Analytical framework to evaluate and optimize the use of imperfect diagnostics to inform outbreak response: Application to the 2017 plague epidemic in Madagascar.

During outbreaks, the lack of diagnostic "gold standard" can mask the true burden of infection in the population and hamper the allocation of resources required for control. Here, we present an analytical framework to evaluate and optimize the use of diagnostics when multiple yet imperfect diagnostic tests are available. We apply it to laboratory results of 2,136 samples, analyzed with 3 diagnostic tests (based on up to 7 diagnostic outcomes), collected during the 2017 pneumonic (PP) and bubonic plague (BP) outbreak in Madagascar, which was unprecedented both in the number of notified cases, clinical presentation, and spatial distribution. The extent of these outbreaks has however remained unclear due to nonoptimal assays. Using latent class methods, we estimate that 7% to 15% of notified cases were Yersinia pestis-infected. Overreporting was highest during the peak of the outbreak and lowest in the rural settings endemic to Y. pestis. Molecular biology methods offered the best compromise between sensitivity and specificity. The specificity of the rapid diagnostic test was relatively low (PP: 82%, BP: 85%), particularly for use in contexts with large quantities of misclassified cases. Comparison with data from a subsequent seasonal Y. pestis outbreak in 2018 reveal better test performance (BP: specificity 99%, sensitivity: 91%), indicating that factors related to the response to a large, explosive outbreak may well have affected test performance. We used our framework to optimize the case classification and derive consolidated epidemic trends. Our approach may help reduce uncertainties in other outbreaks where diagnostics are imperfect.

Excess risk of subsequent infection in hospitalized children from a community cohort study in Cambodia and Madagascar.

BACKGROUND: Children in low- and middle-income countries are particularly vulnerable in the months following an initial health event (IHE), with increased risk of mortality caused mostly by infectious diseases. Due to exposure to a wide range of environmental stressors, hospitalization in itself might increase child vulnerability at discharge. The goal of this study was to disentangle the role of hospitalization on the risk of subsequent infection. METHODS: Data from a prospective, longitudinal, international, multicenter mother-and-child cohort were analysed. The main outcome assessed was the risk of subsequent infection within 3 months of initial care at hospital or primary healthcare facilities. First, risk factors for being hospitalized for the IHE (Step 1) and for having a subsequent infection (Step 2) were identified. Then, inpatients were matched with outpatients using propensity scores, considering the risk factors identified in Step 1. Finally, adjusted on the risk factors identified in Step 2, Cox regression models were performed on the matched data set to estimate the effect of hospitalization at the IHE on the risk of subsequent infection. RESULTS: Among the 1312 children presenting an IHE, 210 (16%) had a subsequent infection, mainly lower-respiratory infections. Although hospitalization did not increase the risk of subsequent diarrhoea or unspecified sepsis, inpatients were 1.7 (95% Confidence Intervals [1.0-2.8]) times more likely to develop a subsequent lower-respiratory infection than comparable outpatients. CONCLUSION: For the first time, our findings suggest that hospitalization might increase the risk of subsequent lower-respiratory infection adjusted on severity and symptoms at IHE. This highlights the need for robust longitudinal follow-up of at-risk children and the importance of investigating underlying mechanisms driving vulnerability to infection.

Influence of Sociospatial determinants on knowledge, attitudes and practices related to the plague in a population living in endemic areas in the central highlands, Madagascar.

BACKGROUND: Plague is endemic to the central highlands of Madagascar. Sporadic human cases or outbreaks can occur annually in these areas. In Madagascar, the associations between endemicity and the knowledge, attitudes and practices (KAP) of the population with regard to this disease remain poorly documented. The aim of this study was to assess KAP related to plague among the population living in the central highlands. METHODS: A cross-sectional survey was conducted in the general population from June to August 2017. Based on the reported cases of plague between 2006 and 2015 in two central highland districts, a KAP questionnaire was administered in the population. Based on the proportion of correct answers provided by respondents, KAP scores were classified into three KAP categories: low (< Mean - SD), medium (Mean ± SD) and good (> Mean + SD). Multivariate analyses were performed to determine the associations between population KAP scores related to plague and sociodemographic and epidemiological factors. In addition, individual interviews and focus groups with health professionals were conducted to assess plague perception. RESULTS: A total of 597 individuals participated in the survey; 20% (n = 119) had a good KAP score, 62% (n = 370) a medium KAP score and 18% (n = 108) a low KAP score. Among the 119 respondents with good KAP scores, 80% (n = 95) resided in Ambositra district, and 20% (n = 24) resided in Tsiroanomandidy district. According to the health professionals in the two districts, populations in endemic areas are well aware of the plague. There were significant associations (p <  0.05) of not owning a mobile phone, having no contact with a former plague case, and living in Tsiroanomandidy district with a lower KAP score. CONCLUSION: The results of the study showed the need to adapt plague control interventions to the local context to allow a better allocation of human and financial resources. Doing so would minimize delays in patient management care and increase community resilience to plague epidemics.

Factors influencing maternal healthcare seeking in a highland region of Madagascar: a mixed methods analysis.

BACKGROUND: In Madagascar, maternal mortality remains stable and high (426 deaths per 100,000 live births). This situation is mainly due to a delay or lack of use of maternal healthcare services. Problems related to maternal healthcare services are well documented in Madagascar, but little information related to maternal healthcare seeking is known. Thus, this paper aims to identify and analyze the factors that influence the utilization of maternal services, specifically, the use of antenatal care (ANC) during pregnancy and the use of skilled birth attendants (SBAs) at delivery. METHOD: We used quantitative and qualitative approaches in the study. Two communes of the Vakinankaratra region, which are located in the highlands, were the settings. Data collection occurred from October 2016 to July 2017. A total of 245 pregnant women were included and followed up in the quantitative survey, and among them, 35 participated in in-depth interviews(IDIs). Logistic regressions were applied to explore the influencing factors of antenatal and delivery healthcare seeking practices through thematic qualitative analysis. RESULTS: Among the 245 women surveyed, 13.9% did not attend any ANC visits. School level, occupation and gravidity positively influenced the likelihood of attending one or more ANC visits. The additional use of traditional caregivers remained predominant and was perceived as potentially complementary to medical care. Nine in ten (91%) women expressed a preference for delivery at healthcare facilities (HFs), but 61% of births were assisted by a skilled birth attendant (SBA).The school level; the frequency of ANCs; the origin region; and the preference between modern or traditional care influenced the use of SBAs at delivery. A lack of preparation (financial and logistics problems) and women's low involvement in decision making at delivery were the main barriers to giving birth at HFs. CONCLUSION: The use of maternal healthcare services is starting to gain ground, although many women and their relatives still use traditional caregivers at the same time. Relatives play a crucial role in maternal healthcare seeking. It would be necessary to target women's relatives for awareness-raising messages about ANC and childbirth in healthcare facilities and to support and formalize collaborations between traditional healers and biomedical caregivers.

Performance of plague rapid diagnostic test compared to bacteriology: a retrospective analysis of the data collected in Madagascar.

BACKGROUND: Plague is a highly fatal disease caused by Yersinia pestis. Late diagnosis hampers disease outcome and effectiveness of control measures, induces death and disease spread. Advance on its diagnosis was the use of lateral flow rapid diagnostic test (RDT). METHODS: We assessed the performance of the plague RDT based on Y. pestis F1 antigen detection more than 15 years after its deployment in Madagascar. We compared the RDT with bacteriological culture results, using data from plague notified cases collected during the periods for which both tests were performed independently and systematically. RESULTS: Used with bubonic plague (BP) patient samples, RDTs had a sensitivity of 100% (95% CI: 99.7-100%), a specificity of 67% (95% CI: 64-70%) with a good agreement between bacteriology and RDT results (86%; κ = 0.70, 95% CI 0.67-0.73). For pneumonic plague (PP), RDT had a sensitivity of 100% (95% CI: 91-100%) and a specificity of 59% (95% CI: 49-68%) and concordance between the bacteriological and plague RDT results was moderate (70%; κ = 0.43, 95% CI 0.32-0.55). Analysis focusing on the 2017-2018 plague season including the unprecedented epidemic of PP showed that RDT used on BP samples still had a sensitivity of 100% (95% CI: 85-100%) and a specificity of 82% (95% CI: 48-98%) with a very good agreement with bacteriology 94% (κ = 0.86, 95% CI 0.67-1); for PP samples, concordance between the bacteriological and plague RDT results was poor (61%; κ = - 0.03, 95% CI -0.17 - 0.10). CONCLUSIONS: RDT performance appeared to be similar for the diagnosis of BP and PP except during the 2017 PP epidemic where RDT performance was low. This RDT, with its good sensitivity on both plague clinical forms during a normal plague season, remained a potential test for alert. Particularly for BP, it may be of great value in the decision process for the initiation of therapy. However, for PP, RDT may deliver false negative results due to inconsistent sample quality. Plague diagnosis could be improved through the development of next generation of RDTs.

Epidemiological characteristics of an urban plague epidemic in Madagascar, August-November, 2017: an outbreak report.

BACKGROUND: Madagascar accounts for 75% of global plague cases reported to WHO, with an annual incidence of 200-700 suspected cases (mainly bubonic plague). In 2017, a pneumonic plague epidemic of unusual size occurred. The extent of this epidemic provides a unique opportunity to better understand the epidemiology of pneumonic plagues, particularly in urban settings. METHODS: Clinically suspected plague cases were notified to the Central Laboratory for Plague at Institut Pasteur de Madagascar (Antananarivo, Madagascar), where biological samples were tested. Based on cases recorded between Aug 1, and Nov 26, 2017, we assessed the epidemiological characteristics of this epidemic. Cases were classified as suspected, probable, or confirmed based on the results of three types of diagnostic tests (rapid diagnostic test, molecular methods, and culture) according to 2006 WHO recommendations. FINDINGS: 2414 clinically suspected plague cases were reported, including 1878 (78%) pneumonic plague cases, 395 (16%) bubonic plague cases, one (<1%) septicaemic case, and 140 (6%) cases with unspecified clinical form. 386 (21%) of 1878 notified pneumonic plague cases were probable and 32 (2%) were confirmed. 73 (18%) of 395 notified bubonic plague cases were probable and 66 (17%) were confirmed. The case fatality ratio was higher among confirmed cases (eight [25%] of 32 cases) than probable (27 [8%] of 360 cases) or suspected pneumonic plague cases (74 [5%] of 1358 cases) and a similar trend was seen for bubonic plague cases (16 [24%] of 66 confirmed cases, four [6%] of 68 probable cases, and six [2%] of 243 suspected cases). 351 (84%) of 418 confirmed or probable pneumonic plague cases were concentrated in Antananarivo, the capital city, and Toamasina, the main seaport. All 50 isolated Yersinia pestis strains were susceptible to the tested antibiotics. INTERPRETATION: This predominantly urban plague epidemic was characterised by a large number of notifications in two major urban areas and an unusually high proportion of pneumonic forms, with only 23% having one or more positive laboratory tests. Lessons about clinical and biological diagnosis, case definition, surveillance, and the logistical management of the response identified in this epidemic are crucial to improve the response to future plague outbreaks. FUNDING: US Agency for International Development, WHO, Institut Pasteur, US Department of Health and Human Services, Laboratoire d'Excellence Integrative Biology of Emerging Infectious Diseases, Models of Infectious Disease Agent Study of the National Institute of General Medical Sciences, AXA Research Fund, and the INCEPTION programme.

Estimating sources and sinks of malaria parasites in Madagascar.

In areas where malaria epidemiology is spatially and temporally heterogeneous, human-mediated parasite importation can result in non-locally acquired clinical cases and outbreaks in low-transmission areas. Using mobility estimates derived from the mobile phone data and spatial malaria prevalence data, we identify travel routes relevant to malaria transmission in Madagascar. We find that the primary hubs of parasite importation are in a spatially connected area of the central highlands. Surprisingly, sources of these imported infections are not spatially clustered. We then related these source locations directly to clinical cases in the low-transmission area of the capital. We find that in the capital, a major sink, the primary sources of infection are along the more populated coastal areas, although these sources are seasonally variable. Our results have implications for targeting interventions at source locations to achieve local or national malaria control goals.

Acquisition of extended spectrum beta-lactamase-producing enterobacteriaceae in neonates: A community based cohort in Madagascar.

In low and middle income countries (LMICs), where the burden of neonatal sepsis is the highest, the spread of extended spectrum beta-lactamase-producing enterobacteriaceae (ESBL-PE) in the community, potentially contributing to the neonatal mortality, is a public health concern. Data regarding the acquisition of ESBL-PE during the neonatal period are scarce. The routes of transmission are not well defined and particularly the possible key role played by pregnant women. This study aimed to understand the neonatal acquisition of ESBL-PE in the community in Madagascar. The study was conducted in urban and semi-rural areas. Newborns were included at birth and followed-up during their first month of life. Maternal stool samples at delivery and six stool samples in each infant were collected to screen for ESBL-PE. A Cox proportional hazards model was performed to identify factors associated with the first ESBL-PE acquisition. The incidence rate of ESBL-PE acquisition was 10.4 cases/1000 newborn-days [95% CI: 8.0-13.4 cases per 1000 newborn-days]. Of the 83 ESBL-PE isolates identified, Escherichia coli was the most frequent species (n = 28, 34.1%), followed by Klebsiella pneumoniae (n = 20, 24.4%). Cox multivariate analysis showed that independent risk factors for ESBL-PE acquisition were low birth weight (adjusted Hazard-ratio (aHR) = 2.7, 95% CI [1.2; 5.9]), cesarean-section, (aHR = 3.4, 95% CI [1.7; 7.1]) and maternal use of antibiotics at delivery (aHR = 2.2, 95% CI [1.1; 4.5]). Our results confirm that mothers play a significant role in the neonatal acquisition of ESBL-PE. In LMICs, public health interventions during pregnancy should be reinforced to avoid unnecessary caesarean section, unnecessary antibiotic use at delivery and low birth weight newborns.

Bacterial Infections in Neonates, Madagascar, 2012-2014.

Severe bacterial infections are a leading cause of death among neonates in low-income countries, which harbor several factors leading to emergence and spread of multidrug-resistant bacteria. Low-income countries should prioritize interventions to decrease neonatal infections; however, data are scarce, specifically from the community. To assess incidence, etiologies, and antimicrobial drug-resistance patterns of neonatal infections, during 2012-2014, we conducted a community-based prospective investigation of 981 newborns in rural and urban areas of Madagascar. The incidence of culture-confirmed severe neonatal infections was high: 17.7 cases/1,000 live births. Most (75%) occurred during the first week of life. The most common (81%) bacteria isolated were gram-negative. The incidence rate for multidrug-resistant neonatal infection was 7.7 cases/1,000 live births. In Madagascar, interventions to improve prevention, early diagnosis, and management of bacterial infections in neonates should be prioritized.

Spatial and temporal dynamics of malaria in Madagascar.

BACKGROUND: Malaria is one of the primary health concerns in Madagascar. Based on the duration and intensity of transmission, Madagascar is divided into five epidemiological strata that range from low to mesoendemic transmission. In this study, the spatial and temporal dynamics of malaria within each epidemiological zone were studied. METHODS: The number of reported cases of uncomplicated malaria from 112 health districts between 2010 and 2014 were compiled and analysed. First, a Standardized Incidence Ratio was calculated to detect districts with anomalous incidence compared to the stratum-level incidence. Building on this, spatial and temporal malaria clusters were identified throughout the country and their variability across zones and over time was analysed. RESULTS: The incidence of malaria increased from 2010 to 2014 within each stratum. A basic analysis showed that districts with more than 50 cases per 1000 inhabitants are mainly located in two strata: East and West. Lower incidence values were found in the Highlands and Fringe zones. The standardization method revealed that the number of districts with a higher than expected numbers of cases increased through time and expanded into the Highlands and Fringe zones. The cluster analysis showed that for the endemic coastal region, clusters of districts migrated southward and the incidence of malaria was the highest between January and July with some variation within strata. CONCLUSION: This study identified critical districts with low incidence that shifted to high incidence and district that were consistent clusters across each year. The current study provided a detailed description of changes in malaria epidemiology and can aid the national malaria programme to reduce and prevent the expansion of the disease by targeting the appropriate areas.