In order to address the public health burden imposed by epilepsy and to stimulate expenditure into its appropriate management, there is a need to demonstrate that the interventions are not only effective and sustainable, but also affordable. Evidence for the cost-effectiveness of epilepsy treatment is currently lacking but has been identified as an important priority (Patchlatko & Beran, 1995). Phenobarbital (PB) is recommended by the World Health Organization (WHO) as a broad-spectrum first-line drug for partial and generalized tonic–clonic seizures (Kwan & Brodie, 2004; Chisholm, 2005). A demonstration project of epilepsy management at primary health level was carried out in rural China under the auspices of the WHO/International League against Epilepsy (ILAE)/International Bureau for Epilepsy (IBE) Global Campaign Against Epilepsy (Wang et al., 2006). It offered an opportunity to obtain data related to resource utilization and costs for PB treatment in primary care settings, which together with clinical outcomes can be used to inform decisions about cost-effectiveness and resource allocation in the context of low-income populations. The target population comprised patients older than 2 years with at least two convulsive seizures in the previous 12 months. Exclusion criteria were described elsewhere (Wang et al., 2006). All participants were asked to provide written informed consent (for children, consent was signed by their parents or guardians). Fifty patients and one hundred patients were randomly selected in rural Shanghai and Ningxia, respectively, with the representative of areas with different levels of economic development. The initial PB dose, 60 mg for adults and 15 mg for children, was given for a period of 2–4 weeks. Doses were then titrated according to clinical needs up to 180 mg for adults and 4–5 mg per kg for children (WHO, 2000). Patients were followed up by rural physicians in village clinics (Shanghai) or town hospitals (Ningxia) every 2 weeks for the first month and monthly thereafter for dose adjustments, adverse events (AEs) assessment, and checks of adherence to treatment, and to receive further supplies of PB. Seizure-free cases were defined as the patients who became seizure free during the last 6 months of follow-up. Improved cases were defined as the patients with ≥50% reduction in the seizure frequency compared to the baseline (WHO, 2000). The efficacy of PB treatment was determined in terms of the percentage of seizure free and improved cases at 1-year follow-up. We calculated 1-year costs of running the primary care level epilepsy demonstration project in rural Shanghai and Ningxia. The unit price of PB was 2.00 yuan (US$ 0.24, at 2004 prices) per hundred 30 mg tablets, (Phenobarbital Tablets, Shanghai Jinshan Pharma Ltd., Jinshan, Shanghai, China). Costs, including drug costs, personnel, training, and operating were calculated and compared between the two study sites. At the first visit, healthcare contacts, related medical expenses, and time during the 12 months before treatment were documented as the baseline data. The same information was then obtained for each subsequent 3-month period. We developed an economic burden questionnaire to establish the costs to patients associated with contact with (and fees paid for) other health-care services (hospital outpatient and inpatient care) at different levels of the health system (village health worker, town hospital doctor, neurologist, and hospital admission) as well as time costs associated with travel and waiting (Chisholm et al., 2000; Srinivasa et al., 2005). Continuous variables were expressed as mean and standard deviation (SD), whereas categorical variables were presented as frequencies (%). Comparisons of changes over time in the health care contacts, access time, and seizure frequency for epilepsy patients with baseline and 1-year follow-up were evaluated with the paired sample t-test or the Wilcoxon signed-ranks test, according to whether or not the data were normally distributed. All p-values and confidence intervals (CIs) were estimated in a two-tailed fashion. Differences were considered to be statistically significant at p < 0.05. The sex and age of people with epilepsy were similar in the two areas, but the education level and yearly income of the patients in rural Ningxia were lower than those in rural Shanghai (Table 1). The total annual cost was 6,267 yuan and 12,370 yuan in rural Shanghai and Ningxia, respectively. PB cost 17% (1,050 yuan) and 23% (2,800 yuan) of the total expenses of the project in rural Shanghai and Ningxia. In the year before the demonstration project, 16 (32%) and 50 (50%) patients in rural Shanghai and Ningxia, respectively, visited town hospital doctors. Only 14 (28%) and 5 (5%) patients, respectively, made outpatient visits to a neurologist. The hospital admission rate was 22% among sampled patients in rural Shanghai, considerably higher than that in rural Ningxia (2%). All the sampled patients in rural Shanghai were followed up by village health workers with an average of 13.56 visits. For town hospital doctor and neurologist visits, the average visits per year decreased by 2.24 and 1.72, respectively, because fewer patients were seen by town hospital doctors (11 cases) and neurologists (4 cases). As the patients in rural Ningxia were all followed up by town hospital doctors, the average visits to town hospital doctors increased from 5.76 to 13.58. The average visits to neurologist decreased by 0.07 because fewer patients (2 cases) visited neurologist during the 1-year follow-up time (Table 2). The mean costs of health care contact with town hospital doctors in rural Shanghai decreased by 103.22 yuan. Even after allowing for the additional costs of village health worker contacts, overall expenditure was still lower than that of the baseline. In rural Ningxia, epilepsy patients were followed up by physicians in town hospitals, thus the time costs to town hospital doctors significantly increased by 9.61 as compared to that of the baseline (p < 0.001). The health care costs to town hospital doctors, however, decreased by 12.36 as compared to that of the baseline. Costs related to neurologist contacts dropped in both study sites. In rural Shanghai and Ningxia, the total 1-year expenses per treated patient before the demonstration project were 1,494.30 yuan and 213.09 yuan, respectively, and these expenses decreased to 91.52 yuan and 45.90 yuan, respectively, during the demonstration project (p < 0.05). We found that 42 (84%) patients had been effectively treated and 23 (46%) patients became seizure free after one year of the PB treatment in rural Shanghai, whereas 73% patients have been effectively treated and 44% became seizure-free cases within the 100 Ningxia patients. Seizure frequency of the epilepsy patients decreased by about 60% with statistical significant in both study sites after 1 year of the demonstration project (Table 3). A small number of cost-effectiveness studies have been carried out in high-income countries (Heaney et al., 2002), while fewer studies in developing countries provide economic data related to epilepsy treatment. In India, the annual direct cost of epilepsy per patient was US$62–93, mainly using PB, phenytoin, carbamazepine, and sodium valproate (Thomas et al., 2001; Krishnan et al., 2004). The cost-effectiveness of first-line antiepileptic drugs (AED) treatments in the developing world was studied at population level as part of the WHO's choosing interventions that are cost-effective (CHOICE) project. In the Western Pacific region, to which China belongs, the annual patient cost of PB treatment per year was estimated at 93 international dollars (an international dollar has the same purchasing power as the U.S. dollar has in the United States) (Chisholm, 2005). Apart from this study, our study is the only economic evaluation study (cost-outcome study) of PB treatment conducted in Asian countries, reflecting the situation in rural China by using two study areas with different economic levels. Furthermore, and unlike other cost-effectiveness studies that have relied to a varying extent on assumptions, expert opinion, and modeling of underlying cost and effectiveness (Heaney & Begley, 2002), this prospective study has presented the cost-outcome of standard PB therapy by using the exact survey data obtained from the demonstration project. The efficacy of PB treatment in the study needs to be interpreted cautiously because the current study was not a controlled study design, meaning that the observed improvements could be at least partly accounted for by factors uncontrolled for (including the spontaneous rate of remission for epilepsy patients plus any placebo effect present). However, the placebo effect for first-line AED treatment is likely to be low, while the expected rate for spontaneous remission among untreated chronic epilepsy patients has been recently estimated to be as low as 5% (Chisholm, 2005). Therefore, it is unlikely that the observed substantial improvements in seizure frequency in both study sites occurred by chance. Our results demonstrate that the use of PB for the treatment of epilepsy has considerable appeal, especially in low-income regions, as there are not only clinical improvements, but also lower costs in the target population. More analysis considering cost-minimization, cost-effectiveness, and cost-benefit may be done in the future to strengthen the evidence for the generalization of PB therapy to resource-poor regions, which contain 60% of the world's population. We thank the village health workers and the town hospital doctors in rural Shanghai and Ningxia for their efforts in the patient follow-up in this study. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
No takes yet. Share an insight, caveat, or question.
Ding et al. (2008) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: