Economic costs analysis of uncomplicated malaria case management in the Peruvian Amazon

The good economic performance together with targeted social spending and enhanced anti-poverty programs led Peru to have significant success in improving health outcomes during the Countdown to the 2015 Millennium Development Goals (MDGs) [1]. Currently, the 2030 Sustainable Development Goals (SDGs) “to ensure healthy lives and promote wellbeing for all at all ages” [2] challenge the health sector, not least because Peruvian health expenditure are among the lowest in the Americas (about 5% of Gross Domestic Product-GDP) [3], substantial health inequities persist [4], and communicable diseases such as malaria resurge [5, 6].

Malaria remains an important public health problem in the country despite several decades of intense control efforts [7]. Peru had the third highest increase in malaria incidence (after Venezuela and Nicaragua) since 2010 in the Americas, [8], having reported about 72% more malaria cases in 2017 (54,309 cases) compared with 2010 (31,545 cases) [9]. The Amazon Region, mainly the department of Loreto, is commonly affected by malaria due to both Plasmodium vivax and Plasmodium falciparum (P. vivax/P. falciparum ratio: 4/1), accounting for more than 95% of Peru’s malaria cases in 2017 (52,280 cases) [9]. The complex association between poverty and malaria is well known, which likely operates in both directions in the poorest districts of Loreto: poor households are more exposed to infectious mosquitoes due to occupational hazards (i.e. subsistence farming and hunting) and less able to afford prevention, and the higher burden of malaria may push these same households deeper into poverty (due to high productivity losses following multiple episodes throughout a year) [10, 11]. Indeed, a former economic evaluation conducted in 1998 from the perspective of society found that malaria costs were mainly borne by the families [11].

In the last two decades, the Peruvian Ministry of Health (MoH) has implemented strategies to improve the population’s access to healthcare, especially for people not covered by the contributory social health insurance system (EsSalud), the Armed Forces (FFAA), National Police (PNP), or the private sector [12]. Thus, the Integrated Health Insurance scheme (IHIS), launched in 2009 by MoH, allowed for the expansion of former benefit packages beyond maternal and child health, and the increase of healthcare coverage to the poor and vulnerable populations of all ages [13‐15]. Currently, the IHIS and EsSALUD cover about 65% and 17% of the total population in Loreto, respectively [16]; however, in poor Amazonian districts that consist mostly of rural communities, the IHIS is often the only possibility to provide health services to the population. The IHIS has reinforced the main pillar interventions for malaria control in Loreto, such as malaria surveillance and case identification through passive case detection (PCD, i.e. malaria detection at health facility or by community health workers (CHW) using standard diagnosis among symptomatic individuals seeking healthcare) or active case detection (ACD, i.e. malaria detection in the community through household visits) [17, 18], and the effective, free, and timely management of confirmed malaria infections. In the past, those interventions were mainly supported by funding specifically assigned to the National Malaria Control Program (NMCP) at MoH [11].

From an economic perspective, it was described that the healthcare-seeking behaviours among individuals with malaria-like symptoms can significantly affect the direct and indirect costs associated with malaria detection and management [19]. In areas like Loreto, early and appropriate healthcare-seeking behaviours can save direct costs by halting transmission from confirmed malaria infections, and by avoiding additional costs associated with complications of such infections [11, 20]. Conversely, delayed behaviours, self-medication, and seeking healthcare from traditional healers can increase the latter direct medical costs, impacting household budgets due to the lost wages and transportation costs during the care-seeking process and the illness period [20]. This study aimed to estimate the economic costs associated with the PCD and management of uncomplicated malaria episodes in riverine communities in the Peruvian Amazon from both the provider’s and patient’s perspectives, and explore whether healthcare-seeking behaviours significantly influence those costs.

This research provides total costs and average costs per episode associated with the detection and management of uncomplicated malaria in Mazan, an exemplar district in the Peruvian Amazon where poverty, limited accessibility from communities to health facilities, and healthcare-seeking behaviour challenge malaria control efforts. Estimated total costs for the 1344 uncomplicated malaria episodes registered in Mazan in 2017 were US$ 217,006 with most of these costs (78%) corresponding to indirect costs incurred by patients. Average costs per episode were higher for P. falciparum (US$ 180) than for P. vivax (US$ 156), mainly due to differences in costs of the anti-malarial treatment and in time losses due to illness. Indeed, the length of time lost due to illness in a patient, found to be associated with delayed confirmatory malaria diagnosis, was the parameter that caused most uncertainty in the average costs per malaria episode for both species.

As in Colombia [38], Brazil [39] and other endemic areas outside the Americas [19, 40, 41], uncomplicated malaria episodes in Mazan imposes significant costs on households, pushing poor families further into poverty. In several Amazonian communities with high risk of malaria transmission, the situation can become catastrophic due to the occurrence of repeated episodes in the same individuals and/or the same households every year [11, 38, 39, 42, 43]. In contrast to the low costs of transport and medications paid out of pocket by families, indirect costs of malaria episodes were the key determinant of the disease’s overall costs. This is because adults have to interrupt their normal activities to care for ill family members, or the disease directly afflicts the economically active population [11, 40, 44].

The indirect costs of a P. vivax episode is markedly higher in Mazan in 2017 (US$ 260.45 with GDP; US$ 122.56 with minimum wage) than in Afghanistan, Ethiopia, Indonesia and Vietnam in 2016 [45] and Papua New Guinea in 2013 [45]. This is partially due to Peru having a higher GDP per capita, which was used to estimate the lost wages of patients and their carers/companions. Specifically, in Papua, Indonesia, malaria associated costs may have been underestimated since they were calculated based on data collected in localities with high access to health facilities by road [40]. In the study area, the limited access to health facilities would also have increased the time lost due to illness with corresponding uncertainty for economically active patients and carers/companions during the period in which healthcare is sought for diagnosis, treatment and follow-up. As a result, the costs of case management of one P. vivax episode in Mazan (US$ 313.50 based on GDP per capita, US$ 155.77 based on minimum wages) were higher than the estimates in Papua, Indonesia (US$ 44.50 based on mean wage) [40].

The association between longer time losses in a malaria patient and delayed malaria diagnosis could reflect a more compromised clinical case (without the need for hospitalization) that requires more recovery time. Indeed, several studies have shown an increase in malaria severity with a delay in malaria case management [20, 46, 47], thus the WHO’s recommendation of insuring access to early diagnosis, and prompt, effective treatment within 24-48 h of the onset of malaria symptoms [30]. The reduced group of participants (37.1%) with malaria diagnosis within 48 h of onset of symptoms suggests that the delay in the malaria diagnosis remains a pitfall of the malaria case management in the rural and widely dispersed population of the Peruvian Amazon, as previously found by a longitudinal study (2001–2003) in riverine and road-associated rural populations of peri-Iquitos districts [48, 49].

It was not surprising to find that the availability of RDTs in the first place to which enrolled participants sought for healthcare (i.e. CHW or health posts) reduced the time from the onset of symptoms to malaria diagnosis. The impact of incorporating the use of RDTs to get timely diagnosis and appropriate treatment has already been demonstrated in rural Amazonian communities [48, 49]. Where RDTs were not available in the study communities, unreliable or irregular fluvial transport and limited access to the health centre in MT may have delayed any patient reference and/or sending of blood smear samples for malaria confirmation [50]. However the findings show that even with an easy access to malaria diagnosis, an early malaria case management of participants was not guaranteed [48]. The time to diagnosis of three or more days in several participants who went directly to health facility or CHW with RDTs available, suggest a delay in seeking care. The community’s knowledge, misperceptions (e.g. perceiving malaria as a common, uneventful, and a mild and self-limiting disease), the lack of household involvement in community control of malaria, and initial self-treatment (e.g. use of antipyretics and refreshing herbs at home) reported in the KAP survey may be associated with such delay to seek care [51‐53].

Several studies have highlighted the need for economic research on the household costs of illness, household responses, and their implications for poverty [44]. The lack of knowledge about the economic burden of illness for households and/or the poor understanding about the effects on the quality of life and well-being in the population may lead health stakeholders to plan and deliver health services with the main goals of increasing coverage and reducing disease burden. This decision-making process may be more obvious in the case of services that are provided free-of-charge like the detection, diagnosis and treatment of malaria provided by the Peruvian MoH in the entire country [54, 55]. Increased access to prevention measures (e.g. delivery of long lasting insecticide treated bed nets) and malaria behaviour change communication (BCC) strategies would remain as the best options to avoid indirect and direct household costs due to malaria.

Similar to other studies conducted in Peru [11, 49], the provider in Mazan had to use more resources for delivering diagnosis and treatment services to an individual with a P. falciparum episode than one with a P. vivax episode. This was also observed in Papua, Indonesia, where the cost differences between both species were associated with additional consultations, transportation to the health facility and lost wages per episode [40]. The difference in costs of first-line treatment regimens between species (US$ 8.76 for P. falciparum and US$ 1.32 for P. vivax) explains well the higher provider costs of a P. falciparum episode. Delays in adjusting the daily dosage of one of the drugs (MQ) that make up the anti-malarial combined regimen for P. falciparum (MQ + AS + PQ) according to national policies have made difficult and expensive the procurement of separate presentations of these drugs because low volumes are less interesting for pharmaceutical companies, ceteris paribus. These delays have also prevented the purchase of cheaper drug presentations (AS-MQ fixed dose combinations instead of separate drug blisters) [56], which are used in other South-American countries such as Bolivia, Brazil and Venezuela [57, 58]. The uncertainty in the unit cost of the anti-malarial treatment for P. falciparum likely depends on these procurement issues.

This study has a number of limitations. First, budget constraints did not allow to include all the 70 communities of Mazan and to visit households more than three times to maximize participation in the socio-economic survey; however, the chosen communities (accounting for 80% of malaria incidence) and the cost parameters obtained from the survey likely reflect the case-management´s situation in the district. Second, verbal report of time losses due to illness and costs may affect the accuracy of the cost estimates from the patient perspective. This possibility was reduced with a short recall. Third, productivity changes may occur because of absence from work (absenteeism) or because of reduced productivity while at work (presenteeism). This information may be derived from existing data sources, such as registrations from occupational health service companies. Nonetheless, such registrations are typically available only for salaried workers and not for self-employed workers, such as subsistence farmers [59]. Fourth, from a patient perspective, indirect costs for children were considered zero as they should not be economically active. However, children represent a risk group and as 41% of the population were children, the methods for valuing the productivity costs could lead to biased estimates; this could be due to educational losses. Fifth, malaria infection can cause macroeconomic costs, which cannot be assessed in this study at the household level. For example, malaria can affect foreign direct investments, international trade and tourism. Sixth, mental stress and social costs of families with sick members were not included, which are in general very difficult to evaluate over a short period of time (and are typically not monetised in health economic evaluations). Finally, as this study was focused on uncomplicated malaria, costs may have been underestimated due to the exclusion of the cost of primaquine-induced haemolysis and the cost of adverse effects. Primaquine-induced haemolysis was not included in any of the estimates due to uncertainty related to its frequency and severity. In countries that are prescribing primaquine without testing for glucose-6-phosphate dehydrogenase (G6PD) deficiency, as in Peru, it is possible that haemolysis already occurs, however in previous studies, Peru has reported low (< 2%) G6PD deficiency [60].

Access to prompt diagnosis and treatment is one of the principal strategies for malaria control and elimination. This includes two components, one dependent upon the patient to seek care when sick and the other being the availability of accessible diagnosis and treatment facilities. Although malaria case management is provided free-of-charge by the MoH in Peru, this study confirmed that malaria poses a significant economic burden on rural households and individuals in terms of indirect costs due to the loss of working days. These costs are not well perceived by society and the substantial financial impact of the disease may be undervalued in public policy planning. Malaria control policies need to be integrated into development and poverty reduction programmes as socioeconomic development may be an effective and sustainable intervention against malaria in the long term.