Prevalence of malaria relapse: systematic review with meta-analysis

ABSTRACT Objective: systematic review with a meta-analysis of the prevalence of malaria relapse. Method: it consisted in a search for cross-sectional studies, carried out in three databases, without application of filters. A total of 1,924 articles were identified, selected based on eligibility criteria. The selection was made in pairs from the reading of the titles, abstracts and text. The meta-analysis was performed with a statistical program. Results: a sample of 1,308 patients with malaria, ranging from 70 to 586 patients in the study. Relapse was estimated at 0.47%, with a 95% confidence interval and 99.04% of squared i. In the included studies, the prevalence of relapse ranged from 17.00% to 92.85%. The result of the meta-analysis is considered relevant, despite the heterogeneity. Conclusion: relapse is a phenomenon that can contribute to the maintenance of the endemicity of malaria in the world and to introduce it in non-affected areas. In addition, there is the need for advancement in the production of knowledge regarding this disease, to qualify the research methods on prevalence.

Investigating and measuring the prevalence rate of malaria relapse is an important indicator for health services and for disease control programs.
This information can help nurses in the planning of interventions, guide managers in the application of financial resources, and foster research needed to explain the factors that determine the phenomenon at the local level and its endemic permanence (5,10) . In addition, this may be an important data for conducting economic evaluations studies in malaria control programs.
The objective of this study was to perform a systematic review, with meta-analysis, of observational studies that estimated the prevalence of malaria relapse.

Method
This is a systematic review with a meta-analysis of observational studies, conducted according to the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyzes (PRISMA) (11) . The study was guided by a structured question in the format of structured clinical question, as follows:  The reading and selection of the articles was carried out by two reviewers independently, starting by the titles, followed by the abstracts and full texts.
In the selection of the titles, we included all those that presented one of the following terms: relapse, prevalence, malaria and Plasmodium. When the application of the eligibility criteria was not sufficient to establish whether or not the article would be included in a phase, it was retained for reading the abstract.
The last phase of the selection was made by reading the full text of the articles.  were cross-sectional studies (12) .

Results
Five studies were included for the review and meta-analysis, according to Figure 2. All articles that did not present prevalence data of relapse were excluded because they were out of the scope of the research or because of the type of design.
Data were extracted from the studies selected for review, according to Figure 3. Two studies were excluded due to the use of secondary data obtained from health surveillance information systems. The prevalence of relapse was 0.47 and ranged from 17 to 92.85%, with a confidence interval of 95% (15)(16)(17)(18)(19) . The methodological procedures for recruitment and confirmation of relapse cases varied among the studies, and all five presented laboratory test as standard.
The meta-analysis expresses the prevalence of relapse, according to Figure 4, with a confidence interval of 95%.
The heterogeneity has I 2 of 99.04%, with p <0.001 and 95% confidence interval (0.434, 0.509). The difference between the studies can be considered high.

Discussion
The results show that malaria relapse is a worrisome phenomenon, since it can contribute to the disease remaining endemic in the affected areas and introduce it in non-affected areas, which can exacerbate the epidemiological picture and generate impacts on economic and social aspects (16) .
A study carried out in the city of Porto Velho, in the state of Rondônia, Brazil, indicated that the risk of relapse was estimated at 45.1/100 inhabitants and that this risk can be considered a high rate for a city of this size when compared to the rate for the Amazon region, estimated at 20.8% (8) . In Ethiopia, a study showed that 77% of secondary malaria cases are due to relapses (20) .
It is most likely that the cause of relapse is the reactivation of the hypnozoites and the great capacity of Plasmodium vivax to adapt biologically, with phenotypic changes (7) .
Also related to the biological factor of the parasite, a study carried out in Nepal estimated the prevalence rate of relapse at 17%, stating that there may be a relationship between the specific genotypes of Plasmodium vivax, which may differ between the different geographical regions, considering the biological aspects of the immunity of specific populations (19) .
Knowing the prevalence of relapse is crucial for public health, since it extends the possibilities of management by malaria control programs, allowing the understanding of the magnitude of the disease and, therefore, the adequate planning of actions by services, conferring greater effectiveness to health interventions.
From the epidemiological point of view, it allows users who remain with the disease, thus being sources of infection, to be identified in a timely manner, interrupting the transmission cycle (21) .
In order to verify the prevalence of relapse, it is necessary that Cure Slide Verification (CSV) be performed after the end of treatment. This is, therefore, a technology that should integrate the actions of the programs, since it is an important indicator for the health services for being able point operational problems of epidemiological surveillance, inform which users are sources of infection, "besides being useful to differentiate a new infection (new case)" from a relapse (21) .
Because it is a neglected disease, including its financing, there is an intimate relationship between its is of utmost importance, since in some territories this is the only health professional that is present and capable of establishing control and prevention measures (4) .
The result found in the meta-analysis of the prevalence of relapse shows high heterogeneity, 99.04%.
This is due to the fact that the original studies present significant differences in their methodological designs.
There is still a lack of methodological standardization for these types of studies, even when appropriate  A more specific look at relapse cases is also necessary due to its relevance for the maintenance of the endemic disease, its high prevalence rate and consequent challenges for health surveillance (25) .
The study presents limitations because the gray literature was not consulted and because of its high heterogeneity.
The results of this work can be used as a basis for conducting research on economic evaluation of health technologies to intervene in malaria.