between different parasite infection states and the effect of treatment on infection over a 55-year time horizon (treatment commenced at five years of age, with follow-up until 60 years). The authors reported that a societal perspective was adopted.

Effectiveness data:

Clinical and effectiveness data was from a systematic review conducted by the authors. PubMed, MEDLINE and EMBASE were searched for studies in English published after 1982. The search strategy was reported. Eligible studies were those in which both single and double praziquantel were given for Schistosoma mansoni or Schistosoma haematobium infection involving population-based or subpopulation-based drug treatment. Randomised controlled trials and observational studies were included in the review. As the proportion of total variation in pooled study estimates varied widely, meta-analytical methods were not used to calculate pooled estimates for cure rates. Instead, the range of observed outcomes was reported. The main estimate used in the model was cure rate (positive to negative conversion in parasite egg detection assays).

Monetary benefit and utility valuations:

The health states in the model were different levels of infection (uninfected, light, or heavy). Utility estimates for these states were from using the EQ-5D questionnaire, and were obtained from previously published studies on schistosomiasis-related disability.

Measure of benefit:

The measures of benefit were the number of infection years averted, egg years averted, and quality-adjusted life-years (QALYs) gained. Future QALYs were discounted at an annual rate of 3%.

Cost data:

The direct costs were treatment (number and price of pills needed at each age) and drug delivery. The cost of treatment was taken from a recent study with pricing for bulk purchased generic praziquantel. The cost of delivery was estimated from similar community health projects in African countries. All costs were reported in US $. As costs could be incurred over a long time period, future costs were discounted at an annual rate of 3%.

Analysis of uncertainty:

One-way sensitivity analyses were performed by varying all model parameters over their 95% confidence intervals or a plausible range of values. Results of the analysis were presented in a Tornado diagram.

The lifetime QALYs gained with no treatment was 26.96. For patients who received one annual dose of praziquantel, the QALYs gained were 27.46 for the treatment of Schistosoma mansoni infection and 27.47 for treatment of Schistosoma haematobium infection. For patients who received two annual doses of praziquantel, the QALYs gained were 27.52 for Schistosoma mansoni infection and 27.52 for treatment of Schistosoma haematobium infection.

The annual average cost of treatment for one annual dose of praziquantel was $23.01 and $46.03 for two annual doses of praziquantel.

Costs and outcomes were combined using an incremental cost-utility ratio (the additional cost per QALY gained). When compared with no treatment, the incremental cost per QALY gained for one annual dose of praziquantel was $47.90 for the treatment of Schistosoma mansoni infection and $45.58 for the treatment of Schistosoma haematobium infection. When compared with one annual dose of praziquantel, two annual doses of praziquantel were associated with an incremental cost-utility ratio of $291.07 per QALY gained for treatment of Schistosoma mansoni infection and $432.76 per QALY gained for treatment of Schistosoma haematobium infection.

The sensitivity analysis showed that the results were most sensitive to changes in the costs of medication delivery and price of praziquantel; the probability of infection had limited impact on the results.

The authors concluded that treating children with two annual doses of praziquantel could be a cost-effective treatment in areas with high initial prevalence of Schistosoma mansoni or Schistosoma haematobium.

Interventions:

The interventions under study were reported clearly and in detail.

Effectiveness/benefits:

Clinical and effectiveness data came from a systematic review conducted by the authors. The review methods were appropriately reported including the databases searched, search terms, inclusion and exclusion criteria, and the results. The review was limited to studies in English, but it was unlikely that major studies were excluded or omitted. The authors did not report the design of included studies.

The severity of infection may have been an appropriate health outcome for the disease, and utilities were obtained for these states from disease related studies, but there was no comparison of the source populations for the utilities and the infection levels.

Costs:

Although the authors reported that a societal perspective was adopted in the analysis, only the costs of drugs and their delivery were reported. There was no indication that other disease and complications costs (such as hospitalisations and doctor visits) were included in the analysis. The exclusion of other infection-related costs would probably underestimate the cost-effectiveness of praziquantel treatment. The opportunity costs of free volunteer time were incorporated. The price year was not reported, which could hamper future inflationary exercises. The time horizon and discount rate used were reported.

Analysis and results:

Costs and outcomes were combined using a decision-tree Markov model. Appropriate details of the model structure were provided, including a diagram. An exhaustive series of one-way sensitivity analyses were undertaken and clearly reported. However, a probabilistic sensitivity analysis would have been a better way to evaluate overall model uncertainty. As main limitation to their analysis the authors reported that the model included some simplifying assumptions that might limit the generalisability of the results.

Concluding remarks:

The quality of the study methods was adequate. Although a very limited costing analysis was performed, the authors’ conclusions appear to be valid given the scope of the study.

The study was supported by the Schistosomiasis Consortium for Operational Research and Evaluation (SCORE) through the University of Georgia Research Foundation, USA and a National Science Foundation Ecology of infectious Diseases Program grant, National Institutes of Health, USA.