Design, Implementation, and Evaluation of a School Insecticide-Treated Net Distribution Program in Cross River State, Nigeria.

BACKGROUND: In 2013, the World Health Organization recommended distribution through schools, health facilities, community health workers, and mass campaigns to maintain coverage with insecticide-treated nets (ITNs). We piloted school distribution in 3 local government areas (LGAs) of Cross River State, Nigeria. METHODS: From January to March 2011, all 3 study sites participated in a mass ITN campaign. Baseline data were collected in June 2012 (N=753 households) and school distribution began afterward. One ITN per student was distributed to 4 grades once a year in public schools. Obubra LGA distributed ITNs in 2012, 2013, and 2014 and Ogoja LGA in 2013 and 2014 while Ikom LGA served as a comparison site. Pregnant women in all sites were eligible to receive ITNs through standard antenatal care (ANC). Endline survey data (N=1,450 households) were collected in March 2014. Data on ITN ownership, population access to an ITN, and ITN use were gathered and analyzed. Statistical analysis used contingency tables and chi-squared tests for univariate analysis, and a concentration index was calculated to assess equity in ITN ownership. RESULTS: Between baseline and endline, household ownership of at least 1 ITN increased in the intervention sites, from 50% (95% confidence interval [CI]: 44.7, 54.3) to 76% (95% CI: 71.2, 81.0) in Ogoja and from 51% (95% CI: 35.3, 66.7) to 78% (95% CI: 71.5, 83.1) in Obubra, as did population access to ITN, from 36% (95% CI: 32.0, 39.5) to 53% (95% CI: 48.0, 58.0) in Ogoja and from 34% (95% CI: 23.2, 45.6) to 55% in Obubra (95% CI: 48.4, 60.9). In contrast, ITN ownership declined in the comparison site, from 64% (95% CI: 56.4, 70.8) to 43% (95% CI: 37.4, 49.4), as did population ITN access, from 47% (95% CI: 40.0, 53.7) to 26% (95% CI: 21.9, 29.9). Ownership of school ITNs was nearly as equitable (concentration index 0.06 [95% CI: 0.02, 0.11]) as for campaign ITNs (-0.03 [95% CI: -0.08, 0.02]), and there was no significant oversupply or undersupply among households with ITNs. Schools were the most common source of ITNs at endline and very few households (<2%) had nets from both school and ANC. CONCLUSION: ITN distribution through schools and ANC provide complementary reach and can play an effective role in achieving and maintaining universal coverage. More research is needed to evaluate the cost-effectiveness of such continuous distribution channels in combination with, or as a potential replacement for, subsequent mass campaigns.

Artemisinin combination therapy mass drug administration in a setting of low malaria endemicity: programmatic coverage and adherence during an observational study in Zanzibar.

BACKGROUND: Mass drug administration (MDA) appears to be effective in reducing the risk of malaria parasitaemia. This study reports on programmatic coverage and compliance of MDA using artemisinin-based combination therapy (ACT) in four shehias (smallest administration unit) that had been identified as hotspots through Zanzibar's malaria case notification surveillance system. METHODS: Mass drug administration was done in four shehias selected on the basis of: being an established malaria hot spot; having had mass screening and treatment (MSaT) 2-6 weeks previously; and exceeding the epidemic alert threshold of 5 cases within a week even after MSaT. Communities were sensitized and MDA was conducted using a house-to-house approach. All household members, except pregnant women and children aged less than 2 months, were provided with ACT medicine. Two weeks after the MDA campaign, a survey was undertaken to investigate completion of ACT doses. RESULTS: A total of 8816 [97.1% of eligible; 95% confidence interval (CI) 96.8-97.5] people received ACT. During post MDA surveys, 2009 people were interviewed: 90.2% reported having completed MDA doses; 1.9% started treatment but did not complete dosage; 4.7% did not take treatment; 2.0% were absent during MDA and 1.2% were ineligible (i.e. infants <2 months and pregnant women). Main reasons for failure to complete treatment were experience of side-effects and forgetting to take subsequent doses. Failure to take treatment was mainly due to fear of side-effects, reluctance due to lack of malaria symptoms and caregivers forgetting to give medication to children. CONCLUSION: Mass drug administration for malaria was well accepted by communities at high risk of malaria in Zanzibar, with high participation and completion rates. Further work to investigate the potential of MDA in accelerating Zanzibar's efforts towards malaria elimination should be pursued.

Pregnant women and infants as sentinel populations to monitor prevalence of malaria: results of pilot study in Lake Zone of Tanzania.

BACKGROUND: As malaria control interventions are scaled-up, rational approaches are needed for monitoring impact over time. One proposed approach includes monitoring the prevalence of malaria infection among pregnant women and children at the time of routine preventive health facility (HF) visits. This pilot explored the feasibility and utility of tracking the prevalence of malaria infection in pregnant women attending their first antenatal care (ANC) visit and infants presenting at 9-12 months of age for measles vaccination. METHODS: Pregnant women attending first ANC and infants nine to 12 months old presenting for measles vaccination at a non-probability sample of 54 HFs in Tanzania's Lake Zone (Mara, Mwanza and Kagera Regions) were screened for malaria infection using a malaria rapid diagnostic test (RDT) from December 2012 to November 2013, regardless of symptoms. Participants who tested positive were treated for malaria per national guidelines. Data were collected monthly. RESULTS: Overall 89.9 and 78.1 % of expected monthly reports on malaria infection prevalence were received for pregnant women and infants, respectively. Among 51,467 pregnant women and 35,155 infants attending routine preventive HF visits, 41.2 and 37.3 % were tested with RDT, respectively. Malaria infection prevalence was 12.8 % [95 % confidence interval (CI) 11.3-14.3] among pregnant women and 11.0 % (95 % CI 9.5-12.5) among infants, and varied by month. There was good correlation of the prevalence of malaria among pregnant women and infants at the HF level (Spearman rho = 0.6; p < 0.001). This approach is estimated to cost $1.28 for every person tested, with the RDT accounting for 72 % of the cost. CONCLUSIONS: Malaria infection was common and well correlated among pregnant women and infants attending routine health services. Routine screening of these readily accessible populations may offer a practical strategy for continuously tracking malaria trends, particularly seasonal variation. Positivity rates among afebrile individuals presenting for routine care offer an advantage as they are unaffected by the prevalence of other causes of febrile illness, which could influence positivity rates among febrile patients presenting to outpatient clinics. The data presented here suggest that in addition to contributing to clinical management, ongoing screening of pregnant women could be used for routine surveillance and detection of hotspots.

Taking local ownership: government and household contribution to indoor residual spraying in Zanzibar and mainland Tanzania.

BACKGROUND: While donor funding is instrumental in initiation and implementation of malaria control efforts, national government contributions are key to local ownership and sustainability. This study explored in-kind contributions of local government and households towards the cost of indoor residual spraying (IRS) interventions in Tanzania. METHODS: Data were collected through interviews with local government officials and technical teams in the IRS project. Household contribution was based on provision of water for IRS. Government contributions included government-provided warehouse and office space, vehicles, and staff labour. In-kind contributions were aggregated at the district, regional and national level. Calculations were based on proportion of total costs of IRS from 2010 to 2012. RESULTS: The mainland government provided larger amounts of in-kind contribution in absolute value (mean of US$454 200) compared to Zanzibar (US$89 163). On average, in-kind contribution was 5.5% of total costs in Zanzibar and 2.9% in mainland. The proportion of government in-kind contribution was higher in Zanzibar versus the mainland (86% vs 50%) while household contribution was higher in mainland compared to Zanzibar (50% vs 14%). CONCLUSION: Government involvement, particularly through budgetary allocations and increased in-kind contribution, needs to be encouraged for malaria control efforts to be locally owned, managed and sustained.

Use of insecticide quantification kits to investigate the quality of spraying and decay rate of bendiocarb on different wall surfaces in Kagera region, Tanzania.

BACKGROUND: Bendiocarb was introduced for the first time for Indoor Residual Spraying (IRS) in Tanzania in 2012 as part of the interim national insecticide resistance management plan. This move followed reports of increasingly alarming levels of pyrethroid resistance across the country. This study used the insecticide quantification kit (IQK) to investigate the intra-operational IRS coverage and quality of spraying, and decay rate of bendiocarb on different wall surfaces in Kagera region. METHODS: To assess intra-operational IRS coverage and quality of spraying, 104 houses were randomly selected out of 161,414 sprayed houses. A total of 509 samples (218 in Muleba and 291 in Karagwe) were obtained by scraping the insecticide samples from wall surfaces. To investigate decay rate, 66 houses (36 in Muleba and 30 in Karagwe) were selected and samples were collected monthly for a period of five months. Laboratory testing of insecticide concentration was done using IQK(TM) [Innovative Vector Control Consortium]. RESULTS: Of the 509 samples, 89.5% met the World Health Organization (WHO) recommended concentration (between 100-400 mg/m(2)) for IRS target dosage. The proportion of samples meeting WHO standards varied between Karagwe (84.3%) and Muleba (96.3%) (p < 0.001). Assessment of quality of spraying at house level revealed that Muleba (84.8%) had a significantly higher proportion of households that met the expected target dosage (100-400 mg/m(2)) compared to Karagwe (68.9%) (p < 0.001). The quality of spraying varied across different wall substrates in both districts. Evaluation of bendiocarb decay showed that the proportion of houses with recommended concentration declined from 96.9%, 93.5% and 76.2% at months one, two, and three post IRS, respectively (p-trend = 0.03). The rate of decay increased in the fourth and fifth month post spraying with only 55.9% and 26.3% houses meeting the WHO recommendations, respectively. CONCLUSION: IQK is an important tool for assessing IRS coverage and quality of spraying. The study found adequate coverage of IRS; however, residual life of bendiocarb was observed to be three months. Results suggest that in order to maintain the recommended concentrations with bendiocarb, a second spray cycle should be carried out after three months.

The impact of different sprayable surfaces on the effectiveness of indoor residual spraying using a micro encapsulated formulation of lambda-cyhalothrin against Anopheles gambiae s.s.

BACKGROUND: The type of sprayable surface impacts on residual efficacy of insecticide used in indoor residual spraying (IRS). However, there is limited data on common types of wall surfaces sprayed in Zanzibar and mainland Tanzania where IRS began in 2006 and 2007 respectively. The study investigated residual efficacy of micro-encapsulated lambda-cyhalothrin sprayed on common surfaces of human dwellings and domestic animal shelters in Zanzibar and mainland Tanzania. METHODS: An experimental hut was constructed with different types of materials simulating common sprayable surfaces in Zanzibar and mainland Tanzania. Surfaces included cement plastered wall, mud-daub, white-wash, wood, palm-thatch, galvanized iron-sheets, burnt-bricks, limestone and oil-paint. The World Health Organization (WHO) procedure for IRS was used to spray lambda-cyhalothrin on surfaces at the dose of 20-25 mg/m(2). Residual efficacy of insecticide was monitored through cone bioassay using laboratory-reared mosquitoes; Kisumu strain (R-70) of Anopheles gambiae ss. Cone bioassay was done every fortnight for a period of 152 days. The WHO Pesticide Evaluation Scheme (WHOPES) threshold (80% mortality) was used as cut-off point for acceptable residual efficacy. RESULTS: A total of 5,800 mosquitoes were subjected to contact cone bioassay to test residual efficacy of lambda-cyhalothrin. There was a statistically significant variation in residual efficacy between the different types of wall surfaces (r = 0.24; p < 0.001). Residual efficacy decreased with increasing pH of the substrate (r = -0.5; p < 0.001). Based on WHOPES standards, shorter residual efficacy (42-56 days) was found in wall substrates made of cement, limestone, mud-daub, oil paint and white wash. Burnt bricks retained the residual efficacy up to 134 days while galvanized iron sheets, palm thatch and wood retained the recommended residual efficacy beyond 152 days. CONCLUSION: The study revealed a wide variation in residual efficacy of micro encapsulated formulation of lambda-cyhalothrin across the different types of wall surfaces studied. In areas where malaria transmission is bimodal and wall surfaces with short residual efficacy comprise > 20% of sprayable structures, two rounds of IRS using lambda-cyhalothrin should be considered. Further studies are required to investigate the impact of sprayable surfaces on residual efficacy of other insecticides commonly used for IRS in Zanzibar and mainland Tanzania.

Co-occurrence and distribution of East (L1014S) and West (L1014F) African knock-down resistance in Anopheles gambiae sensu lato population of Tanzania.

OBJECTIVE: Insecticide resistance molecular markers can provide sensitive indicators of resistance development in Anopheles vector populations. Assaying these makers is of paramount importance in the resistance monitoring programme. We investigated the presence and distribution of knock-down resistance (kdr) mutations in Anopheles gambiae s.l. in Tanzania. METHODS: Indoor-resting Anopheles mosquitoes were collected from 10 sites and tested for insecticide resistance using the standard WHO protocol. Polymerase chain reaction-based molecular diagnostics were used to genotype mosquitoes and detect kdr mutations. RESULTS: The An. gambiae tested were resistance to lambdacyhalothrin in Muheza, Arumeru and Muleba. Out of 350 An. gambiae s.l. genotyped, 35% were An. gambiae s.s. and 65% An. arabiensis. L1014S and L1014F mutations were detected in both An. gambiae s.s. and An. arabiensis. L1014S point mutation was found at the allelic frequency of 4-33%, while L1014F was at the allelic frequency 6-41%. The L1014S mutation was much associated with An. gambiae s.s. (χ(2) = 23.41; P < 0.0001) and L1014F associated with An. arabiensis (χ(2) = 11.21; P = 0.0008). The occurrence of the L1014S allele was significantly associated with lambdacyhalothrin resistance mosquitoes (Fisher exact P < 0.001). CONCLUSION: The observed co-occurrence of L1014S and L1014F mutations coupled with reports of insecticide resistance in the country suggest that pyrethroid resistance is becoming a widespread phenomenon among our malaria vector populations. The presence of L1014F mutation in this East African mosquito population indicates the spreading of this gene across Africa. The potential operational implications of these findings on malaria control need further exploration.

HIV transmission risk behavior among HIV-infected adults in Uganda: results of a nationally representative survey.

OBJECTIVE: Evaluate factors associated with HIV transmission risk behavior among HIV-infected adults in Uganda. DESIGN: Cross-sectional and nationally representative study (2004-2005 Uganda HIV/AIDS Sero-Behavioral Survey) tested 18,525 adults (15-59 years old) for HIV and herpes simplex virus type 2 (HSV-2). METHODS: Data were weighted to obtain nationally representative results. Sexual-risk behavior by HIV-infected individuals was defined as unprotected sex at last encounter. Using multivariate analysis, we identified factors associated with being sexually active, knowing HIV status, and using contraception and condoms. RESULTS: Of 1092 HIV-infected respondents, 64% were female (median age was 33 years), 84% had HSV-2, and 13% reported one lifetime partner (1% of men and 23% of women). Twenty-one percent of adults knew their HIV status and 9% knew their partners'. Seventy-seven percent were sexually active, of whom 27% reported condom use at last sex. Of last unprotected sexual encounters, 84% were with spouses and 13% with steady partners. Of cohabitating persons, 40% had an HIV-negative spouse. Those who knew their HIV status were three times more likely to use a condom at last sex encounter [adjusted odds ratio (AOR), 3.0; 95% confidence intervals (CI), 1.9-4.7] and those who knew their partners' HIV status were 2.3 times more likely to use condoms (AOR, 2.3; 95% CI, 1.2-4.3). CONCLUSIONS: A minority of HIV-infected adults in Uganda knew they had HIV infection; nearly half were in an HIV-discordant relationship, and few used condoms. Knowledge of HIV status, both one's own and one's partner's, was associated with increased condom use. Interventions to support HIV-infected persons and their partners to be tested are urgently needed.

The estimated burden of HIV/AIDS in Uganda, 2005-2010.

OBJECTIVES: To estimate the burden of HIV disease in Uganda and the effect of HIV/AIDS control programmes to mitigate it. DESIGN: Mathematical modelling and projecting using surveillance and census data. METHODS: Using antenatal clinic surveillance (1986-2002) and a recent population-based survey (2004-2005) data, we modelled the adult national HIV prevalence over time (1981-2004), and kept prevalence constant at 6.4% for the years 2004-2010. Using Spectrum software and census data, we estimated the national burden of HIV disease and the effect of selected HIV-related prevention and treatment programmes. RESULTS: In 2005, we estimated that there were 135,300 new HIV infections (adult HIV incidence 0.96%), 691,900 asymptomatic prevalent infections, 88 100 AIDS cases, and 76 400 AIDS deaths. An estimated 647,000 (80%) HIV-infected adults were unaware of their infection; one third of all adult deaths were HIV related. As a result of population growth, by 2008 a similar number of people will be HIV infected (1.1 million) as during the peak of the epidemic in 1994. Although antiretroviral therapy (ART) coverage is expected to rise from 67,000 (2005) to 160,000 (2010), the number of persons needing but not receiving ART will decrease only slightly from 127,600 (2005) to 111,100 (2010). The use of single-dose in 2005 nevirapine probably averted only 4% of the estimated 20 400 vertical infections. CONCLUSION: HIV/AIDS continues to be a leading cause of adult disease and death in Uganda. Universal ART access is probably unachievable. With the absolute burden of HIV disease approaching the historic peak in the early 1990s, more effective prevention programmes are of paramount importance.