Capacity for vector intervention deployment (includes surveillance and control)
Of the 35 participating countries (in Africa (n = 18), Asia-Pacific (n = 14) and the Americas (n = 3)), seven were classified as “eliminating” based on their inclusion in the E2020 (n = 6) with one (Sri Lanka) certified as malaria-free in 2016, while the remaining 28 countries were categorised as controlling malaria. Hereafter the term, eliminating countries, refers to the E2020 countries and Sri Lanka. Surveys were completed between 1 November 2017 and 19 November 2018. Overall, 91 % of participating countries distributed LLINs, 31 % implemented IRS and 41 % practiced LSM at the time of the survey. For more details regarding the scale and scope of vector control and surveillance operations, see Burkot et al. .
Only 8 % of NMCPs reported having sufficient capacity to implement vector surveillance. In contrast, 57 %, 56 % and 28 % of NMCPs had the capacity to implement LLINs, IRS and LSM activities, respectively. When the capacity limitations were analysed by the country malaria status, the countries controlling malaria more frequently expressed limitations than countries that were eliminating malaria (χ2 = 47.77, df = 3, p <0.0001). In other words, the respondents from the controlling countries, more frequently expressed the existence of NMCP capacity limitations than those in eliminating countries.
Intervention capacity by programmatic inputs
Vector surveillance implementation by NMCPs was limited by governance (42 %), human resources (40 %), finance (20 %), information systems (20 %) and logistics and resources (14 %). For vector control (LLINs, IRS and LSM), the proportional responses in these same categories differed with the majority of respondents highlighting limitations in logistics and resources (53 %), followed by human resources (43 %), funding (36 %) and governance (10 %) (χ2 = 21.48, df = 4, p = 0.0002; Figure 2). Thus, governance (strategic planning) was much more limiting for vector surveillance activities than for other vector control deployment activities. Respondents did not identify information systems as a limitation for LLINs, IRS or LSM deployment.
The main subcategories of programmatic inputs that limited vector surveillance activities were strategic plans, operational staff, professional staff and the budget (Figure 3). Many of the respondents specifically noted that the strategic plan was limited in the scale or scope of vector surveillance activities. While for LLINs, IRS and LSM, the main input limiting subcategories were budget, training, equipment and supplies, transport and operational staff (Figure 3).
Resource limitations for vector control activities differed significantly at the national and subnational levels for both vector surveillance and the individual interventions used for control (χ2 = 20.21, df = 8, p = 0.009; Figure 4). Subnational (i.e. provincial and district level) malaria control programmes more frequently had shortfalls in supplies, equipment, transport, computers and office space. At the national level, supplies, office space and transport were inadequate (Figure 5).
NMCPs in 60 % (n = 21) of countries had access to an entomology laboratory (e.g. molecular (PCR) or immunology (ELISA) capacity). All countries with ELISA capabilities (31 % (n = 11)) also were PCR capable (42 % (n = 15)). Insectaries were maintained in 57 % (n = 18) of the countries in which colonies of An. arabiensis, An. funestus s.s., An. gambiae s.s. and An. merus in Africa; An. aconitus, An. balabacensis, An. dirus, An. maculatus, An. minimus, An. sinensis and An. sundaicus in the Asia-Pacific; and An. albimanus in the Americas were maintained. In addition, 31 % (n = 11) of NMCPs had semi-field facilities.
Malaria control programs collaborate with external partners or organizations such as universities, multilateral agencies or U.S. government agencies to varying degrees (Figure 5). Partners supporting laboratories differed from those supporting surveys of intervention access and use (e.g., LLIN and IRS surveys) (χ2 = 60.39, df = 7, p <0.0001) with entomological laboratories mainly supported by national research institutes (Figure 5). LLIN and IRS use and coverage surveys were primarily conducted by the Ministry of Health (Figure 5). External support for surveys in the Asia-Pacific was uncommon and significantly less than in Africa for LLINs (χ2 = 11.64, df = 4, p = 0.020), and almost significant for IRS (χ2 = 9.04, df = 4, p = 0.060).
Staffing and training for capacity building
Eliminating countries were better staffed compared with countries controlling malaria (β = -0.028, se = 0.011, p = 0.019). The median number of staff in eliminating programs was 28 per 1 million people at risk, while control countries had a median of 4 staff per 1 million people at risk (Figure 6). Within each program, staff work at the national, subnational (provincial/district), field and laboratory postings, and the ratios of these staffing allocations was not significantly different between eliminating and control programs (F(1,31) = 0.397, p = 0.258; Figure 7). However, as eliminating programs had more staff, the number of field and provincial staff was also greater than in control programs.
Reflecting the high resource requirements for elimination programs, all eliminating country respondents opined that present staff numbers were not adequate to undertake all vector surveillance and control activities, while only 55 % of control countries indicated that their programs were under-staffed (χ2 = 3.62, df = 1, p = 0.057). There was no difference in the perceived relative need for additional staff for countries controlling or eliminating malaria (F(1,31) = 1.24, p = 0.291) with both control and eliminating countries indicating a need to double the number of staff to attain sufficient staffing capacity. Overwhelmingly the greatest need for additional staff was at the subnational (provincial/district) and field positions (Figure 8). Over half of control and eliminating country staff were engaged in both malaria and dengue control (62 % for countries controlling malaria and 71 % for eliminating countries).
More eliminating countries (86 %) had an established system for staff training and capacity building compared to countries controlling malaria (50 %) but this difference was not significant (χ2 = 1.604, df = 1, p = 0.205). The primary opportunities for training or capacity building was through ad hoc on the job training (11 countries), as well as regional or national vector control courses (10 countries). A limited number of countries had mechanisms for post-graduate training (n = 3), and 1 country had a structured training-of -trainers program. Staff that participated in training were field entomologists and vector control officers (12 countries each) followed by program managers (9 countries) and provincial entomologists (10 countries).
 Note that on the job training included field based training and in house training, and vector control officers included spray personnel.