Early warning systems for malaria outbreaks in Thailand: an anomaly detection approach.

BACKGROUND: Malaria continues to pose a significant health threat. Rapid identification of malaria infections and the deployment of active surveillance tools are crucial for achieving malaria elimination in regions where malaria is endemic, such as certain areas of Thailand. In this study, an anomaly detection system is introduced as an early warning mechanism for potential malaria outbreaks in countries like Thailand. METHODS: Unsupervised clustering-based, and time series-based anomaly detection algorithms are developed and compared to identify abnormal malaria activity in Thailand. Additionally, a user interface tailored for anomaly detection is designed, enabling the Thai malaria surveillance team to utilize these algorithms and visualize regions exhibiting unusual malaria patterns. RESULTS: Nine distinct anomaly detection algorithms we developed. Their efficacy in pinpointing verified outbreaks was assessed using malaria case data from Thailand spanning 2012 to 2022. The historical average threshold-based anomaly detection method triggered three times fewer alerts, while correctly identifying the same number of verified outbreaks when compared to the current method used in Thailand. A limitation of this analysis is the small number of verified outbreaks; further consultation with the Division of Vector Borne Disease could help identify more verified outbreaks. The developed dashboard, designed specifically for anomaly detection, allows disease surveillance professionals to easily identify and visualize unusual malaria activity at a provincial level across Thailand. CONCLUSION: An enhanced early warning system is proposed to bolster malaria elimination efforts for countries with a similar malaria profile to Thailand. The developed anomaly detection algorithms, after thorough comparison, have been optimized for integration with the current malaria surveillance infrastructure. An anomaly detection dashboard for Thailand is built and supports early detection of abnormal malaria activity. In summary, the proposed early warning system enhances the identification process for provinces at risk of outbreaks and offers easy integration with Thailand's established malaria surveillance framework.

An artesunate pharmacometric model to explain therapeutic responses in falciparum malaria.

BACKGROUND: The artemisinins are potent and widely used antimalarial drugs that are eliminated rapidly. A simple concentration-effect pharmacometric model does not explain why dosing more frequently than once daily fails to augment parasite clearance and improve therapeutic responses in vivo. Artemisinins can induce a temporary non-replicative or 'dormant' drug refractory state in Plasmodium falciparum malaria parasites which may explain recrudescences observed in clinical trials despite full drug susceptibility, but whether it explains the dosing-response relationship is uncertain. OBJECTIVES: To propose a revised model of antimalarial pharmacodynamics that incorporates reversible asexual parasite injury and temporary drug refractoriness in order to explain the failure of frequent dosing to augment therapeutic efficacy in falciparum malaria. METHODS: The model was fitted using a Bayesian Markov Chain Monte Carlo approach with the parasite clearance data from 39 patients with uncomplicated falciparum malaria treated with artesunate from western Cambodia and 40 patients from northwestern Thailand reported previously. RESULTS: The revised model captured the dynamics of parasite clearance data. Its predictions are consistent with observed therapeutic responses. CONCLUSIONS: A within-host pharmacometric model is proposed in which it is hypothesized that some malaria parasites enter a temporary drug refractory state after exposure to artemisinin antimalarials, which is followed by delayed parasite death or reactivation. The model fitted the observed sequential parasite density data from patients with acute P. falciparum malaria, and it supported reduced ring stage activity in artemisinin-resistant infections.

How to model the impact of vaccines for policymaking when the characteristics are uncertain: A case study in Thailand prior to the vaccine rollout during the COVID-19 pandemic.

Thailand faced a dilemma of which groups to prioritise with a limited first tranche of COVID-19 vaccinations in early 2021, at a time when there was low incidence and low mortality in the country. A mathematical modelling analysis was performed to compare the potential short-term impact of allocating the available doses to either the high severity group (over 65-year-olds) or the high transmission group (aged 20-39). At the time of the analysis, there was uncertainty about the precise characteristics of the vaccines available, in terms of their potential impact on transmission and reductions to the severity of infection. As such, a range of vaccine characteristic scenarios, with differing levels of severity and transmission reductions were explored. Using the evidence available at the time regarding severity reduction of infection due to the vaccines, the model suggested that vaccinating high severity group should be the priority if reductions in deaths is the priority. Vaccinating this group was found to have a direct impact on reducing the number of deaths, while the incidence and hospitalisations remained unchanged. However, the model found that vaccinating the high transmission group with a vaccine with sufficiently high protection against infection (more than 70%) could provide enough herd effects to delay the expected epidemic peak, resulting in both case and death reductions in both target groups. The model explored a 12-month time horizon. These analyses helped to inform the vaccination strategy in Thailand throughout 2021 and can inform future modelling studies for policymaking when the characteristics of vaccines are uncertain.

Assessing the cost-effectiveness of COVID-19 vaccines in a low incidence and low mortality setting: the case of Thailand at start of the pandemic.

OBJECTIVE: This study aimed to assess the cost-effectiveness of COVID-19 vaccines, preferred COVID-19 vaccine profiles, and the preferred vaccination strategies in Thailand. METHODS: An age-structured transmission dynamic model was developed based on key local data to evaluate economic consequences, including cost and health outcome in terms of life-years (LYs) saved. We considered COVID-19 vaccines with different profiles and different vaccination strategies such as vaccinating elderly age groups (over 65s) or high-incidence groups, i.e. adults between 20 and 39 years old who have contributed to more than 60% of total COVID-19 cases in the country thus far. Analyses employed a societal perspective in a 1-year time horizon using a cost-effectiveness threshold of 160,000 THB per LY saved. Deterministic and probabilistic sensitivity analyses were performed to identify and characterize uncertainty in the model. RESULTS: COVID-19 vaccines that block infection combined with social distancing were cost-saving regardless of the target population compared to social distancing alone (with no vaccination). For vaccines that block infection, the preferred (cost-effective) strategy was to vaccinate the high incidence group. Meanwhile, COVID-19 vaccines that reduces severity (including hospitalization and mortality) were cost-effective when the elderly were vaccinated, while vaccinating the high-incidence group was not cost-effective with this vaccine type. Regardless of vaccine type, higher vaccination coverage, higher efficacy, and longer protection duration were always preferred. More so, vaccination with social distancing measures was always preferred to strategies without social distancing. Quarantine-related costs were a major cost component affecting the cost-effectiveness of COVID-19 vaccines. CONCLUSION: COVID-19 vaccines are good value for money even in a relatively low-incidence and low-mortality setting such as Thailand, if the appropriate groups are vaccinated. The preferred vaccination strategies depend on the type of vaccine efficacy. Social distancing measures should accompany a vaccination strategy.

Classification of elderly pain severity from automated video clip facial action unit analysis: A study from a Thai data repository.

Data from 255 Thais with chronic pain were collected at Chiang Mai Medical School Hospital. After the patients self-rated their level of pain, a smartphone camera was used to capture faces for 10 s at a one-meter distance. For those unable to self-rate, a video recording was taken immediately after the move that causes the pain. The trained assistant rated each video clip for the pain assessment in advanced dementia (PAINAD). The pain was classified into three levels: mild, moderate, and severe. OpenFace© was used to convert the video clips into 18 facial action units (FAUs). Five classification models were used, including logistic regression, multilayer perception, naïve Bayes, decision tree, k-nearest neighbors (KNN), and support vector machine (SVM). Out of the models that only used FAU described in the literature (FAU 4, 6, 7, 9, 10, 25, 26, 27, and 45), multilayer perception is the most accurate, at 50%. The SVM model using FAU 1, 2, 4, 7, 9, 10, 12, 20, 25, and 45, and gender had the best accuracy of 58% among the machine learning selection features. Our open-source experiment for automatically analyzing video clips for FAUs is not robust for classifying pain in the elderly. The consensus method to transform facial recognition algorithm values comparable to the human ratings, and international good practice for reciprocal sharing of data may improve the accuracy and feasibility of the machine learning's facial pain rater.

Sensitive and accurate DNA metabarcoding of parasitic helminth mock communities using the mitochondrial rRNA genes.

Next-generation sequencing technologies have accelerated the pace of helminth DNA metabarcoding research, enabling species detection in bulk community samples. However, finding suitable genetic markers with robust species-level resolution and primers targeting a broad species range among parasitic helminths are some of the challenges faced. This study aimed to demonstrate the potential use of the mitochondrial 12S and 16S rRNA genes for parasitic helminth (nematodes, trematodes, cestodes) DNA metabarcoding. To demonstrate the robustness of the 12S and 16S rRNA genes for DNA metabarcoding, we determined the proportion of species successfully recovered using mock helminth communities without environment matrix and mock helminth communities artificially spiked with environmental matrices. The environmental matrices are human fecal material, garden soil, tissue, and pond water. Our results revealed the robustness of the mitochondrial rRNA genes, through the high sensitivity of the 12S rRNA gene, and the effectiveness of the 12S and 16S primers targeting platyhelminths. With the mitochondrial rRNA genes, a broad range of parasitc helminths were successfully detected to the species level. The potential of the mitochondrial rRNA genes for helminth DNA metabarcoding was demonstrated, providing a valuable gateway for future helminth DNA metabarcoding applications like helminth detection and biodiversity studies.

The potential use of mitochondrial ribosomal genes (12S and 16S) in DNA barcoding and phylogenetic analysis of trematodes.

BACKGROUND: Genetic markers like the nuclear ribosomal RNA (rRNA) genes, internal transcribed spacer regions, mitochondrial protein-coding genes, and genomes have been utilized for molecular identification of parasitic trematodes. However, challenges such as the design of broadly applicable primers for the vast number of species within Digenea and the genetic markers' ability to provide sufficient species-level resolution limited their utility. This study presented novel and broadly applicable primers using the mitochondrial 12S and 16S rRNA genes for Digenea and aimed to show their suitability as alternative genetic markers for molecular identification of orders Plagiorchiida, Echinostomida, and Strigeida. RESULTS: Our results revealed that the mitochondrial 12S and 16S rRNA genes are suitable for trematode molecular identification, with sufficient resolution to discriminate closely related species and achieve accurate species identification through phylogenetic placements. Moreover, the robustness of our newly designed primers to amplify medically important parasitic trematodes encompassing three orders was demonstrated through successful amplification. The convenience and applicability of the newly designed primers and adequate genetic variation of the mitochondrial rRNA genes can be useful as complementary markers for trematode molecular-based studies. CONCLUSIONS: We demonstrated that the mitochondrial rRNA genes could be alternative genetic markers robust for trematode molecular identification and potentially helpful for DNA barcoding where our primers can be widely applied across the major Digenea orders. Furthermore, the potential of the mitochondrial rRNA genes for molecular systematics can be explored, enhancing their appeal for trematode molecular-based studies. The novelty of utilizing the mitochondrial rRNA genes and the designed primers in this study can potentially open avenues for species identification, discovery, and systematics in the future.

Assessing the suitability of mitochondrial and nuclear DNA genetic markers for molecular systematics and species identification of helminths.

BACKGROUND: Genetic markers are employed widely in molecular studies, and their utility depends on the degree of sequence variation, which dictates the type of application for which they are suited. Consequently, the suitability of a genetic marker for any specific application is complicated by its properties and usage across studies. To provide a yardstick for future users, in this study we assess the suitability of genetic markers for molecular systematics and species identification in helminths and provide an estimate of the cut-off genetic distances per taxonomic level. METHODS: We assessed four classes of genetic markers, namely nuclear ribosomal internal transcribed spacers, nuclear rRNA, mitochondrial rRNA and mitochondrial protein-coding genes, based on certain properties that are important for species identification and molecular systematics. For molecular identification, these properties are inter-species sequence variation; length of reference sequences; easy alignment of sequences; and easy to design universal primers. For molecular systematics, the properties are: average genetic distance from order/suborder to species level; the number of monophyletic clades at the order/suborder level; length of reference sequences; easy alignment of sequences; easy to design universal primers; and absence of nucleotide substitution saturation. Estimation of the cut-off genetic distances was performed using the 'K-means' clustering algorithm. RESULTS: The nuclear rRNA genes exhibited the lowest sequence variation, whereas the mitochondrial genes exhibited relatively higher variation across the three groups of helminths. Also, the nuclear and mitochondrial rRNA genes were the best possible genetic markers for helminth molecular systematics, whereas the mitochondrial protein-coding and rRNA genes were suitable for molecular identification. We also revealed that a general gauge of genetic distances might not be adequate, using evidence from the wide range of genetic distances among nematodes. CONCLUSION: This study assessed the suitability of DNA genetic markers for application in molecular systematics and molecular identification of helminths. We provide a novel way of analyzing genetic distances to generate suitable cut-off values for each taxonomic level using the 'K-means' clustering algorithm. The estimated cut-off genetic distance values, together with the summary of the utility and limitations of each class of genetic markers, are useful information that can benefit researchers conducting molecular studies on helminths.

Determinants of MDA impact and designing MDAs towards malaria elimination.

Malaria remains at the forefront of scientific research and global political and funding agendas. Malaria models have consistently oversimplified how mass interventions are implemented. Here, we present an individual based, spatially explicit model of P. falciparum malaria transmission that includes all the programmatic implementation details of mass drug administration (MDA) campaigns. We uncover how the impact of MDA campaigns is determined by the interaction between implementation logistics, patterns of human mobility and how transmission risk is distributed over space. Our results indicate that malaria elimination is only realistically achievable in settings with very low prevalence and can be hindered by spatial heterogeneities in risk. In highly mobile populations, accelerating MDA implementation increases likelihood of elimination; if populations are more static, deploying less teams would be cost optimal. We conclude that mass drug interventions can be an invaluable tool towards malaria elimination in low endemicity areas, specifically when paired with effective vector control.

Factors affecting the electrocardiographic QT interval in malaria: A systematic review and meta-analysis of individual patient data.

BACKGROUND: Electrocardiographic QT interval prolongation is the most widely used risk marker for ventricular arrhythmia potential and thus an important component of drug cardiotoxicity assessments. Several antimalarial medicines are associated with QT interval prolongation. However, interpretation of electrocardiographic changes is confounded by the coincidence of peak antimalarial drug concentrations with recovery from malaria. We therefore reviewed all available data to characterise the effects of malaria disease and demographic factors on the QT interval in order to improve assessment of electrocardiographic changes in the treatment and prevention of malaria. METHODS AND FINDINGS: We conducted a systematic review and meta-analysis of individual patient data. We searched clinical bibliographic databases (last on August 21, 2017) for studies of the quinoline and structurally related antimalarials for malaria-related indications in human participants in which electrocardiograms were systematically recorded. Unpublished studies were identified by the World Health Organization (WHO) Evidence Review Group (ERG) on the Cardiotoxicity of Antimalarials. Risk of bias was assessed using the Pharmacoepidemiological Research on Outcomes of Therapeutics by a European Consortium (PROTECT) checklist for adverse drug events. Bayesian hierarchical multivariable regression with generalised additive models was used to investigate the effects of malaria and demographic factors on the pretreatment QT interval. The meta-analysis included 10,452 individuals (9,778 malaria patients, including 343 with severe disease, and 674 healthy participants) from 43 studies. 7,170 (68.6%) had fever (body temperature ≥ 37.5°C), and none developed ventricular arrhythmia after antimalarial treatment. Compared to healthy participants, patients with uncomplicated falciparum malaria had shorter QT intervals (-61.77 milliseconds; 95% credible interval [CI]: -80.71 to -42.83) and increased sensitivity of the QT interval to heart rate changes. These effects were greater in severe malaria (-110.89 milliseconds; 95% CI: -140.38 to -81.25). Body temperature was associated independently with clinically significant QT shortening of 2.80 milliseconds (95% CI: -3.17 to -2.42) per 1°C increase. Study limitations include that it was not possible to assess the effect of other factors that may affect the QT interval but are not consistently collected in malaria clinical trials. CONCLUSIONS: Adjustment for malaria and fever-recovery-related QT lengthening is necessary to avoid misattributing malaria-disease-related QT changes to antimalarial drug effects. This would improve risk assessments of antimalarial-related cardiotoxicity in clinical research and practice. Similar adjustments may be indicated for other febrile illnesses for which QT-interval-prolonging medications are important therapeutic options.

Model evaluation of target product profiles of an infant vaccine against respiratory syncytial virus (RSV) in a developed country setting.

Respiratory syncytial virus (RSV) is a major cause of lower respiratory tract disease in children worldwide and is a significant cause of hospital admissions in young children in England. No RSV vaccine has been licensed but a number are under development. In this work, we present two structurally distinct mathematical models, parameterized using RSV data from the UK, which have been used to explore the effect of introducing an RSV paediatric vaccine to the National programme. We have explored different vaccine properties, and dosing regimens combined with a range of implementation strategies for RSV control. The results suggest that vaccine properties that confer indirect protection have the greatest effect in reducing the burden of disease in children under 5 years. The findings are reinforced by the concurrence of predictions from the two models with very different epidemiological structure. The approach described has general application in evaluating vaccine target product profiles.

Towards malaria elimination in Savannakhet, Lao PDR: mathematical modelling driven strategy design.

BACKGROUND: The number of Plasmodium falciparum malaria cases around the world has decreased substantially over the last 15 years, but with the spread of resistance against anti-malarial drugs and insecticides, this decline may not continue. There is an urgent need to consider alternative, accelerated strategies to eliminate malaria in countries like Lao PDR, where there are a few remaining endemic areas. A deterministic compartmental modelling tool was used to develop an integrated strategy for P. falciparum elimination in the Savannakhet province of Lao PDR. The model was designed to include key aspects of malaria transmission and integrated control measures, along with a user-friendly interface. RESULTS: Universal coverage was the foundation of the integrated strategy, which took the form of the deployment of community health workers who provided universal access to early diagnosis, treatment and long-lasting insecticidal nets. Acceleration was included as the deployment of three monthly rounds of mass drug administration targeted towards high prevalence villages, with the addition of three monthly doses of the RTS,S vaccine delivered en masse to the same high prevalence sub-population. A booster dose of vaccine was added 1 year later. The surveillance-as-intervention component of the package involved the screening and treatment of individuals entering the simulated population. CONCLUSIONS: In this modelling approach, the sequential introduction of a series of five available interventions in an integrated strategy was predicted to be sufficient to stop malaria transmission within a 3-year period. These interventions comprised universal access to early diagnosis and adequate treatment, improved access to long-lasting insecticidal nets, three monthly rounds of mass drug administration together with RTS,S vaccination followed by a booster dose of vaccine, and screening and treatment of imported cases.

A Dynamic Stress Model Explains the Delayed Drug Effect in Artemisinin Treatment of Plasmodium falciparum.

Artemisinin resistance constitutes a major threat to the continued success of control programs for malaria, particularly in light of developing resistance to partner drugs. Improving our understanding of how artemisinin-based drugs act and how resistance manifests is essential for the optimization of dosing regimens and the development of strategies to prolong the life span of current first-line treatment options. Recent short-drug-pulse in vitro experiments have shown that the parasite killing rate depends not only on drug concentration but also the exposure time, challenging the standard pharmacokinetic-pharmacodynamic (PK-PD) paradigm in which the killing rate depends only on drug concentration. Here, we introduce a dynamic stress model of parasite killing and show through application to 3D7 laboratory strain viability data that the inclusion of a time-dependent parasite stress response dramatically improves the model's explanatory power compared to that of a traditional PK-PD model. Our model demonstrates that the previously reported hypersensitivity of early-ring-stage parasites of the 3D7 strain to dihydroartemisinin compared to other parasite stages is due primarily to a faster development of stress rather than a higher maximum achievable killing rate. We also perform in vivo simulations using the dynamic stress model and demonstrate that the complex temporal features of artemisinin action observed in vitro have a significant impact on predictions for in vivo parasite clearance. Given the important role that PK-PD models play in the design of clinical trials for the evaluation of alternative drug dosing regimens, our novel model will contribute to the further development and improvement of antimalarial therapies.

Identifying artemisinin resistance from parasite clearance half-life data with a simple Shiny web application.

The emergence of artemisinin-resistant Plasmodium falciparum malaria is a major threat to malaria elimination. New tools for supporting the surveillance of artemisinin resistance are critical for current and future malaria control and elimination strategies. We have developed an open-access, user-friendly, web-based tool to analyse parasite clearance half-life data of P. falciparum infected patients after treatment with artemisinin derivatives, so that resistance to artemisinin can be identified. The tool can be accessed at bit.ly/id_artemisinin_resistance.

Predicting the relative impacts of maternal and neonatal respiratory syncytial virus (RSV) vaccine target product profiles: A consensus modelling approach.

BACKGROUND: Respiratory syncytial virus (RSV) is the major viral cause of infant and childhood lower respiratory tract disease worldwide. Defining the optimal target product profile (TPP) is complicated due to a wide range of possible vaccine properties, modalities and an incomplete understanding of the mechanism of natural immunity. We report consensus population level impact projections based on two mathematical models applied to a low income setting. METHOD: Two structurally distinct age-specific deterministic compartmental models reflecting uncertainty associated with the natural history of infection and the mechanism by which immunity is acquired and lost were constructed. A wide range of vaccine TPPs were explored including dosing regime and uptake, and effects in the vaccinated individual on infectiousness, susceptibility, duration of protection, disease severity and interaction with maternal antibodies and natural induced immunity. These were combined with a range of vaccine implementation strategies, targeting the highest priority age group and calibrated using hospitalization data from Kilifi County Hospital, Kenya. FINDINGS: Both models were able to reproduce the data. The impact predicted by the two models was qualitatively similar across the range of TPPs, although one model consistently predicted higher impact than the other. For a proposed realistic range of scenarios of TPP combinations, the models predicted up to 70% reduction in hospitalizations in children under five years old. Vaccine designs which reduced the duration and infectiousness of infection were predicted to have higher impacts. The models were sensitive to the coverage and rate of loss of vaccine protection but not to the interaction between vaccine and maternal/naturally acquired immunity. CONCLUSION: The results suggest that vaccine properties leading to reduced virus circulation by lessening the duration and infectiousness of infection upon challenge are of major importance in population RSV disease control. These features should be a focus for vaccine development.

Modelling malaria elimination on the internet.

BACKGROUND: Unprecedented efforts are underway to eliminate malaria. Mathematical modelling can help to determine the optimal strategies for malaria elimination in different epidemiological settings. This is necessary as there is limited scope for expensive and time-consuming field studies and failure of planned elimination strategies is likely to discourage ongoing investment by funders. However, there has been very little modelling of malaria elimination and little direct involvement of policymakers in its development. There is thus an urgent need for user-friendly and accessible models purpose-designed in collaboration with policymakers to answer pertinent questions arising from the field. RESULTS: An internet site is presented with a simple mathematical modelling platform for population level models of malaria elimination. It is freely accessible to all and designed to be flexible so both the platform and models can be developed through interaction with users. The site is an accessible introduction to modelling for a non-mathematical audience, and lessons learned from the project will help inform future development of mathematical models and improve communication of modelling results. Currently it hosts a simple model of strategies for malaria elimination and this will be developed, and more models added, over time. The iterative process of feedback and development will result in an educational and planning tool for non-modellers to assist with malaria elimination efforts worldwide. CONCLUSIONS: By collaboration with end users, iterative development of mathematical models of malaria elimination through this internet platform will maximize its potential as an educational and public health policy planning tool. It will also assist with preliminary optimisation of local malaria elimination strategies before commitment of valuable resources.

Intrahost modeling of artemisinin resistance in Plasmodium falciparum.

Artemisinin-resistant Plasmodium falciparum malaria has emerged in western Cambodia. Resistance is characterized by prolonged in vivo parasite clearance times (PCTs) following artesunate treatment. The biological basis is unclear. The hypothesis that delayed parasite clearance results from a stage-specific reduction in artemisinin sensitivity of the circulating young asexual parasite ring stages was examined. A mathematical model was developed, describing the intrahost parasite stage-specific pharmacokinetic-pharmacodynamic relationships. Model parameters were estimated using detailed pharmacokinetic and parasite clearance data from 39 patients with uncomplicated falciparum malaria treated with artesunate from Pailin (western Cambodia) where artemisinin resistance was evident and 40 patients from Wang Pha (northwestern Thailand) where efficacy was preserved. The mathematical model reproduced the observed parasite clearance for each patient with an accurate goodness of fit (rmsd: 0.03-0.67 in log(10) scale). The parameter sets that provided the best fits with the observed in vivo data consist of a highly conserved concentration-effect relationship for the trophozoite and schizont parasite stages, but a variable relationship for the ring stages. The model-derived assessment suggests that the efficacy of artesunate on ring stage parasites is reduced significantly in Pailin. This result supports the hypothesis that artemisinin resistance mainly reflects reduced ring-stage susceptibility and predicts that doubling the frequency of dosing will accelerate clearance of artemisinin-resistant parasites.

The role of mathematical modelling in guiding the science and economics of malaria elimination.

Unprecedented efforts are now underway to eliminate malaria from many regions. Despite the enormous financial resources committed, if malaria elimination is perceived as failing it is likely that this funding will not be sustained. It is imperative that methods are developed to use the limited data available to design site-specific, cost-effective elimination programmes. Mathematical modelling is a way of including mechanistic understanding to use available data to make predictions. Different strategies can be evaluated much more rapidly than is possible through trial and error in the field. Mathematical modelling has great potential as a tool to guide and inform current elimination efforts. Economic modelling weighs costs against characterised effects or predicted benefits in order to determine the most cost-efficient strategy but has traditionally used static models of disease not suitable for elimination. Dynamic mathematical modelling and economic modelling techniques need to be combined to contribute most effectively to ongoing policy discussions. We review the role of modelling in previous malaria control efforts as well as the unique nature of elimination and the consequent need for its explicit modelling, and emphasise the importance of good disease surveillance. The difficulties and complexities of economic evaluation of malaria control, particularly the end stages of elimination, are discussed.

Hyperparasitaemia and low dosing are an important source of anti-malarial drug resistance.

BACKGROUND: Preventing the emergence of anti-malarial drug resistance is critical for the success of current malaria elimination efforts. Prevention strategies have focused predominantly on qualitative factors, such as choice of drugs, use of combinations and deployment of multiple first-line treatments. The importance of anti-malarial treatment dosing has been underappreciated. Treatment recommendations are often for the lowest doses that produce "satisfactory" results. METHODS: The probability of de-novo resistant malaria parasites surviving and transmitting depends on the relationship between their degree of resistance and the blood concentration profiles of the anti-malarial drug to which they are exposed. The conditions required for the in-vivo selection of de-novo emergent resistant malaria parasites were examined and relative probabilities assessed. RESULTS: Recrudescence is essential for the transmission of de-novo resistance. For rapidly eliminated anti-malarials high-grade resistance can arise from a single drug exposure, but low-grade resistance can arise only from repeated inadequate treatments. Resistance to artemisinins is, therefore, unlikely to emerge with single drug exposures. Hyperparasitaemic patients are an important source of de-novo anti-malarial drug resistance. Their parasite populations are larger, their control of the infection insufficient, and their rates of recrudescence following anti-malarial treatment are high. As use of substandard drugs, poor adherence, unusual pharmacokinetics, and inadequate immune responses are host characteristics, likely to pertain to each recurrence of infection, a small subgroup of patients provides the particular circumstances conducive to de-novo resistance selection and transmission. CONCLUSION: Current dosing recommendations provide a resistance selection opportunity in those patients with low drug levels and high parasite burdens (often children or pregnant women). Patients with hyperparasitaemia who receive outpatient treatments provide the greatest risk of selecting de-novo resistant parasites. This emphasizes the importance of ensuring that only quality-assured anti-malarial combinations are used, that treatment doses are optimized on the basis of pharmacodynamic and pharmacokinetic assessments in the target populations, and that patients with heavy parasite burdens are identified and receive sufficient treatment to prevent recrudescence.