Tifariti

Seroepidemiological survey of Rift Valley fever
and Peste des Petits Ruminants in the Sahrawi
Arab Democratic Republic, Western Sahara
Di Nardo
1
A,
2
D,
3
S,
3
S,
Rossi
Mohammed Lamin Saleh Mohammed Lejlifa
4
5
5
6
Sabatini M , Di Gennaro A , Savini G , Thrusfield MV
Hatri Hamdi
3
S,
1 Institute
for Animal Health, Ash Road, Pirbright, Woking, Surrey, GU24 0NF, United Kingdom
2 Africa 70 (International Non-Governmental Organization), Rabouni, Algeria
3 Dirección Nacional de Veterinaria, Ministerio de Salud Pública, República Árabe Saharaui Democrática, Rabouni, Algeria
4 Faculty of Veterinary Medicine, University of Perugia, Via S. Costanzo, 06126 Montebello (PG), Italy
5 Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale”, Via Campo Boario, 64100 Teramo, Italy
6 University of Edinburgh, Veterinary Clinical Sciences, The Royal (Dick) of Veterinary Studies, Easter Bush Veterinary Centre, Roslin, Midlothian, EH25 9RG, United Kingdom
INTRODUCTION
METHODS
The seroprevalence of Rift Valley fever (RVF)
and peste des petits ruminants (PPR) in
sheep, goat and camel populations of the
Sahrawi Arab Democratic Republic (SADR),
Western Sahara (WS) was estimated for the
first time. Sera were collected from a field
study conducted in the whole SADR territory
between March and April 2007 (Fig. 1).
The survey was based on two-stage cluster sampling of 23 clusters (Fig. 2), considering a finite
population as per the 2007 census (Tab. 1). The sample size was calculated assuming the expected
prevalence (Pexp) = 15%, with the absolute precision (d) = ±5%, and the between-cluster variance (Vc) =
0.0039 for RVF, where for PPR a Pexp = 18%, with d = ±5%, and Vc = 0.0026 was assumed. A total of 982
samples were collected from sheep (n=461), goats (n=463) and camels (n=58). The sample size
obtained was tested in CSurvey (UCLA, Los Angeles, USA) to assess the acceptability of the
parameters. Animals were
randomly
selected
using
Survey Toolbox 1.0b (ACIAR
Canberra, AU), generating
tables of random numbers,
according to the total
population data for each
cluster. Sera samples were
tested using the inhibitionELISA for RVF and the
Fig. 2 – Sampling site location in the study area. The red symbols
show the Wilayas’ sampling sites, whereas the green ones shows
Tab.1 – SADR animal population 2007 census data by region and
competitive-ELISA
for
PPR.
the Liberated Territories sampling sites.
animal species.
4
Sheep
3
2
Camel
TOT
1
WILAIA
23
27 Febrero
Smara
Awserd
El Aaiun
Dahkla
TOT Wilayas animal population
MILITARY REGION
Bir Lehlou
Tifariti
Mehaires
Mijek
Agwanit
Dougaj
6
15
16
9
14 10
13 12
11
7
5
8
18
17
20
21
19
22
Fig. 1 – Study area (Sahrawi Arab Democratic Republic).
Goat
355
7628
4675
7424
3729
23811
378
7537
5035
7657
3518
24125
12
92
89
140
125
458
745
15257
9799
15221
7372
48394
845
9940
5700
18000
2990
1395
600
11500
3800
8000
2610
1014
1052
5090
3000
12000
2860
1715
2497
26530
12500
38000
8460
4124
TOT ‘liberated territories’ animal
population
38870
27524
25717
92111
TOT SADR Animal Population
62681
51649
26175
140505
RESULTS
Low RVF seroprevalence was reported (1%; 95%CI 0.9% to 1.1%), where 11 of 982 samples tested positive;
whereas, 264 of 976 animals were found positive for PPR (29.3%: 95%CI 28.9% to 29.75) (Tab. 2). Analysis of
serological results by species revealed a significant increase of seroprevalence in goats (RVF, p=0.0022; PPR,
p<0.0001) and older animals (RVF, p=0.02; PPR, p<0.001). A high prevalence of RVF was reported in Tifariti Tab.2 – Summary of results for the survey in SADR.
region
(5%; 95%CI 3.7% to 6.6%) (Fig. 5), where spatial analysis revealed a high prevalence in the Tifariti site 10 (7.7%; 95%CI 5.2% to 10.2%) and in the
r
Mehaires site 15 (7.1%; 95%CI 3.5% to 12%) (Fig. 3, 6), reporting higher finding in goats in both of clusters, 15.4% (95%CI 11.6% to 20.1%) and 14.3%
(95%CI 7.1% to 23.4%), respectively.
Positive animals/
No sampled
Fig. 3 – Rift Valley fever spatial distribution.
Fig. 4 – Box plot of RVF seroprevalence for all species sampled. PI
value of ≥36.1 indicates a positive result in camels; ≥38.4 a
positive result in sheep; ≥41.4 a positive result in goats.
Seroprevalence
(95% CI)
Within-region
prevalence range
Within-site
prevalence range
RVF
11/982
1% (0.9% to 1.1%)
0% to 5%
0% to 7.7%
PPR
264/976
29.3% (28.9% to 29.7%)
0% to 45.2%
0% to 50%
Fig. 5 – Box plot of RVF seroprevalence for all regions sampled. PI
value of ≥36.1 indicates a positive result in camels; ≥38.4 a
positive result in sheep; ≥41.4 a positive result in goats.
Fig. 6 – Box plot of RVF seroprevalence for all clusters sampled. PI
value of ≥36.1 indicates a positive result in camels; ≥38.4 a
positive result in sheep; ≥41.4 a positive result in goats.
A high prevalence of PPR was reported in the Wilaya (30.1%; 95%CI 29.7% to 30.6%), Bir Lehlou (33.5%; 95%CI 30.3% to 36.9%), Tifariti (45.2%; 95%CI 42%
to 48.5%) and Agwanit (31%; 95%CI 28.4% to 33.7%) regions (Fig. 8), where spatial analysis revealed high prevalence in 27 Febrero site 1 (34.5%; 95%CI
31.1% to 38%), Awserd site 3 (40%; 95%CI 42% to 44%), El Aaiun site 4 (32.7%; 95%CI 32% to 33.5%), Bir Lehlou site 5 (33.3%; 95%CI 29.2% to 37.7%),
Tifariti site 9 (39.3%; 95%CI 33.8% to 45.1%), Tifariti site 10 (49.6%; 95%CI 45.5% to 53.7%) and Agwanit site 20 (34.5%; 95%CI 31.2% to 38%) (Fig. 9).
Fig. 10 – Location of spatial clusters of PPR detected by the spatial
discrete Poisson model. The rank order correspond with the detailed
information for each cluster (C) given in Table 3.
Fig. 7 – Box plot of PPR seroprevalence for all species sampled. PI
value of ≥50 indicates a positive result
Fig. 9 – Box plot of RVF seroprevalence for all clusters sampled. PI
value of ≥50 indicates a positive result.
Fig. 8 – Box plot of PPR seroprevalence for all regions sampled. PI
value of ≥50 indicates a positive result.
Cluster No
Significant (p=0.0283) local spatial clustering of PPR seroprevalence was identified by using the
spatial discrete Poisson model, which includes the Bir Lehlou site 8, Tifariti site 9, Tifariti site 10, and
Tifariti site 11 (Fig. 10). The centre of the cluster was located in Tifariti region, reporting a PPR
relative risk of 1.753 higher in Tifariti than in other regions (Tab. 3).
Location
Region
Radius (km) Observed-to-expected ratio Relative Risk
P-value
1
Bir Lehlou (site 8)
Tifariti (site 9)
Tifariti (site 10)
Tifariti (site 11)
Tifariti
27.10
1.619
1.753
0.00283
2
27 Febrero (site 1)
Awserd (site 3)
El Aaiun (site 4)
Wilayas
49.10
1.241
1.371
0.2874
3
Agwanit (site 19)
Agwanit
0.00
1.806
1.861
0.3780
Tab.3 – Spatial clusters detected by the spatial discrete Poisson model for PPR in the SADR.
CONCLUSIONS
Data about RVF and PPR prevalence and distribution are reported for WS for the first time. Although the overall prevalence reported for RVF is not
alarming, the presence of clusters with high prevalence deserves more attention because it suggests RVF activity in the Maghreb region, where the high
prevalence reported for PPR is likely to suggest its endemic occurrence in WS. The high seroprevalence found in Tifariti may indicates a hot spot for RVF
and PPR risk from this region.