Economic implications of three strategies for the control of taeniasis

Authors


Corresponding Author  Vedantam Rajshekhar, Department of Neurological Sciences, Christian Medical College, Vellore 632 004, India. Fax: +91 416 2232103; E-mail: rajshekhar@cmcvellore.ac.in

Summary

Objective  To evaluate the cost-effectiveness of three strategies for the control of taeniasis in a community, in terms of cost per case treated.

Methods  A study was conducted in South India to determine the prevalence of taeniasis by screening stool samples from 653 randomly chosen subjects, for coproantigens. The costs incurred in the project were used to estimate the cost per case screened and treated. A one-way sensitivity analysis was carried out for varying rates of taeniasis, different screening strategies and mass therapy. Further sensitivity analysis was carried out with different manpower and test costs.

Results  The rate of taeniasis as detected by ELISA for coproantigen was 3 per 1000 (2 of 653 samples). Our study showed that mass therapy without screening for taeniasis would be the most economical strategy in terms of cost per case treated if field workers are employed exclusively for either mass therapy or screening. For each strategy, costs per case treated are higher at low prevalence of taeniasis, with a sharp rise below 15%.

Conclusions  In places that are endemic for taeniasis and neurocysticercosis, mass therapy or screening for taeniasis should be considered. Screening by stool microscopy is not cost-effective in terms of cost per case of taeniasis treated owing to its low sensitivity. Although the cost per case of taeniasis treated is high at low prevalence of taeniasis for all options, incorporating mass therapy into existing mass drug distribution programmes might prove to be the most cost-effective control strategy.

Abstract

Objectif:  Evaluer la rentabilité de trois stratégies pour le contrôle du téniasis dans une communauté, en termes de coût par cas traité.

Méthodes:  Une étude a été menée dans le sud de l’Inde afin de déterminer la prévalence de téniasis par dépistage de coproantigènes à partir d’échantillons de selles provenant de 653 sujets choisis aléatoirement. Les frais engagés dans le projet ont été utilisés pour estimer le coût par cas dépisté et traité. Une analyse de sensibilité a été réalisée pour des taux variables de téniasis, différentes stratégies de dépistage et le traitement de masse. Une analyse de sensibilité supplémentaire a été effectuée avec différents coûts de la main-d’œuvre et des tests.

Résultats:  Le taux de téniasis détectés par ELISA pour coproantigènes était de 3 pour 1000 (2 échantillons sur 653). Notre étude a montré que la thérapie de masse sans dépistage de téniasis serait la stratégie la plus rentable en termes de coût par cas traité, si les agents de terrain sont employés exclusivement pour le traitement de masse ou de dépistage. Pour chaque stratégie, les coûts par cas traité sont plus élevés pour des prévalences faibles de téniasis, avec une forte hausse en dessous de 15%.

Conclusions:  Dans les régions endémiques pour le téniasis et la neurocysticercose, le traitement de masse ou le dépistage de téniasis devrait être envisagé. Le dépistage par examen microscopique des selles n’est pas rentable en termes de coût par cas de téniasis traité en raison de sa faible sensibilité. Bien que le coût par cas traité est élevéà faible prévalence de téniasis pour toutes les options, l’intégration d’un traitement de masse dans les programmes existants de distribution de masse de médicaments pourrait s’avérer être la stratégie de contrôle la plus rentable.

Abstract

Objetivo:  Evaluar la costo-efectividad de tres estrategias para el control de la teniasis en una comunidad, en términos de coste por caso tratado.

Métodos:  Se realizó un estudio en el sur de India para determinar la prevalencia de teniasis mediante la criba de muestras de heces de 653 sujetos elegidos al azar a los que se les realizó la detección de coproantígenos. Los costes en los que se incurrió durante el proyecto se utilizaron para estimar el coste por caso incluido en la criba y tratado. Se realizó un análisis de sensibilidad para tasas variables de teniasis, diferentes estrategias de cribaje y terapia masiva. Un análisis de sensibilidad más profundo se realizó incluyendo los diferentes costes de recursos humanos y de las pruebas.

Resultados:  La tasa de teniasis detectada mediante ELISA para el coproantígeno era de 3 por 1000 (2 de 653 muestras). Nuestro estudio ha demostrado que la terapia masiva sin cribaje para la teniasis sería la estrategia más económica, en términos de coste por caso tratado, si los trabajadores de campo se dedican exclusivamente para el tratamiento masivo o para el cribaje. Para cada estrategia, el coste por caso tratado es mayor en zonas con baja prevalencia de teniasis, con un incremento acusado por debajo del 15%.

Conclusiones:  En lugares que son endémicos para la teniasis y la neurocisticercosis, la terapia masiva o el cribaje para teniasis debería ser considerado. El cribaje mediante microscopía de heces no es costo efectivo en términos de coste por caso de teniasis tratado debido a su baja sensibilidad. Aunque el coste por caso de teniasis tratado es alto, para todas las opciones, en zonas con una baja prevalencia de teniasis, incorporar la terapia masiva en los programas existentes de distribución masiva de medicamentos podría ser la estrategia de control más costo-efectiva.

Introduction

Persons who harbour the adult Taenia solium worm are an essential link in the transmission of cysticercosis in a community. Thus, one of the strategies for the control of cysticercosis is to identify and treat taenia carriers in the community. One approach is screening for taeniasis using either the simpler but less sensitive stool microscopy (Young et al. 1979; Allan et al. 1993, 1996) or the highly sensitive ELISA for coproantigens (Wilson & Nakane 1978; Allan et al. 1996; Rodriguez-Canul et al. 1999), followed by targeted therapy for those who test positive. The second approach is universal treatment with niclosamide or praziquantel in areas where the prevalence of taeniasis is high (Allan et al. 1997; Sarti et al. 2000; Sarti & Rajshekhar 2003; Garcia et al. 2006).

Studies from rural communities in India have shown prevalence of taeniasis of 9.7% (Vora et al. 2008) and 18.6% (Prasad et al. 2007) by stool microscopic examination. These rates are higher than those reported from other developing countries such as Peru (Garcia et al. 2003), Mexico (Sarti et al. 2000), Guatemala (Allan et al. 1997) and Vietnam (Somers et al. 2006). A cross-sectional study conducted in Vellore district in South India showed a prevalence of active epilepsy of 3.04/1000 and a prevalence of neurocysticercosis of 1.02/1000 population, indicating that it is a major public health problem (Rajshekhar et al. 2006). In this area, 24.2% of the population consume pork and 63.2% live in pig-rearing localities. Pigs being reared in these areas and those being sold for consumption were sampled for cysticercus antibodies by enzyme-linked immunoelectrotransfer blot (EITB) assay (Tsang et al. 1989; Prabhakaran et al. 2004) and circulating cysticercus antigen by antigen detection ELISA (Dorny et al. 2004). The EITB results demonstrated infection (more than three bands positive) in 9.7% (11/112) and exposure to the parasite (1–3 bands positive) in 49.6% (56/112). Circulating antigens were found in 7.1% (8/112) of porcine sera (unpublished data). A research study to estimate the prevalence of taeniasis in this area forms the background for this cost analysis. The aim of this study was to estimate the cost for control of taeniasis using targeted therapy of taenia carriers detected with one of two screening techniques or mass therapy without screening.

Methods

Study population

The study was conducted in Kaniyambadi, a rural block in Tamil Nadu, South India, with an area of 184 km2, 82 villages and a population of 105 886 in 2008 (Census by department of Community Medicine, Christian Medical College, Vellore) and adjoining Vellore town with 300 000 people. Four clusters were selected from 14 clusters (Rajshekhar et al. 2006), which had either pig-rearing communities or freely roaming pigs. From these four clusters, 386 families were selected by random sampling from the database. In the selected families, there were 1305 persons between 2 and 60 years of age who were considered eligible for the study by the criteria of age. Of these, 653 consented and gave stool samples for testing.

Sample collection

Two health workers were involved in visiting the families and identifying eligible subjects. Participants were given stool boxes and disposables for stool collection. Picture models of Taenia segments were also shown to the subjects who were asked whether they had passed similar segments in their stools at any point of time. The stool samples were collected on the next day by health workers and transported to the laboratory on the same day. The process of recruitment and screening was completed in 6 months.

Coproantigen assay and stool microscopy

Antibodies to T. solium somatic antigens were raised in rabbits and purified over Protein A-Sepharose in the laboratory. An aliquot of purified IgG was coupled to horse radish peroxidase (Wilson & Nakane 1978). Antibody titres were optimized for capture of coproantigens with six normal stool samples spiked with 0–1 μg protein of T. solium somatic antigens.

The samples were assayed for coproantigens by a capture ELISA using rabbit antibodies to T. solium somatic antigens by the method of Allan et al. (sensitivity of 98%, specificity of 99.2%) (Allan et al. 1996; Rodriguez-Canul et al. 1999). A positive result for coproantigen ELISA was taken for a value of 1.6-fold greater than mean + 3SD of six negative control samples on each ELISA plate. The stool samples were also tested by microscopy of formalin ethyl acetate concentrates (sensitivity of 38–56%) (Young et al. 1979; Allan et al. 1993, 1996).

Treatment for taeniasis

Persons who were positive on the coproantigen assay were administered niclosamide 2 g (Yomesan™, Bayer) per os along with a laxative (Exelyte; monobasic sodium phosphate + sodium hydrogen phosphate). They were hospitalized for 24 h, for the collection of stools to confirm the diagnosis by examining the stools for Taenia segments.

Costing analysis

We considered three ways to reduce the burden of taeniasis and calculated the costs per case detected and treated for each. All costs are in US dollars for the year 2008 (1 US$ = 39.4 Indian Rupees).

Screening with coproantigen assay and targeted therapy.

The first strategy is screening of stool samples for coproantigens by ELISA and therapy for those who are positive according to the criteria described earlier. The costs were calculated using actual costs incurred during our study. As various studies related to cysticercosis were conducted at the same time, the time spent by the health workers on the collection of stool samples was estimated to be one-fourth of the total travel for the project, and therefore, the cost was one-fourth of the project’s total travel cost. The cost thus obtained was divided by the average cost of hiring a vehicle, to derive a figure for the estimated distance travelled for this study on taeniasis.

As our study was for research purposes, the stool-positive persons were hospitalized for confirmation of diagnosis after a stool purge. For a community wide screening programme, this step may be omitted and persons who are positive for coproantigens directly administered oral niclosamide. Therefore, in our cost analysis, we have not considered the additional cost of hospitalization for the above step, which was US $ 30 per stool-positive subject in our study.

Screening with stool microscopy and targeted therapy.

The second strategy is universal screening using only stool microscopy of formalin ethyl acetate concentrates followed by targeted therapy of stool-positive persons.

Mass therapy without any screening.

The third option would be mass treatment with oral niclosamide, a drug that acts on the intestinal parasite alone, avoiding the side effects possible because of action on the cysts in the brain. A study in Guatemala, where treatment coverage of 75% was achieved for mass treatment with niclosamide, showed significant reduction in taeniasis 10 months after treatment (Allan et al. 1997). The cost of this option for the 653 subjects in our study was calculated assuming that the time, travel and manpower costs would be half of the screening options (3 months), as an additional visit to collect stool samples would be avoided. Sensitivity analysis was also carried out to account for possible variation in rates of taeniasis and inputs for the control strategies (Drummond et al. 2005).

The study on taeniasis was approved by the Institutional Review Board and Ethics Committee of Christian Medical College, Vellore.

Results

None of the 653 subjects had noticed Taenia segments in their stools.

Coproantigen screening and stool microscopy

Two (0.3%) of 653 samples tested positive for coproantigens. Both subjects were treated for taeniasis as described above. However, Taenia segments were not recovered from their stool samples, which were also negative for Taenia ova by stool microscopy.

Costing analysis

Screening with coproantigen assay and targeted therapy.

The cost per person screened by stool testing for coproantigens was US $ 12, and the cost per case of taeniasis detected was US $ 4051 (Table 1).

Table 1.   Screening using coproantigen ELISA testing of stool samples and targeted taeniasis therapy using niclosamide
ItemCosts/unit* (US $)Number of unitsTotal (US $)
  1. 1 US $ = 39.4 Indian Rupees.

Field workers’ salaries339/month2 workers, 6 months4063
Materials
 Pens0.452
 Stickers0.013100013
 Soap bottle1.311.3
 Stool boxes0.038100038
 Plastic sheets0.012100012
 Wooden spatula0.01810008
 Picture models for survey0.542
 Coproantigen ELISA tests56533480
 Cost of treatment with niclosamide, (Yomesan™, Bayer)52 subjects10
Transport0.2/km2329 km473
Total cost  8102
Number tested  653
Number who were stool positive  2
Cost/person screened  12
Cost/case taeniasis treated  4051

Screening with stool microscopy and targeted therapy.

The cost per person screened was US $ 10.8 (Table 2) assuming all costs were the same as for option 1 except for the cost of the stool microscopy. As there were no stool-positive cases by microscopic examination in our study, the cost per case of taeniasis treated was not calculated.

Table 2.   Screening using stool microscopy and targeted therapy for stool-positive subjects
ItemCosts/unit (US $)Number of unitsTotal (US $)
Stool microscopy test3.86532486
Other costs as in strategy 1 (excluding drug costs)  4612
Number who were stool positive 0 
Total cost  7098
Cost/person screened  10.8

Mass therapy without screening.

This option may be more cost-effective per case of taeniasis treated than the others (Table 3), but the acceptability, feasibility and actual costs in India would have to be explored.

Table 3.   Mass treatment with niclosamide without a screening programme
ItemCosts/unit (US $)Number of unitsTotal (US $)
2 g niclosamide (Yomesan™, Bayer)5653 persons3315
Manpower 2 workers, 3 months2031
Travel  236
Total Cost  5582
Number of persons treated 653 
Cost/person treated  8.5
Number of cases of taeniasis who would be treated  2
Cost/case of taeniasis treated  2791

Costs at different prevalence rates and input costs

If the results of prevalence of taeniasis from the other Indian studies (Prasad et al. 2007; Vora et al. 2008) are applied after adjusting for the sensitivity of stool microscopy (38%) and coproantigen ELISA (98%) (Allan et al. 1996), the cost per case detected falls with rising prevalence and is lowest for a mass treatment programme (Table 4). This was confirmed by a sensitivity analysis with prevalence of taeniasis varying from 1% to 100% and assuming a lower monthly salary of US $ 254 for field workers exclusively involved in the control strategies (data not shown).

Table 4.   Sensitivity analysis of cost per case of taeniasis treated at various values of prevalence of taeniasis, for different control strategies
Prevalence of taeniasis from various studies and tool usedOptions for screening using stool samplesMass therapy with niclosamide without screening
Coproantigen ELISA and targeted therapy*Microscopy and targeted therapy
  1. *Assuming coproantigen ELISA (98% sensitivity) detects 2.6 times the number of cases as microscopy (Allan et al. 1996).

If prevalence = 0.3% (this study), Cases = 2 (Coproantigen ELISA)US $ 4051No cases detectedUS $ 2791 (2 cases treated)
If prevalence = 9.7% (Vora et al. 2008), Cases = 63 (stool microscopy)US $ 54 (assuming 164 cases detected)US $ 118US $ 34 (assuming 167 cases treated)
If prevalence = 18.6% (Prasad et al. 2007), Cases = 121 (stool microscopy)US $ 31 (assuming 315 cases detected)US $ 64US $ 18 (assuming 321 cases treated)

If the activities performed by the field worker are incorporated into the routine rural primary care activity, the cost of manpower can be reduced further and only travel costs included in the model. Assuming each worker would spend 2 h a day for 6 months, for either screening or mass therapy, a salary of US $ 63 was taken as a lower estimate of manpower costs. A two-way sensitivity analysis was carried out for each control strategy varying the monthly salary of field workers as US $ 254, US $ 63 or nil salary at various rates of taeniasis (Figures 1 and 2). The cost per case treated of screening by microscopy followed by targeted therapy was higher than mass therapy and screening by coproantigen ELISA at all rates of taeniasis. Cost analysis was also performed varying the cost of stool microscopy to US $ 1.9/test, which is half the rate charged at the tertiary centre where our laboratory analysis was carried out.

Figure 1.

 Cost per case of taeniasis treated at various values of prevalence of taeniasis and manpower costs, for screening by stool microscopy.

Figure 2.

 Cost per case of taeniasis treated at various values of prevalence of taeniasis and manpower costs, for mass treatment.

Even with lower costs of stool microscopy and manpower, the cost of screening using microscopy would be higher per case detected than mass therapy or screening for stool coproantigens (Figure 3). Mass therapy would be the most economical option for all rates of taeniasis in terms of cost per case treated. However, sensitivity analysis showed that if the sensitivity of stool microscopy is above 68%, cost per case treated for screening by microscopy would be the most economical option, for rates of taeniasis below 20% (data not shown).

Figure 3.

 Sensitivity analysis of various control strategies at different rates of taeniasis and costs of microscopy, assuming field workers’ monthly salary as US $ 63.

The monetary benefit of control of taeniasis was also estimated with the assumption that a patient with epilepsy because of neurocysticercosis would incur a cost of US $ 355 per year, based on a previous study on the economic burden of epilepsy (Thomas et al. 2001), for 3 years. It was also assumed that a third of epilepsy is because of neurocysticercosis (Rajshekhar et al. 2006). Sensitivity analysis was carried out for various rates of taeniasis and neurocysticercosis calculated based on our study and another study carried out in a community with high rates of these conditions (18.6% taeniasis and 6.6% epilepsy) (Prasad et al. 2007), assuming a linear relationship between taeniasis and neurocysticercosis (Garcia et al. 2003). It was found that at a prevalence of neurocysticercosis above 5.5/1000 (estimated to correspond to a prevalence of taeniasis of 5%), monetary benefits gained would exceed the cost of mass chemotherapy at the lower limits of manpower costs assumed in our analysis. The monetary benefits would exceed costs of screening by coproantigen ELISA if the prevalence of neurocysticercosis is above 7.7/1000 (corresponding to a taeniasis rate of 7%). As microscopy would not detect many cases of taeniasis because of low sensitivity, we could not estimate the number of cases of neurocysticercosis prevented and monetary benefits gained.

Discussion

Reducing the burden of taeniasis is an important method of controlling the problem of neurocysticercosis. The burden of taeniasis varies widely in India as seen from our study as well as other Indian studies. Although the prevalence of taeniasis in this study was low, the burden of epilepsy and neurocysticercosis was high (Rajshekhar et al. 2006). This fact may justify measures to control taeniasis in such areas, albeit at a higher cost per case treated than in places with higher rates of taeniasis.

Both the options of mass therapy or screening for taeniasis would have to be conducted repeatedly to eradicate cysticercosis. As the coverage of either mass therapy or screening is unlikely to be 100%, more than 11 rounds of intervention may be required at a frequency of 90 days (Gonzalez et al. 2002).

Testing stool samples for coproantigens by ELISA would be an ideal method of screening for taeniasis as it has a high sensitivity, but as seen from our study does not seem to be an economically feasible option for adoption in other parts of the country.

Mass therapy for taeniasis

If a programme similar to our study is to be replicated using field workers exclusively for either mass therapy or collection of samples for screening, mass therapy would be more economical in terms of cost per case treated. As niclosamide has a good safety profile, it could possibly be given as regular mass treatment along the lines of the programme for the distribution of drugs against filariasis. However, mass therapy would be justifiable only after it has been proved through surveys that the region is endemic for cysticercosis/taeniasis.

The other option for taenicidal therapy would be praziquantel instead of niclosamide, as it is cheaper and more widely available. However, there is a possibility of occurrence of neurological symptoms following treatment in persons with occult neurocysticercosis (Flisser et al. 1993; Sarti et al. 2000).

Screening by stool microscopy and targeted therapy

Although screening by stool microscopy would be easier to implement in primary care and cheaper than screening by coproantigen ELISA or mass therapy, it is not cost-effective per case of taeniasis treated.

Even with lower manpower costs and stool microscopy charges than required for our study, screening using stool microscopy would not be cost-effective per case of taeniasis treated. This is because of the low sensitivity (38–56%) of the test, as fewer than half of the tapeworm carriers would be detected (Allan et al. 1993, 1996). However, improving the sensitivity to above 68% may make microscopy a cost-effective screening tool. The other drawback of screening using stool samples is the low response rate, which if improved would also decrease costs per case treated.

Mass therapy for taeniasis along with the filariasis control programme

Lymphatic filariasis is endemic in most of the states in India (Sabesan et al. 2000), and the government has a programme of regular mass administration of albendazole and diethylcarbamazine as per the WHOs recommended strategy for control of lymphatic filariasis (WHO 2001). Incorporation of mass therapy for taeniasis along with this existing programme may further reduce costs, by avoiding extra costs for travel and manpower. However, research would be required on the safety of giving niclosamide along with albendazole in a mass programme.

Limitations

If the response rate for obtaining stool samples was higher, the cost per case treated would be lower for all the strategies. Mass and individual health education may be used to improve the response rate in a screening programme rather than increasing the number of field visits.

As the burden of neurocysticercosis prevented or lives saved by treating Taenia carriers detected by stool microscopy was not known, monetary benefits gained by this method of screening could not be estimated. Thus, a full economic evaluation comparing costs and benefits of all control options could not be estimated. As our study was a research study on estimating the prevalence of taeniasis, the costs involved are probably higher than would be expected in programme settings. We have tried to adjust for the differences by sensitivity analysis, varying the costs of manpower and stool microscopy.

There is also a possibility that the persons who were positive for coproantigens in our study may have been falsely positive, as we did not recover Taenia segments from the stool samples after treatment or detect Taenia ova by microscopy. However, the probability of this is low as the specificity of the test is 99.2% (Allan et al. 1996).

Health education as part of screening for taeniasis

Another alternative that was not explored in this study is screening using interview technique for detecting those who have a history of having passed tapeworm segments in the stool (Sarti & Rajshekhar 2003). Although the interview method alone has a reported sensitivity varying from 2% to 60% (Sarti et al. 1994; Allan et al. 1996), it might be improved by intensive education on taeniasis. The education should include suggestions on safe disposal of stools for tapeworm carriers, sanitation, personal hygiene, safe pig-rearing and pork consumption practices. This strategy may prove to be the most cost-effective method to bring down the carrier rate for taeniasis in India and thereby decrease the burden of neurocysticercosis.

Acknowledgement

This study was funded by a grant from the Flemish Interuniversity Council, Belgium.

Ancillary