This study showed that FMD is still endemic in the majority of the countries/regions in the Eastern Africa region, and that these countries/regions mainly use quarantine and post outbreak ring vaccination as FMD control strategies. Moreover, the majority of the countries in the region estimated that they were below stage 3 on the PCP-FMD, an important tool for endemic countries to progressively reduce the presence of FMD
It has previously been established that disease recognition is essential for any disease control programme
, and this is particularly relevant for FMD due to the seven FMDV serotypes causing clinically indistinguishable disease
 and to FMD being easily confused with other viral diseases
[29, 30]. The present study showed that, in Eastern Africa, the laboratory capacity for FMD, in terms of tests, equipment and skilled manpower, is still limited, and thus all reported outbreaks are not properly serotyped and characterised leading to insufficient knowledge of the regional FMD status. Other factors contributing to unclear FMD status include unwillingness of farmers to pay for diagnosis where it is not a public good (3), and failure (6) or inconsistency (7) in submitting samples to WRLFMD, Pirbright, UK, for free typing, possibly due to logistically complicated and expensive sample shipment
. Moreover, three countries had received negative results from WRLFMD, probably due to poor sample quality. Possible reasons for this could be improper handling during the 1–7 day transit from the field to the NRLs, which in 10 of the 13 countries/regions happened in cool boxes with ice packs rather than in liquid nitrogen, or poor quality sample storage in the laboratories caused by unreliable power supplies in most countries/regions. Poor sampling is less likely, since training in sampling technique had been provided in 12 countries/regions, including the three receiving negative results from WRLFMD.
Though three countries in this survey exclusively collected serum during the acute and subacute phases of outbreaks, the majority of the countries/regions followed the recommendation of the OIE
 and collected samples for demonstration of FMD viral antigen or nucleic acid during the acute phase of outbreaks; moreover, their NRLs had either antigen ELISA or PCR set up. However, success of these tests entirely depends on sample quality determined by timing and handling of samples
, and a number of outbreaks in the region were not serotyped and/or characterised. All but one of the 13 NRLs relied on antibody ELISAs, most likely because these tests are cheap and suitable for working on many samples, require lower level of biocontainment
, and neither depend on cell cultures nor on highly sensitive, expensive and service-requiring PCR-equipment
The most widely used antibody ELISAs were tests for identification of antibodies against FMDV non structural proteins (NSP), probably because they are simple and serotype–independent, and thus good screening tests for exposure to FMDV antigen
[20, 32]. These tests can also differentiate infected from vaccinated animals (DIVA test)
, however, in Eastern Africa, interpretation of NSP-test results is complicated by the frequent use of non-purified vaccines which elicit antibodies against NSPs, thereby limiting the DIVA application of these tests
[20, 34, 35]. Furthermore, antibodies against NSPs do not appear until day 8–9 after infection
, thus, to be useful, these tests should only be used for sera sampled in the late subacute and chronic phases. Moreover, as demonstrated in both small ruminants
 and cattle
, antibodies against NSPs persist for a long time, and thus NSP ELISAs do not differentiate well between present and past infection at individual level.
Eight of the 13 NRLs serotyped antibodies against FMDV using OIE-recommended tests, i.e. VNT (3), LPBE (7) and SPCE (1), and one of these also used a comparable in-house SPBE developed at the National Veterinary Institute, Lindholm, Denmark
. The limited use of VNT is most likely due to lack of cell culture facilities, and possibly also to most NRLs working at BSL 2-level. All available serotype-specific antibody tests, including VNT, show cross reactions
[39–41], which have grave implications for the control of FMD in the Eastern Africa countries that greatly rely on serology for serotyping of outbreaks. In endemic situations such cross reactions may be more pronounced due to repeated vaccination against and/or infection with one or more FMDV serotypes
[42, 43]. Moreover, test related cross-reactivity has been demonstrated for SPBEs in samples collected 1–3 weeks after experimental infection of naive calves (unpublished results) and in field sera from unvaccinated small ruminants
, and for LBPE in sera from bovines
, hence, there is a need to improve the specificity of the existing serotype-specific antibody ELISAs.
Tests for detection of FMDV can either identify the serotype directly (antigen ELISA and sequencing) or in combination with other techniques (VI and PCR). In this study, nine NRLs had the capacity for the detection of FMDV while the remaining four NRLs relied on sending samples abroad for antigen ELISA (3) or to another national laboratory for PCR (1). The most widespread test was the antigen ELISA (7), which like the serotype-specific antibody ELISAs, shows cross reactions between the FMDV serotypes
. The phasing out of the CFT in the region demonstrates a move towards more modern methods, which is also evident from the five NRLs already using or introducing PCR. Only three NRLs used VI despite this being about as sensitive as PCR
[31, 46], most likely for the same reasons as for not using VNT. Equally few NRLs (3) used real time and/or conventional RT-PCR in routine diagnosis, which accords with findings in another endemic country, Brazil, where limited use was attributed to lack of infrastructure, high cost and anticipated problems of maintaining technically complicated and service-demanding PCR machines
The majority of the NRLs (12) collaborated with foreign laboratories, including WRLFMD in UK, OVI in South Africa and FNL in Kenya, to complement their own diagnostic services. However, Rweyemamu et al. maintained that relying on foreign technical assistance to manage disease control programs may not be sustainable in developing countries, and experience from the region confirms this as the number of samples analysed is insufficient to get adequately detailed knowledge of the circulating FMDV strains to implement sufficiently efficient control measures to reduce FMD in the region. Moreover, recent evidence of the transboundary nature of FMD in Eastern Africa
 points to a need for assuming a regional approach to achieve more efficient control of FMD and progress on the PCP-FMD.
Eight NRLs had a quality management system (QMS) in place and had participated in laboratory comparisons, either inter-laboratory tests or proficiency tests. In Europe, QMSs are considered essential for diagnosis of FMD
[47, 48] including focus on competent, motivated staff, organisational management, functional equipment, process control and biosafety/biosecurity
. In the Eastern Africa region, although only one of the 13 NRLs entirely lacked SOPs, all had deficient QMS as equipment was not regularly serviced in five and not calibrated in six NRLs, and six did not monitor sample storage equipment daily. This can lead to unreliable equipment and inconsistent quality of samples, which may affect the results of the performed tests
. Moreover, none of the NRLs had been accredited for FMD diagnosis, including those in countries with vaccine production plants, and most laboratory comparisons were arranged by laboratories outside the region with a more worldwide focus. Thus, QMS efforts could be strengthened substantially by setting up a regional reference laboratory for FMD, which would arrange local comparative inter-laboratory tests with relevance for the region, encourage QMS and promote virus characterisation among the NRLs
Many diagnostic laboratories also have consultative/advisory and disease surveillance roles
, and its recommended to build a team of national experts for these tasks
. However, in this study, 12 NRLs were understaffed, disclosing a clear regional need to address capacity building in terms of laboratory space, equipment and training of professional and technical staff, as is currently being carried out by collaborative projects in Uganda and Kenya (TADEA, DANIDA-funded) and in Tanzania (SADC TADs, Wellcome Trust- funded), and as has been initiated for the entire region by FAO organizing the NRLs into a network (EARLN).
OIE recommends that FMD diagnosis is carried out in OIE class 4 facilities
[25, 50] and this is generally adhered to in the FMD-free countries. However, most Eastern African NRLs were working below BSL 3 including seven NRLs undertaking virological tests to diagnose FMD. Moreover, the recommendations for developing biosafety manuals and adopting biosafety policies
, were only implemented at five NRLs. The low level of biosafety and biosecurity measures could result in escape of FMDV as happened in the 2007 UK FMD outbreak
, and it may be speculated that a (presumably small) proportion of the outbreaks in Eastern Africa may be due to poor laboratory biocontainment.