|Study type||Objectives (outcomes measures)||Results and conclusions||Learning points and recommendations: transmission dynamics, vectorial capacity, co-infection||Main limitations, comments||Future research, public health policy & practice||Funding|
|Araujo et al., 2015 ||
|Spatial modelling||To assess incidence of dengue and its association with multiple environmental and socioeconomic factors||
Dengue incidence was higher in areas of high land surface temperature (28–32 °C) and in slum-like areas. Most cases occurred in areas with low vegetation cover.|
Laboratory vector experiments indicated that higher temperature was favourable for Ae. aegypti proliferation.
Incidence mostly affected by land surface temperature (LST), due to Ae. aegypti proliferation.|
Recovering vegetation on urban heat island (UHI).
Importance of evaluating LST (environmental inner city conditions), slum-like areas and human activities (man-made reservoirs) for DENV transmission.|
Translational study in dengue.
|Dibo et al., 2008 ||
|Ae. aegypti female abundance (collection) and eggs (oviposition traps); sensitivity of both methods; correlation of entomological indices (positivity and averages of females and eggs) dengue incidence and climate.||Almost 20 000 specimens of culicids were caught, ~ 10% were Ae. aegypti. The ovitrap caught 165 000 eggs in the period and 173 autochthonous cases of dengue were recorded.||Relationship between climatic factors, vector and disease. Use of information about climate for early detection of epidemics and for establishing more effective prevention strategies.||Using information about reported dengue fever (DF) cases may represent only part of the total number of dengue virus infections, rainfall data were obtained for a location not exactly at the mosquito collection site, and temperatures were available only for a neighbouring city.||Widespread application in planning and undertaking surveillance and dengue control activities||NR|
|Teixeira et al., 2011 ||
Rio de Janeiro
|Ecologic spatial modelling||Dengue spatial distribution and its relationship to environmental and socioeconomic variables||Direct association was found between dengue incidence and rainfall, in both final generalized linear model and some monthly CAR models. Direct association was also found between dengue incidence and a 1-month rainfall time lag.||
Significant direct association between dengue incidence and rainfall +lag, Gini index, and Breteau index for Ae. albopictus, but only the Gini index showed strong association.|
Importance of socio-environmental variables in the dynamics of dengue transmission.
A methodological limitation was data georeferencing, due to the incomplete notification data.|
Use of Breteau index as an indicator of vector infestation since it was considered precarious and not a reliable indicator.
|Need to investigate and analyze the association between dengue incidence and explanatory variables through more complex models (complete Bayesian model) capable of simultaneously capturing the spatial and temporal autocorrelation structures.||NR|
|Corrêa et al., 2005 ||
|Ecologic October 1997– May 2001||Link between vector infestation index and dengue incidence||Ae. aegypti infestation was positively associated with dengue incidence.||Higher building infestation rates were associated with higher risk of disease. This correlation was significant but weak.||Building infestation rates have limitations for estimating vector infestation.||To reduce and maintain the house infestation index below the 1% threshold.||NG|
|Teixeira et al., 2002 ||
|Prospective populational study, seroprevalence and incidence after 1 year of follow-up||Relationship between intensity of viral circulation and standard of living, environmental quality and vector density||Wide range of prevalence of previous dengue infection (16.2% to 97.6%) in the area. Areas with a lower prevalence at baseline presented higher incidence of dengue infection. High seroincidence was observed in areas with low vector indices.||Increase in herd immunity to two different serotypes of dengue, and maintenance of the environmental conditions necessary and sufficient for transmission of new serotypes, created conditions for occurrence of epidemic haemorrhagic dengue.||Some years after the virus’ introduction, transmission continues to occur, but with a decrease in the perception of the problem by individuals and health professionals and hence reduced sensitivity of the surveillance system.||Further studies are required to identify the behavioural and environmental risk factors (public and private domains) that have the greatest influence on transmission.||NG|
|Méndez et al., 2006 ||
Valle del Cauca
Cities: Cali, Palmira, Tulua, Buenaventura
|Prospective study||Study dynamics during and after epidemics; mosquito infection rates, cases, serotypes (infection rates)||Infections in students: asymptomatic cases outnumber symptomatic (silent circulation of DENV). Infections in mosquitoes: Ae. aegypti adult mosquito and larvae house indexes were not found to be associated with increased burden of disease.||Virological surveillance for detecting infected Aedes mosquitoes as early warning system for outbreaks.||Asymptomatic infections potential source of subsequent outbreaks; to include pooled mosquito infection rates in entomologic surveillance.||Need to understand a) inter-outbreak virus circulation and b) Ae. albopictus vectorial capacity.||NG|
|Estallo et al., 2014 ||
|Descriptive||Spatio-temporal dynamics of DF outbreak 2009||
Spatio-temporal cluster of dengue cases. Outbreak started because of imported cases from neighbouring provinces.|
Ae. aegypti infestation levels were not associated with occurrence of the cases.
|Significant risk factor for dengue emergence: travel, migration and displacement within and outside the country. Dengue virus spread may be related to human movements.||Applications of vector control measures (focal intervention at dengue positive houses) may have interrupted DENV transmission.||Need to develop innovative strategies of vector control and arboviral surveillance to prevent future outbreaks.||NG, U|
|Martinez-Vega et al., 2015 ||
2 endemic communities
|Prospective cohort with spatial modelling||To determine risk in individuals living near dengue cases (exposed cohort); vector data||At baseline, seroprevalence of infection was ~ 44% (5–14 yrs) and 92% (> 35 years) indicating high exposure levels. 3-fold increased risk of dengue infection among the exposed cohort.||Asymptomatic subjects could participate in transmission; peri-domestic transmission was important for ~ 3 months.||Spread of infection within the community mainly depended on human mobility; limited role of vector (50-m ratio) for dengue spread.||Clarify the role of asymptomatic individuals in transmission. Private health centres should contribute to notifications.||NG, PI|
|Reiter et al., 2003 ||
|Two serosurveys||Investigations of 1999 outbreak affecting cross border urban area of Laredo, USA, and Nuevo Laredo, Mexico.||
Incidence of dengue was higher in Mexico (16%) than on USA side (1.3%).|
Vector was more abundant in Texas where dengue transmission was lower.
|Prevalence of dengue in Texas was primarily due to economic, rather than climatic, factors.||Need to explore role of other animals as feeding source for Aedes mosquitoes.||Growing economy supporting use of air conditioning likely to decrease dengue rates.||NG|
|Barrera et al., 2011 ||
|Longitudinal modelling study||Influence of weather and human factors on Aedes vector and cases (numbers)||Containers for water storage and discarded tires were considered important breeding sites. Peak in mosquitoes’ density preceded increase in dengue incidence.||Weather and human factors driving Ae. aegypti dynamics and oviposition.||Passive surveillance system captures small number of symptomatic cases and misses most asymptomatic ones.||Consider using ovitraps for surveillance.||IG|
|Sanchez et al., 2006 ||
|Case control study||Study Aedes larval indices and risk of Havana outbreak (2000)||
Effect of Breteau index (BI) within a radius of 100 m at 2-month intervals.|
Larval indices used in study predicted transmission
with 78% sensitivity and 63% specificity.
|Analysis of BI at local level, with human defined boundaries, could be introduced in control programs to identify high risk areas.||Underestimating larvae; increased capture after dengue reported; ‘missing’ cases reported; lack of transferability to other epidemic settings or locations.||Similar studies in future epidemics and in other settings are necessary to verify general applicability.||NG, IG|
|Fouque et al., 2004 ||
|Descriptive||Epidemiology of dengue after the 1st epidemic of dengue hemorrhagic fever (DHF) (1991–1993) and during endemic period (1993–1995)||
DENV1, 2 and 4 were isolated from dengue cases. DENV4 from mosquitoes.|
Widespread circulation of DENV in urban and rural areas.
Small outdoor containers were most common vector breeding grounds.
|Vertical transmission of DENV indicated that the vector can be a reservoir of dengue virus.||Females deposit small numbers of eggs in several breeding containers. Not possible to estimate fluctuations in Ae. aegypti populations.||Understanding of endemic circulation of dengue is a prerequisite for development of a dengue early warning system.||NG|
|Wu et al., 2007 ||
|Time series modelling||Association between weather variability and dengue fever||Incidence of DF was negatively associated with monthly temperature deviation and relative humidity; time lag of 2 months.||
Weather variables (temperature and humidity) were significantly associated with dengue incidence rate.|
Vector density did not appear to be a good predictor of disease occurrence.
|Change in: disease diagnostic or classification criteria, urbanization status/land use, vector control program, and personal protection.||Include development and verification of analytical models appropriate for predicting influences of global warming on changes in disease patterns at regional level.||NG|
|Chang et al., 2015 ||
Tainan (non-endemic areas)
|Time series regression modelling study||Temporal relation between local weather, entomology, and confirmed cases||Risk of increased number of cases was associated with entomologic indices on a lag of 2 weeks or 1 month.||Proposed 2-stage model suitable to identify best set of lag effects to generate outbreak prediction.||
Complex relation between human hosts, environmental factors and dynamic changes in mosquito density.|
Difficult to extrapolate to other settings with distinct weather conditions, human immunity profiles and vector distributions.
|Suggestion for other countries to apply the authors’ proposed 2-stage modelling system.||NG|
|Ali et al., 2003 ||
|Spatial modelling study||Association of spatial clusters and vector during large outbreak (distribution; relative risk)||Dengue clusters far less observed in major hospitals than within or around the households; importance of Ae. albopictus in the transmission.||Case clusters and Ae. albopictus were linked in densely populated area.||Hospital-based surveillance underestimates disease burden; self-reporting cases limit precision.||Controlling the Aedes mosquito; improving case management.||NG, IG, NGO|
|Anders et al., 2015 ||
Ho Chi Minh City
|Prospective spatial modelling study||Distribution of dengue risk (prevalence, odds ratios)||At baseline, there was 2-fold higher risk in case clusters than controls. Follow-up (14 days) suggested no excess risk for dengue infection in clusters. Prevalence of DENV infection in Ae. aegypti was similar in cases and control houses, and low (1%).||Evidence for some household dengue risk clustering, on short temporal scale rather than sustained chains of localized transmission.||
Short follow-up: longer focal transmission chains missed? Self-reporting: underestimating symptomatic infections?|
Vector control applied after onset of illness causing reduced clustering observed?
|Reactive peri-focal insecticide spraying unlikely to prevent many additional infections.||U|
|Yang et al., 2009 ||
|Descriptive of dengue outbreak||Factors determining the dengue outbreak and measures to avoid additional epidemic outbreaks||Dengue incidence had no relationship with local meteorological factors. Ae. albopictus was the main vector.||Social factors and hygienic conditions for endemic villagers and immigrant workers.||Not reported.||Public education to include handling of environmental factors (artificial containers with stagnant water). Socioeconomic conditions should be taken into account to interrupt transmission.||NG|
|Peng et al., 2012 ||
|Descriptive of outbreak||Describe epidemic, risk factors and control measures||
Attack rate: ~ 51/100000|
Ae. albopictus was only vector species responsible for the outbreak. DENV1 serotype.
Dengue outbreak initiated by imported case from Southeast Asia.
|Urbanization, population density, habitats and habits favouring Ae. albopictus; high infestation, poor housing.||Underreporting and/or misclassification of dengue.||
Importance of a surveillance system for infectious diseases control.|
Surveillance needed in rapidly urbanized areas and among immigrants.
|Sang et al., 2014 ||
|Modelling study||Predicting local dengue transmission in Guangzhou, China, influence of imported cases, vector density, climate variability||
Imported dengue cases, mosquito density and climate variability are important in dengue transmission.
|Imported DF cases, mosquito density were critical for DENV transmission.||Relation between herd immunity, mosquito-human interaction, virus strain and mosquito daily survival rate.||Establishment of an early warning system (incorporating imported cases, mosquitoes density and climate variability), using existing surveillance datasets will help to control and prevent dengue locally.||NG|
|Seidahmed et al., 2012 ||
|Longitudinal study||Spatial and temporal patterns of dengue transmission||Spatial distribution of dengue was irregular in the city. IgM seroprevalence ranged between 3 and 8% among healthy residents and incidence rate was 35 new clinical cases per 10 000 individuals.||
Main determinants of dengue outbreaks were increased maritime traffic and weather variation.|
It should be feasible to carry out timely vector control measures to prevent or reduce dengue transmission in this coastline area.
Further research is needed to define whether there is a true disappearance of Ae. aegypti and, if so, how and from where the vector is reintroduced.|
Study impact of climate and socioeconomic changes on dengue emergence in the Red Sea region.
|Study type||Objectives (outcomes measures)||Results and conclusions||Learning points and recommendations: transmission dynamics, vectorial capacity, co-infection||Main limitations, comments||Future research, public health policy & practice||Funding|
|Woyessa et al., 2004 ||
|Descriptive||Study malaria transmission in Akaki town, at 2110 m altitude; parasitaemia prevalence; malaria/mosquito species frequency||Parasitaemia in 3.7% of 2136 blood films/3 months (69% vivax, 31% falciparum), none in last month (dry season). An. arabiensis/An. chrystyi predominant species (fairly low quantities), suggesting indigenous transmission.||Malaria risk in urbanized African highlands increasing, especially during rains. Possible link to climate change. Long periods of non-transmission reduce population immunity, with risk of recurrent epidemics.||Low case numbers in dry season possibly a result of bednet use/interventions. Malaria transmission low in cities (water pollution) but short-term increase from extra breeding sites during rainy season.||Address vulnerability of highland population of short period malaria transmission and associated epidemics. Apply sustainable and integrated vector control (breeding sites)/case management to prevent epidemics.||U|
|Ebenezer et al., 2016 ||
|Descriptive, modelling, spatial||Correlation of An. gambiae entomological inoculation rate (EIR) and malaria prevalence and incidence rates in various ecozones||Man-biting rate high (6.9) in mangrove coastal water, vs. fresh/brackish swamp water; An. gambiae s.s. infection rate 13% = most efficient and anthropophilic vector (EIR = 70); PR and IR can predict PR-EIR (71%) or IR-EIR (64%).||
MBR differs across different eco-vegetational zones of Bayelsa State.|
Authors propose EIR is a more direct measure of malaria transmission intensity, compared to PR or IR alone.
Methods unclear whether any malaria species or FM only.|
EIR not suitable for inter-age/population comparisons, due to variable factors such as host biomass.
|When assessing efficacy of transmission control measures, both entomological (EIR) and clinical (prevalence and incidence) data need to be considered.||NR|
|El Sayed et al., 2000 ||
|Descriptive||Transmission in low-income peri-urban Ed dekheinat vs. suburban El manshia, in Khartoum (prevalence rates, mosquito density)||An. arabiensis (only vector species) showed higher MBR and indoor density in low vs. high-income areas, especially in rainy season. Parasite screening: FM in high-income areas only, also other species in low-income areas; most FM rates seen in < 15 year age.||Transmission in semi-arid Khartoum unstable, with seasonal pattern and higher risk in low-income peri-urban areas of urban expansion (bringing residential areas closer to cultivated and irrigated land).||Low-income or proximity to agriculture responsible for increased transmission in Ed dekheinat, or interaction? Authors describe major climate, economic, social and political changes to Sudanese capital in recent decades.||
Need for improved malaria control to reflect increasing urbanization and changing malaria epidemiology in Africa.|
Aim for sustained decrease in malaria morbidity and mortality from epidemics.
|Ivan et al., 2012 ||Malaria and helminth co-infection (HIV infected women > 12 m ART in Rwanda)||Descriptive||Malaria (Plasmodium species), helminth or dual infection; cross-sections of peri-urban and rural pregnant women (n = 338); prevalence/OR||Malaria prevalence lower in peri-urban (12%) vs. rural (30%) areas; also helminth infections (33% vs. 45%; n.s.) high co-infection rates (5% vs. 15%), Nematodes found: Ascaris (21%), Trichuris (9%) and hookworm (1%);||Possible differential effect of ART regimen type on P. falciparum/Trichuris co-infection.||Cross-sectional study: no temporal/causality; Study too small to measure true differences, especially potential differential effect of some ART regimens.||Complex immunological profile of co-infection, effect on anaemia (caused by each of the three infections). Malaria/helminth co-infection might facilitate HIV acquisition.||NG, IG|
|Müller et al., 2001 ||
|Descriptive||Multiplicity Plasmodium falciparum in (FM) protective effect in moderate- transmission West Africa (IR)||61% of Riboque population cross-section PCR positive. The msp-2 genotype multiply-infected had less FM/non-FM over subsequent 3 months.||Multiple P. falciparum infections protect against super-infecting parasites; FM prevalence highest in 5–10 year-old children.||Only passive follow-up for febrile illness as possible source of bias (non-attendance, or longterm antimalarial treatment).||Investigate pathophysiology of multiplicity infections: frequently found in asymptomatic infection; protective against superinfection?||NG|
|Peterson et al., 2009 ||
|Descriptive, modelling, spatial||Small-scale spatio-temporal study in 2003 epidemic in Adama (incidence rate ratio)||Kulldorff scan: spatial cluster ≤350 m of breeding site. Other risk factors included: poor housing; proximity to vegetation; ↗temp/rain.||Benefits of identifying temporally stable clusters and risk factors to target cost-effective interventions in transmission hotspots.||Patient may have self-treated or visited other health centres; no data collection in temporary breeding sites.||Small-scale mapping or prediction models to make urban malaria control in Africa more effective.||NR|
|Sissoko et al., 2015 ||
|Descriptive, modelling, spatial||Prevalence of (a)symptomatic malaria/mosquito in 2 areas with different trans-mission/vector distribution||Anopheles density/malaria parasitaemia spatial clusters observed in dry season, sometimes associated; but high Anopheles density or parasite carriage also found outside the hotspots.||Mosquito density and parasitaemia spatial clusters in small villages (low or mesoendemic malaria transmission), best detected during dry season.||Study areas fairly small and mosquitoes possibly outreaching boundaries. Study done only one month after rainy season.||Efforts to maximize impact of interventions by targeting areas of more intense transmission likely limited by lack of suitability, since high parasite prevalence was detected outside hotspots.||NG, NGO|
|Ye et al., 2009 ||
|Descriptive, modelling||Predict FM in cohort of children ≤5 years, living in holendemic area (weather/vector data modelling)||Model represented well FM incidence for three ecological settings, including 595 person-years (n = 676) over 1 year.||Model predicted seasonal variation increasing vector abundance with increasing temperature 14 days after rainfall; in P. falciparum malaria infection incidence.||Model of parasitological data in children ≤5 years, while entire population contributing to malaria transmission.||Local-scale FM prediction beneficial to guide control; incidence depends on daily vector mortality and human parasite clearance rate, both targets of control measures.||IG, NGO|
|Dev et al., 2004 ||
|Descriptive||Fever surveys for malaria incidence and risk factors: distance-breeding sites, healthcare facility (IR/RR); vector EIR and weather data||Malaria throughout the year, more in rainy season; mainly FM; incidence higher near streams and foothills/forest areas; lower in areas < 5 km to nearest healthcare facility. EIR and % malaria among fever cases not correlated.||Areas with low-to-moderate EIR could reduce malaria significantly by using campaigns and other tools in combination with GIS methods to target intervention and save costs.||
Potential confounding: health centres likely based in plains.|
Higher P. falciparum rates during monsoon (likely due to increased temperatures promoting parasite development in vector).
|Health planners and policy makers to consider characteristics of malaria transmission and risk factors in vaccine trials and other, newer approaches for malaria control in this part of the world.||NR|
|Dhiman et al., 2013 ||
|Descriptive||Impact of altitude on monthly malaria incidence and vector density||↗ temperature coinciding with peak malaria incidence. Malaria transmission window decreased by 1 month with 400 m increase in altitude.||Reduced transmission windows and different vector composition in the highlands.||Malaria incidence based on convenience data from health centres (rather than cross-sectional study of parasitaemia population).||Highland urban areas to be considered vulnerable for malaria transmission, especially due to environmental changes.||NR|
|Lee et al., 2009 ||
|Descriptive||Malaria trends, epidemiological characteristics, local transmission and control measures (incidence)||
1983–2007 incidence ranged from 3 to 11 per 100 000 population, with deaths in 92% due to FM, and 8% vivax malaria, and a sharp decline after 1997.|
One P. knowlesi outbreak.
|> 90% of cases imported from other Asian countries; migrants often associated with larger outbreaks in relapsing P. vivax cases.||Medical practitioners to highlight risk of malaria to travellers visiting endemic areas and also to consider possibility of simian malaria.||Singapore vulnerable to reintroduction of malaria, requiring high vigilance (e.g. screening; educating on prophylaxis). Consider simian malaria if no travel history.||NR|
|Zhang et al., 2012 ||
|Descriptive, modelling, spatial||Time series regression of 2006–2010 vivax malaria, vector density, weather variables.||Strong seasonal pattern; peak during 2nd half of year; visual spatial association with average temperature; Tmax/ average humidity (1 m lag); previous month’s incidence.||Increasing An. sinensis density likely to contribute only little to malaria incidence in low transmission areas.||Predictive model does not account for human interventions since difficult to measure.||Direct public resources to control infection, rather than vector, when incidence is low. Use surveillance and vary control efforts according to incidence.||NG|
|Zhao et al., 2013 ||
|Descriptive||Epidemiology of malaria in Ningbo city. Data from case and vector surveillance, local weather and 2008 outbreak analysis||95% of cases were imported vivax malaria from domestic endemic areas, leading to limited local transmission, determined by An. sinensis vector density.||
Domestic endemic areas are important source for limited, local transmission of vivax malaria.|
Strengthen monitoring for imported malaria, ensure timely diagnosis/treatment.
|No data on floating population (might have significant impact on incidence). Study failed to show rain as precipitating factor for monthly density of female An. sinensis.||
Future studies to determine impact of floating population on dynamics of local malaria incidence.|
Focus on timely case detection, diagnosis and treatment.
|Girod et al., 2011 ||
|Descriptive||Explore malaria vectors and associated transmission dynamics||An. darlingi density high at (though variable between) 3 study locations; only few Plasmodium positive and none at the village with highest rates of human cases.||Variable relationships between malaria incidence, An. darling density, rainfall, and nearest river water levels.||
Low numbers of infected|
An. darling mosquitoes due to traps located at places away from where transmission was occurring.
|Not specifically reported; of note is the strong entomological focus of this work.||NR|
|Moreno et al., 2007 ||
|Descriptive/ longitudinal study||Determine anopheline mosquito characteristics and climate factors/ malaria incidence (prevalence)||Transmission throughout the year, with malaria prevalence between 12.5 to 21.4 per 1000 population; An. darlingii and An. marajoara important vectors, more abundant during rainy season.||Asymptomatic carriers are important reservoir of parasites for persistently high levels of transmission.||Both Anopheles species are indoor and outdoor biters but displayed strong exophagic behaviour in villages located in forested areas.||To identify bionomics of Anopheles species relevant to malaria transmission in Venezuela for planning and implementing vector control programs.||NG, IG|
|Camargo-Neves, 2001 ||
Visceral leishmaniasis (VL)|
|Spatial||Spatial analysis surveillance tools for VL; Araçatuba, São Paulo, 1998–1999 (incidence)||Heterogeneous transmission: human cases more frequent in areas with high canine rates. Vector dispersion restricted to a few houses.||Benefit of investigating vector distribution and covariates, in a field study based on house sampling.||Limited information on sampling methods; no modelling of vector density.||Need for new spatial analysis tools and redefinition of protocols for control of endemic disease in urban areas.||NR|
|Salomón et al., 2006 ||
Tegu-mentary leishmaniasis (TL)|
|Outbreak investigation||Distribution of vectors and cases and the risk factors during the 2003 TL outbreak in Bella Vista, Corrientes.||31 cases (25 ± 14 years old), m:f sex ratio = 1.8; compared to matched controls. Clusters in 2 contiguous neighbourhoods; risk higher in peri-urban (96%) than in peripheral (4%) areas.||Urban transmission risk at ecotone-border: changes in human distribution and activities, patches of secondary vegetation, peri-urban streams, rainfall, and river period floods.||Possible over-matching (controls selected in same block of cases) and small sample size: interpret risk factors associated with leishmaniasis infection with caution.||
Local-based surveillance system for entomological surveillance, diagnosis, and treatment at sentinel sites.|
Risk assessment needed for any project involving change in land use at city borders.
|Thomaz-Soccol et al., 2009 ||
Cutaneous leishmaniasis (CL)|
|Ecological||Epidemiological and entomological aspects of CL endemicity||61% (n = 100) were male rural workers ➔ extradomiciliar transmission. 29% of CL houses had antibody positive dogs. Lutzomyia intermedia s.l. most prevalent vector for Leishmania (V.) braziliensis.||Diversity of risk factors including domestic dogs, occupational hazards; deforestation to increase agricultural and pastoral areas likely important factor.||Future exploration could emphasize sample size and consider distributional studies.||Further eco-epidemiological/ biogeographical approaches needed to clarify relationships found in the present study, to guide targeted interventions.||NR|
|Uranw, 2013 ||
|Outbreak investigation||Investigate possible urban transmission of VL in Dharan town, Eastern Nepal, 2000–2008||115 VL cases (448 controls) strongly clustered in 3/19 (70%) neighbourhoods; independent risk factors include poor housing and low income.||Transmission of anthroponotic VL in urban houses; P. argentipes vector and L. donovani parasite both identified inside town.||Retrospective nature of study: possible recall bias (> 10 years) during interview in 2009.||Appropriate surveillance and control measures to be adopted not only in rural areas but also in urban areas.||IG|
|Medrano-Mercado et al., 2008 ||
|Ecological||Demographics of Chagas infections in 5–13 year-old children (n = 2218) in Cochabamba, Bolivia, 1995–1999||High seroprevalence in both South (25%) and North (19%) zones, 3× higher OR in children aged 10–13 years in NZ: higher triatomine vector infection rate (79% in South and 37% in North), but not density.||Evidence of significant exposure leading to severe disease, already early in life.||T. cruzi infection needs to be considered an urban health problem, which is not restricted to rural areas and small villages of Bolivia.||NGO|
|Salazar et al., 2007 ||
|Ecological||T. cruzi antibody seroprevalence and associated risk factors in individuals < 18 year in Veracruz, Mexico, 2000–2001||14/1544 samples confirmed positive. Risk factors included seeing reduviid bugs inside house and fissures in roof. Bug infestation index, colonization, and natural infection: 11%, 50%, and 9%, respectively.||Active vertical transmission of infection confirmed, with 0.91% seroprevalence in people <18y, which required attentative follow-up on seroprevalence at population level.||Call for improved housing, vector control measures and surveillance to be put in place.||IG|
|West Nile virus|
|Godsey et al., 2012 ||
West Nile virus (WNV)|
|Outbreak||Comparison of entomology within and outside WNV outbreak area in Arizona, 2010 (incidence)||Higher Cx. quinquefasciatus abundance and proportion of human blood meals detected in 6 outbreak vs. 6 control areas (similar demographics).||Cx. quinquefasciatus main vector for transmission, with blood meal host preference and availability increasing risk of WNV infection.||Vector blood meal host preference needs cautious interpretation: represents snapshot during outbreak; also, no bird/mammal census was conducted.||Similar multi-faceted studies during periods of increased WNV transmission to aid development of effective outbreak intervention strategies.||NG|
|Nielsen et al., 2012 ||
|Spatial modeling||Risk factors during 2006 outbreak in northern California residential community (incidence)||16/1355 (1.2%) pools of Culex. WNV positive. Significant spatial-temporal clustering of infected dead birds, positive Culex and areas of human cases.||Residential areas had warm night temperatures and WNV-positive Cx. tarsalis, likely critical factors during initiation of the outbreak.||Mosquito trapping not conducted throughout the urban landscape.||Need for spatial detection and emergency adult control in locations with WNV activity: interrupt transmission before human infections occur.||U|
|Ho et al., 2011 ||
|Outbreak study||Epidemiology of chikungunya establishing itself as endemic disease||2006–2009: 812/1072 (76%) were indigenous cases, imported from India and Malaysia; high risk group: foreign contract workers.||Initially sporadic cases imported into Singapore, then local transmission by Ae. aegypti as predominant urban vector.||Introduction of mutant viral strain (A226V) in 2008 resulted in rapid spread by Ae. albopictus as primary vector.||Identify and address demographic vulnerability (international travellers and foreign workers). Vigilance needed to detect and respond early.||NG|
|Rezza et al., 2007 ||
|Outbreak study||Identify primary source of infection and modes of transmission (risk ratio and attack rate)||Presumed index case was a symptomatic man from India visiting relatives in Italy. Laboratory confirmation was obtained for 175/205 cases (32 PCR only; 135 serology only; 8 both)||High similarity between strains found in Italy and those identified during earlier outbreak in Indian Ocean islands.||Outbreaks in non-tropical areas unexpected; need for preparedness and response to emerging infectious threats in era of globalization.||NR|
|Gould et al., 2008 ||
Yellow fever (YF)|
|Outbreak investi-gation||Confirm cause and further describe outbreak (incidence)||605 cases (163 deaths; CFR 27%); in 177 (29%) illness suggesting YF. Positive IgM-antibodies in 5/18 recently symptomatic, and 4/16 asymptomatic unvaccinated persons. Also chikungunya IgM detected in a few cases, (both ill and asymptomatic).||Serology evidence for both chikungunya and YF during outbreak. Migration and drought likely contributors to outbreak; Ae. aegypti most abundant, Ae. luteocephalus (important sylvatic YF to both monkeys and humans.||
Outbreak diagnosis and response delayed/limited. True YF and chikungunya cases likely underestimated.|
Limited clinical information and possible recall bias (retrospective study).
Need for timely/adequate response in future outbreaks; improved surveillance for YFV/other arboviruses;|
Promote YF vaccination
in endemic areas and mosquito control for both arboviruses and malaria.
|Vasconcelos et al., 2001 ||
|Outbreak study||YF outbreaks in Goias and Bahia states, 2000||77 cases reported in 8 Brazilian states with Haemagogus janthinomys as main vector. Climate changes with increasing rainfall associated with epidemic and epizootic episodes.||Populations at risk: agricultural workers, tourists, carpenters, fishermen, truck drivers.||
Unclear what other factors, besides rainfall, were associated with spread of YF in the area.|
Further virology studies needed to determine single vs. multiple genotypes.
|Need for YF vaccination in all areas recently affected, as well as Ae. aegypti control programs for the whole country, to avoid virus re-urbanization.||NR|
|Ross River virus|
|Carver et al., 2008 ||
Ross River virus (RRV)|
|Ecological||Relationship between amplifying host (M. musculus mouse) abundance and RRV notifications||Mouse index data (proxy for abundance) at Symes Farm matched with RRV incidence in rural townships (postcode) in cereal-growing Walpeup region, Southern Australia.||Mice and RRV notifications significantly related in autumn and autumn+summer when Culex annulirostris mosquitoes are abundant.||Cause and effect (mice and RRV cases) not proven: possible other factors required to enable transmission between the two hosts.||More research needed on potential causal relationship (mice and RRV) to justify targeted public health interventions to reduce mice and arboviral disease burden.||NR|
|Pham et al., 2009 ||
|Ecological||Environmental factors and human plague in Vietnam central highland plateau, 1997–2002 (risk ratio)||472 plague cases; 4 main flee and 3 rodent species. Increasing risk during dry and hot months; decreasing rainfall; increasing monthly flea index/rodent density.||Flea index, rodent density, and rainfall could be used as ecological indicators of plague risk in Vietnam’s central highlands plateau.||No wider animal sampling; hence, collected rodents in confirmed cases might not be those causing outbreak. Ecological investigation: no inference to individuals.||Plague outbreaks likely due to multiple ecological factors linked to classical reservoir and vector of bubonic plague; further research warranted.||NG|
|Souza et al., 2015 ||
Brazilian spotted fever (BSF)|
|Case-controlled multivariate regression||Risk factors associated with BSF, 2003–2013, Piracicaba River basin, São Paulo State (ORs)||Confirmed vs. discarded BSF cases were associated with increasing age, urban areas, presence of A. sculptum ticks, ticks collected from horses, presence of capybaras, and dirty pasture environment.||High incidence areas in São Paulo State reinforce trend of urbanization of BSF in peri-urban peripheries; presence of ticks and capybaras requires future field investigations.||Potential misclassification through use of discarded cases as controls: false negative cases (no timely collection) ➔ OR biased towards null hypothesis (no association). .||Control measures needed focusing on elimination of dirty pastures and prevention of human contact with areas of host and vector species, to control transmission of BSF.||NG|