Disparities in mobile phone ownership reflect inequities in access to healthcare.

Human movement and population connectivity inform infectious disease management. Remote data, particularly mobile phone usage data, are frequently used to track mobility in outbreak response efforts without measuring representation in target populations. Using a detailed interview instrument, we measure population representation in phone ownership, mobility, and access to healthcare in a highly mobile population with low access to health care in Namibia, a middle-income country. We find that 1) phone ownership is both low and biased by gender, 2) phone ownership is correlated with differences in mobility and access to healthcare, and 3) reception is spatially unequal and scarce in non-urban areas. We demonstrate that mobile phone data do not represent the populations and locations that most need public health improvements. Finally, we show that relying on these data to inform public health decisions can be harmful with the potential to magnify health inequities rather than reducing them. To reduce health inequities, it is critical to integrate multiple data streams with measured, non-overlapping biases to ensure data representativeness for vulnerable populations.

Feasibility of satellite image and GIS sampling for population representative surveys: a case study from rural Guatemala.

BACKGROUND: Population-representative household survey methods require up-to-date sampling frames and sample designs that minimize time and cost of fieldwork especially in low- and middle-income countries. Traditional methods such as multi-stage cluster sampling, random-walk, or spatial sampling can be cumbersome, costly or inaccurate, leading to well-known biases. However, a new tool, Epicentre's Geo-Sampler program, allows simple random sampling of structures, which can eliminate some of these biases. We describe the study design process, experiences and lessons learned using Geo-Sampler for selection of a population representative sample for a kidney disease survey in two sites in Guatemala. RESULTS: We successfully used Epicentre's Geo-sampler tool to sample 650 structures in two semi-urban Guatemalan communities. Overall, 82% of sampled structures were residential and could be approached for recruitment. Sample selection could be conducted by one person after 30 min of training. The process from sample selection to creating field maps took approximately 40 h. CONCLUSION: In combination with our design protocols, the Epicentre Geo-Sampler tool provided a feasible, rapid and lower-cost alternative to select a representative population sample for a prevalence survey in our semi-urban Guatemalan setting. The tool may work less well in settings with heavy arboreal cover or densely populated urban settings with multiple living units per structure. Similarly, while the method is an efficient step forward for including non-traditional living arrangements (people residing permanently or temporarily in businesses, religious institutions or other structures), it does not account for some of the most marginalized and vulnerable people in a population-the unhoused, street dwellers or people living in vehicles.

Investigating persistent measles dynamics in Niger and associations with rainfall.

Measles is a major cause of child mortality in sub-Saharan Africa. Current immunization strategies achieve low coverage in areas where transmission drivers differ substantially from those in high-income countries. A better understanding of measles transmission in areas with measles persistence will increase vaccination coverage and reduce ongoing transmission. We analysed weekly reported measles cases at the district level in Niger from 1995 to 2004 to identify underlying transmission mechanisms. We identified dominant periodicities and the associated spatial clustering patterns. We also investigated associations between reported measles cases and environmental drivers associated with human activities, particularly rainfall. The annual and 2-3-year periodicities dominated the reporting data spectrum. The annual periodicity was strong with contiguous spatial clustering, consistent with the latitudinal gradient of population density, and stable over time. The 2-3-year periodicities were weaker, unstable over time and had spatially fragmented clustering. The rainy season was associated with a lower risk of measles case reporting. The annual periodicity likely reflects seasonal agricultural labour migration, whereas the 2-3-year periodicity potentially results from multiple mechanisms such as reintroductions and vaccine coverage heterogeneity. Our findings suggest that improving vaccine coverage in seasonally mobile populations could reduce strong measles seasonality in Niger and across similar settings.

High cholera vaccination coverage following emergency campaign in Haiti: Results from a cluster survey in three rural Communes in the South Department, 2017.

Oral cholera vaccine (OCV) has increasingly been used as an outbreak control measure, but vaccine shortages limit its application. A two-dose OCV campaign targeting residents aged over 1 year was launched in three rural Communes of Southern Haiti during an outbreak following Hurricane Matthew in October 2016. Door-to-door and fixed-site strategies were employed and mobile teams delivered vaccines to hard-to-reach communities. This was the first campaign to use the recently pre-qualified OCV, Euvichol. The study objective was to estimate post-campaign vaccination coverage in order to evaluate the campaign and guide future outbreak control strategies. We conducted a cluster survey with sampling based on random GPS points. We identified clusters of five households and included all members eligible for vaccination. Local residents collected data through face-to-face interviews. Coverage was estimated, accounting for the clustered sampling, and 95% confidence intervals calculated. 435 clusters, 2,100 households and 9,086 people were included (99% response rate). Across the three communes respectively, coverage by recall was: 80.7% (95% CI:76.8-84.1), 82.6% (78.1-86.4), and 82.3% (79.0-85.2) for two doses and 94.2% (90.8-96.4), 91.8% (87-94.9), and 93.8% (90.8-95.9) for at least one dose. Coverage varied by less than 9% across age groups and was similar among males and females. Participants obtained vaccines from door-to-door vaccinators (53%) and fixed sites (47%). Most participants heard about the campaign through community 'criers' (58%). Despite hard-to-reach communities, high coverage was achieved in all areas through combining different vaccine delivery strategies and extensive community mobilisation. Emergency OCV campaigns are a viable option for outbreak control and where possible multiple strategies should be used in combination. Euvichol will help alleviate the OCV shortage but effectiveness studies in outbreaks should be done.

Delayed second dose of oral cholera vaccine administered before high-risk period for cholera transmission: Cholera control strategy in Lusaka, 2016.

BACKGROUND: In April 2016, an emergency vaccination campaign using one dose of Oral Cholera Vaccine (OCV) was organized in response to a cholera outbreak that started in Lusaka in February 2016. In December 2016, a second round of vaccination was conducted, with the objective of increasing the duration of protection, before the high-risk period for cholera transmission. We assessed vaccination coverage for the first and second rounds of the OCV campaign. METHODS: Vaccination coverage was estimated after each round from a sample selected from targeted-areas for vaccination using a cross-sectional survey in to establish the vaccination status of the individuals recruited. The study population included all individuals older than 12 months residing in the areas targeted for vaccination. We interviewed 505 randomly selected individuals after the first round and 442 after the second round. Vaccination status was ascertained either by vaccination card or verbal reporting. Households were selected using spatial random sampling. RESULTS: The vaccination coverage with two doses was 58.1% (25/43; 95%CI: 42.1-72.9) in children 1-5 years old, 59.5% (69/116; 95%CI: 49.9-68.5) in children 5-15 years old and 19.9% (56/281; 95%CI: 15.4-25.1) in adults above 15 years old. The overall dropout rate was 10.9% (95%CI: 8.1-14.1). Overall, 69.9% (n = 309/442; 95%CI: 65.4-74.1) reported to have received at least one OCV dose. CONCLUSIONS: The areas at highest risk of suffering cholera outbreaks were targeted for vaccination obtaining relatively high vaccine coverage after each round. However, the long delay between doses in areas subject to considerable population movement resulted in many individuals receiving only one OCV dose. Additional vaccination campaigns may be required to sustain protection over time in case of persistence of risk. Further evidence is needed to establish a maximum optimal interval time of a delayed second dose and variations in different settings.

Measles seroprevalence after reactive vaccination campaigns during the 2015 measles outbreak in four health zones of the former Katanga Province, Democratic Republic of Congo.

BACKGROUND: Measles continues to circulate in the Democratic Republic of Congo, and the country suffered from several important outbreaks over the last 5 years. Despite a large outbreak starting in the former province of Katanga in 2010 and the resulting immunization activities, another outbreak occurred in 2015 in this same region. We conducted measles seroprevalence surveys in four health zones (HZ) in the former Katanga Province in order to assess the immunity against measles in children 6 months to 14 years after the 2015 outbreak. METHODS: We conducted multi-stage cluster surveys stratified by age group in four HZs, Kayamba, Malemba-Nkulu, Fungurume, and Manono. The age groups were 6-11 months, 12-59 months, and 5-14 years in Kayamba and Malemba-Nkulu, 6-59 months and 5-14 years in Manono and Fungurume. The serological status was measured on dried capillary blood spots collected systematically along with vaccination status (including routine Extended Program of Immunization (EPI), and supplementary immunization activities (SIAs)) and previous self-reported history of suspected measles. RESULTS: Overall seroprevalence against measles was 82.7% in Kayamba, 97.6% in Malemba-Nkulu, 83.2% in Manono, and 74.4% in Fungurume, and it increased with age in all HZs. It was 70.7 and 93.8% in children 12-59 months in Kayamba and Malemba-Nkulu, and 49.8 and 64.7% in children 6-59 months in Fungurume and Manono. The EPI coverage was low but varied across HZ. The accumulation of any type of vaccination against measles resulted in an overall vaccine coverage (VC) of at least 85% in children 12-59 months in Kayamba and Malemba-Nkulu, 86.1 and 74.8% in children 6-59 months in Fungurume and Manono. Previous measles infection in 2015-early 2016 was more frequently reported in children aged 12-59 months or 6-59 months (depending on the HZ). CONCLUSION: The measured seroprevalence was consistent with the events that occurred in these HZs over the past few years. Measles seroprevalence might prove a valuable source of information to help adjust the timing of future SIAs and prioritizing support to the EPI in this region as long as the VC does not reach a level high enough to efficiently prevent epidemic flare-ups.

Evaluation of the SD bioline cholera rapid diagnostic test during the 2016 cholera outbreak in Lusaka, Zambia.

OBJECTIVE: To assess the performance of the SD Bioline Cholera Ag O1/O139 rapid diagnostic test (RDT) compared to a reference standard combining culture and PCR for the diagnosis of cholera cases during an outbreak. METHODS: RDT and bacterial culture were performed on site using fresh stools collected from cholera suspected cases, and from stools enriched in alkaline peptone water. Dried stool samples on filter paper were tested for V. cholerae by PCR in Lusaka (as part of a laboratory technology transfer project) and at a reference laboratory in Paris, France. A sample was considered positive for cholera by the reference standard if any of the culture or PCR tests was positive for V. cholerae O1 or O139. RESULTS: Among the 170 samples tested with SD Bioline and compared to the reference standard, the RDT showed a sensitivity of 90.9% (95% CI: 81.3-96.6) and specificity of 95.2% (95% CI: 89.1-98.4). After enrichment, the sensitivity was 95.5% (95% CI: 87.3-99.1) and specificity 100% (95% CI: 96.5-100). CONCLUSION: The observed sensitivity and specificity were within recommendations set by the Global Task Force for Cholera Control on the use of cholera RDT (sensitivity = 90%; specificity = 85%). Although the sample size was small, our findings suggest that the SD Bioline RDT could be used in the field to rapidly alert public health officials to the likely presence of cholera cases when an outbreak is suspected.

Highly targeted cholera vaccination campaigns in urban setting are feasible: The experience in Kalemie, Democratic Republic of Congo.

INTRODUCTION: Oral cholera vaccines are primarily recommended by the World Health Organization for cholera control in endemic countries. However, the number of cholera vaccines currently produced is very limited and examples of OCV use in endemic countries, and especially in urban settings, are scarce. A vaccination campaign was organized by Médecins Sans Frontières and the Ministry of Health in a highly endemic area in the Democratic Republic of Congo. This study aims to describe the vaccine coverage achieved with this highly targeted vaccination campaign and the acceptability among the vaccinated communities. METHODS AND FINDINGS: We performed a cross-sectional survey using random spatial sampling. The study population included individuals one year old and above, eligible for vaccination, and residing in the areas targeted for vaccination in the city of Kalemie. Data sources were household interviews with verification by vaccination card. In total 2,488 people were included in the survey. Overall, 81.9% (95%CI: 77.9-85.3) of the target population received at least one dose of vaccine. The vaccine coverage with two doses was 67.2% (95%CI: 61.9-72.0) among the target population. The vaccine coverage was higher during the first round (74.0, 95%CI: 69.3-78.3) than during the second round of vaccination (69.1%, 95%CI: 63.9-74.0). Vaccination coverage was lower in male adults. The main reason for non-vaccination was to be absent during the campaign. No severe adverse events were notified during the interviews. CONCLUSIONS: Cholera vaccination campaigns using highly targeted strategies are feasible in urban settings. High vaccination coverage can be obtained using door to door vaccination. However, alternative strategies should be considered to reach non-vaccinated populations like male adults and also in order to improve the efficiency of the interventions.

Temporo-spatial dynamics and behavioural patterns of 2012 cholera epidemic in the African mega-city of Conakry, Guinea.

BACKGROUND: Cholera is endemic in Guinea, having suffered consecutive outbreaks from 2004 to 2008 followed by a lull until the 2012 epidemic. Here we describe the temporal-spatial and behavioural characteristics of cholera cases in Conakry during a three-year period, including the large-scale 2012 epidemic. METHODS: We used the national and African Cholera Surveillance Network (Africhol) surveillance data collected from every cholera treatment centre in Conakry city from August 2011 to December 2013. The prevalence of suspect and confirmed cholera cases, the case fatality ratio (CFR), and the factors associated with suspected cholera were described according to three periods: pre-epidemic (A), epidemic 2012 (B) and post epidemic (C). Weekly attack rates and temporal-spatial clustering were calculated at municipality level for period B. Cholera was confirmed by culture at the cholera national reference laboratory. RESULTS: A total of 4559 suspect cases were reported: 66, 4437, and 66 suspect cases in periods A, B and C, respectively. Among the 204 suspect cases with culture results available, 6%, 60%, and 70% were confirmed in periods A, B, and C, respectively. With 0.3%, the CFR was significantly lower in period B than in periods A (7.6%) and C (7.1%). The overall attack rate was 0.28% in period B, ranging from 0.17% to 0.31% across municipalities. Concomitantly, a cluster of cases was identified in two districts in the northern part of Conakry. At 14%, rice water stools were less frequent in period A than in period B and C (78% and 84%). Dehydration (31% vs 94% and 89%) and coma (0.4% vs 3.1% and 2.9%) were lower during period B than in periods A and C. The treatment of drinking water was less frequent in period A, while there were more reports of recent travel in period C. CONCLUSIONS: The epidemic dynamic and the sociological description of suspect cases before, during, and after the large-scale epidemic revealed that the Vibrio cholerae was already present before the epidemic. However, it appeared that infected individuals reacted differently in terms of disease severity as well as their access to treated water and travel habits. Such an in-depth description of cholera epidemics should be systematically carried out in cholera endemic settings in order to prioritize higher risk areas, identify transmission factors, and optimize preventive interventions.

The African cholera surveillance network (Africhol) consortium meeting, 10-11 June 2015, Lomé, Togo.

The fifth annual meeting of the African cholera surveillance network (Africhol) took place on 10-11 June 2015 in Lomé, Togo. Together with international partners, representatives from the 11 member countries -Cameroon, Côte d'Ivoire, Democratic Republic of Congo, Guinea, Kenya, Mozambique, Nigeria, Tanzania, Togo, Uganda, Zimbabwe- and an invited country (Malawi) shared their experience. The meeting featured three sessions: i) cholera surveillance, prevention and control in participating countries, ii) cholera surveillance methodology, such as cholera mapping, cost-effectiveness studies and the issue of overlapping epidemics from different diseases, iii) cholera laboratory diagnostics tools and capacity building. The meeting has greatly benefitted from the input of technical expertise from participating institutions and the observations emerging from the meeting should enable national teams to make recommendations to their respective governments on the most appropriate and effective measures to be taken for the prevention and control of cholera. Recommendations for future activities included collecting precise burden estimates in surveillance sites; modeling cholera burden for Africa; setting up cross-border collaborations; strengthening laboratory capacity for the confirmation of suspected cholera cases and for vaccine impact assessment in settings where oral cholera vaccine would be used; adapting cholera surveillance to concurrent issues (e.g., Ebola); and developing national cholera control plans including rationale vaccination strategies together with other preventive and control measures such as improvements in water, sanitation and hygiene (WASH).

Incidence of urinary tract infections and antibiotic resistance in the outpatient setting: a cross-sectional study.

PURPOSE: In 2012-2013, a cross-sectional survey was conducted in women visiting a general practitioner for urinary tract infection (UTI), to estimate the annual incidence of UTIs due to antibiotic-resistant Escherichia coli (E. coli). METHODS: A sampling design (stratification, stages and sampling weights) was taken into account in all analyses. Urine analyses were performed for each woman and centralised in one laboratory. RESULTS: Among 538 included women, urine culture confirmed UTI in 75.2 % of cases. E. coli represented 82.8 % of species. Among E. coli, resistance (I + R) was most common to amoxicillin [38 % (95 % confidence interval 31.1-44.5)] and to trimethoprim/sulfamethoxazole [18.1 % (12.0-24.1)]. Resistance to ciprofloxacin and cefotaxime was lower [1.9 % in both cases, (0.3-3.5)], as it was for nitrofurantoin [0.4 (0-1.0)] and fosfomycin (0). Extended-spectrum ß-lactamase (ESBL) represented 1.6 % of E. coli (0.2-2.9). Annual incidence rate of confirmed UTI was estimated at 2400 per 100,000 women (1800-3000). Incidence rates of UTI due to fluoroquinolone-resistant and ESBL-producing E. coli were estimated at 102 per 100,000 women (75-129) and at 32 (24-41), respectively. CONCLUSIONS: ESBL had been found in a community population, and even though the rate was low, it represents a warning and confirms that surveillance should continue.

Guillain-Barré Syndrome outbreak associated with Zika virus infection in French Polynesia: a case-control study.

BACKGROUND: Between October, 2013, and April, 2014, French Polynesia experienced the largest Zika virus outbreak ever described at that time. During the same period, an increase in Guillain-Barré syndrome was reported, suggesting a possible association between Zika virus and Guillain-Barré syndrome. We aimed to assess the role of Zika virus and dengue virus infection in developing Guillain-Barré syndrome. METHODS: In this case-control study, cases were patients with Guillain-Barré syndrome diagnosed at the Centre Hospitalier de Polynésie Française (Papeete, Tahiti, French Polynesia) during the outbreak period. Controls were age-matched, sex-matched, and residence-matched patients who presented at the hospital with a non-febrile illness (control group 1; n=98) and age-matched patients with acute Zika virus disease and no neurological symptoms (control group 2; n=70). Virological investigations included RT-PCR for Zika virus, and both microsphere immunofluorescent and seroneutralisation assays for Zika virus and dengue virus. Anti-glycolipid reactivity was studied in patients with Guillain-Barré syndrome using both ELISA and combinatorial microarrays. FINDINGS: 42 patients were diagnosed with Guillain-Barré syndrome during the study period. 41 (98%) patients with Guillain-Barré syndrome had anti-Zika virus IgM or IgG, and all (100%) had neutralising antibodies against Zika virus compared with 54 (56%) of 98 in control group 1 (p<0.0001). 39 (93%) patients with Guillain-Barré syndrome had Zika virus IgM and 37 (88%) had experienced a transient illness in a median of 6 days (IQR 4-10) before the onset of neurological symptoms, suggesting recent Zika virus infection. Patients with Guillain-Barré syndrome had electrophysiological findings compatible with acute motor axonal neuropathy (AMAN) type, and had rapid evolution of disease (median duration of the installation and plateau phases was 6 [IQR 4-9] and 4 days [3-10], respectively). 12 (29%) patients required respiratory assistance. No patients died. Anti-glycolipid antibody activity was found in 13 (31%) patients, and notably against GA1 in eight (19%) patients, by ELISA and 19 (46%) of 41 by glycoarray at admission. The typical AMAN-associated anti-ganglioside antibodies were rarely present. Past dengue virus history did not differ significantly between patients with Guillain-Barré syndrome and those in the two control groups (95%, 89%, and 83%, respectively). INTERPRETATION: This is the first study providing evidence for Zika virus infection causing Guillain-Barré syndrome. Because Zika virus is spreading rapidly across the Americas, at risk countries need to prepare for adequate intensive care beds capacity to manage patients with Guillain-Barré syndrome. FUNDING: Labex Integrative Biology of Emerging Infectious Diseases, EU 7th framework program PREDEMICS. and Wellcome Trust.

Risk factors for resistance in urinary tract infections in women in general practice: A cross-sectional survey.

OBJECTIVES: In 2012 and 2013, a cross-sectional survey was conducted in women visiting a general practitioner for a urinary tract infection (UTI) to i) describe the patterns of antibiotic resistance of Enterobacteriaceae involved in community-acquired UTIs and ii) identify the factors associated with UTIs due to a multi-drug-resistant Enterobacteriaceae (MDREB). METHODS: Urine analyses were performed systematically for all adult women presenting with signs of UTI. Characteristics of women with UTI due to MDREB were compared to those with UTI due to non-MDREB. Weighted logistic regressions were performed to adjust for the sampling design of the survey. RESULTS: Significant factors associated with MDREB included the use of penicillin by the patient in the last three months (OR = 3.1; [1.2-8.0]); having provided accommodation in the previous 12 months to a resident from a country at high risk for drug resistance (OR = 4.0; [1.2-15.1]); and the consumption of raw meat within the previous three months (OR = 0.3; [0.1-0.9]). CONCLUSIONS: In the community, antibiotic use and exposure to a person returning from an area with a high risk of drug resistance are associated with UTIs due to MDREB. The potentially protective role of raw meat consumption warrants further study.