Demographic information was obtained from 6640 individuals including 4464 (67.2%) from Arjo and 2176 (32.8%) from Gambella and can be found in Table 1.
Parasite prevalence in Arjo
Of 4464 blood samples collected in Arjo, the overall malaria prevalence by microscopy was 2.0% (88/4464). The proportion of Plasmodium species by microscopy was 76.1% (67/88) P. falciparum and 23.9% (21/88) P. vivax.
One thousand seven hundred thirteen (1713) smear-negative samples were randomly selected for PCR analysis. Of the total samples analyzed by PCR, 1.2% (21/1713) carried submicroscopic infections. 90.5% (19/21) were afebrile and 9.5% (2/21) were febrile during or 48 h prior to the survey. The number of Plasmodium species in submicroscopic malaria infection was 12 P. falciparum, 6 P. vivax, and 3 P. ovale (Fig. 2).
Malaria prevalence by microscopy showed no significant difference between sugarcane irrigated 2.0% (44/2230) and in non-irrigated clusters 2.0% (44/2234) (p = 0.993).
The prevalence of malaria differs by season, in the wet season higher malaria prevalence was observed than in the dry season in both the sugarcane irrigated (wet: 3.3% (42/1253); dry: 0.2% (2/977); p = 0.0001) and non-irrigated clusters (wet: 2.9% (40/1372); dry: 0.5% (4/862); p = 0.0001). The prevalence of infection in the dry season in both sugarcane-irrigated and non-irrigated clusters was very low, 0.2% (2/977) and 0.5% (4/862) respectively (Table 2.).
Table 2
Malaria infection in irrigated and non-irrigated clusters in Arjo sugar development scheme and Gambella Rice irrigation, Ethiopia (October 2019 and March 2019)
| | Arjo | Gambella |
Season | Total examined | Microscopy Positive | Total examined | Microscopy Positive |
Irrigated | Low transmission | 977 | 2 (0.2%) | 472 | 61 (12.9%) ** |
High transmission | 1253 | 42 (3.3%) *** | 328 | 22 (6.7%) |
Total | 2230 | 44 (2.0%) | 800 | 83 (10.4%) |
Non-irrigated | Low transmission | 862 | 4 (0.5%) | 865 | 25 (2.9%) |
High transmission | 1372 | 40 (2.9%) *** | 511 | 25 (4.9%) |
Total | 2234 | 44 (2.0%) | 1376 | 50 (3.6%) |
p < = 0.01, **: p < 0.001*** |
There was significant difference in parasitemia load between symptomatic and asymptomatic Plasmodium infection by microscopy. The median parasite density/µl for symptomatic and asymptomatic infection in Arjo was: 112, IQR: 88–672 and 80, IQR: 64–96, respectively with W Mann−Whitney = 197.5, n1 = 12, n2 = 76, p = 1.51e-03) (Fig. 3.).
Parasite prevalence in Gambella
The overall malaria prevalence by microscopy was 6.1% (133/2176) Gambella. The proportion of Plasmodium species by microscopy was 75.9% (101/133) P. falciparum and 24.1% (32/133) P. vivax.
Of the 531 smear-negative sub-samples collected from Gambella and analyzed by PCR, 12.8% (68/531) were submicroscopic malaria infection. 92.6% (63/68) were afebrile and 7.4% (5/68) were febrile. Among submicroscopic infections, 48 were P. falciparum, 16 P. vivax, 3 mixed (P. falciparum and P. vivax), and 1 P. ovale (Fig. 2.).
The prevalence of malaria was 10.4% (83/800) and 3.6% (50/1376) in rice irrigated and non-irrigated clusters, respectively (p = 0.0001, Table 2.). In rice irrigated clusters prevalence was higher during the dry season than in the wet season (12.9% (61/472) vs 6.7% (22/328); p = 0.0046). In non-irrigated clusters, marginally significant difference in malaria prevalence was observed between the wet season (4.9% (25/511)) and dry season (2.9% (25/865), p = 0.055, Table 2).
There was significant difference in parasitemia load between symptomatic and asymptomatic Plasmodium infection by microscopy. The median parasite density/µl for symptomatic and asymptomatic infection in Gambella was: 960, IQR: 280–1084 and 96, IQR: 64–144, respectively with W Mann−Whitney = 235.5, n1 = 16, n2 = 117, p = 1.11e-06) (Fig. 3.).
Univariate and multivariate analysis of risk factors for Plasmodium infection
In Arjo, sex, age group, duration of stay in the area, and migrant worker status were not significantly associated with infection. Individuals who were not office workers had significantly higher malaria infection rates than office workers [OR: 5.24, 95%CI: (1.47–18.68), p = 0.010]. Also individuals who never attended school had higher malaria infection than those with higher education [OR: 3.22, 95%CI: (1.27–8.18), p = 0.014] (Table 3.). Stepwise backward elimination of independent variables with the highest p-value resulted in the final model, in which only level of education remained significant predictors of Plasmodium infection with full model AIC = 871.8 and in the final model AIC = 863.5 (Table 4.).
Table 3
Univariate analysis of individual risk factors and malaria infection by microscopy, Ethiopia (March 2019 and October 2019)
Individual risk factor | Arjo | Gambella |
Prevalence | Case | | Prevalence | Case | |
(%) | n | N | OR (95%CI) | (%) | n | N | OR (95%CI) |
Sex | Female | 1.05 | 47 | 2394 | 1 | 2.39 | 52 | 998 | 1 |
Male | 0.92 | 41 | 2070 | 1.01(0.66,1.54) | 3.72 | 81 | 1178 | 1.34(0.94,1.92) |
Age groups | < 15 | 0.74 | 33 | 1678 | 0.80(0.37,1.75) | 0.97 | 21 | 780 | 0.41(0.20,0.83)** |
15_45 | 1.05 | 47 | 2459 | 0.78(0.36,1.66) | 4.55 | 99 | 1191 | 1.34(0.74,2.43) |
> 45 | 0.18 | 8 | 327 | 1 | 0.60 | 13 | 205 | 1 |
Occupation | Office worker | 0.11 | 5 | 309 | 1 | 0.28 | 6 | 114 | 1 |
Student | 0.31 | 14 | 864 | 1.00(0.36,2.80) | 0.78 | 17 | 624 | 0.50(0.19,1.31) |
Non-school child | 0.45 | 20 | 1032 | 1.20(0.45,3.23) | 0.32 | 7 | 336 | 0.38(0.12,1.16) |
Agricultural laborer | 0.83 | 37 | 1673 | 1.38(0.54,3.53) | 2.16 | 47 | 749 | 1.20(0.50,2.88) |
Factory worker¥ | 0.16 | 7 | 523 | 0.82(0.26,2.62) | 2.02 | 44 | 248 | 3.88(1.60,9.40)*** |
Others¶ | 0.11 | 5 | 63 | 5.24(1.47,18.68)** | 0.55 | 12 | 105 | 2.32(0.84,6.43) |
Education status | ≥ Secondary | 0.11 | 5 | 588 | 1 | 1.88 | 41 | 477 | 1 |
Primary | 0.49 | 22 | 1382 | 1.89(0.71,5.00) | 1.79 | 39 | 838 | 0.52(0.33, 0.82)*** |
Preschool age€ | 0.40 | 18 | 896 | 2.39(0.88,6.47) | 0.28 | 6 | 305 | 0.21(0.09, 0.51)*** |
Never attended school | 0.96 | 43 | 1598 | 3.22(1.27, 8.18)0.01 | 2.16 | 47 | 556 | 0.98(0.63, 1.52) |
Duration of stay in the area | > 3 years | 1.50 | 67 | 3567 | 1 | 5.19 | 113 | 1975 | 1 |
1–3 years | 0.18 | 8 | 453 | 0.94(0.45,1.97) | 0.28 | 6 | 87 | 1.22(0.52,2.86) |
7–12 months | 0.18 | 8 | 209 | 2.08(0.98,4.39) | 0.28 | 6 | 66 | 1.65(0.70,3.89) |
≤ 6 months | 0.11 | 5 | 235 | 1.13(0.45,2.84) | 0.37 | 8 | 48 | 3.29(1.51,7.21)*** |
Migrant worker | No | 1.88 | 84 | 4207 | 1 | 3.91 | 85 | 1927 | 1 |
Yes | 0.09 | 4 | 257 | 0.78(0.28,2.13) | 2.21 | 48 | 249 | 5.18(3.53,7.59)*** |
p < 0.05, *: p < = 0.01, **: p < 0.001***, ¥: includes sugarcane and rice plantation workers, ¶: drivers, miners and fisherman, €: children less than 7 years old not yet start formal education |
Table 4
Multivariate analysis of individual risk factor and malaria infection by microscopy in Arjo and Gambella, Ethiopia (March 2019 and October 2019)
Individual risk factor | Prevalence | Case | |
(%) | n | N | AOR (95%CI) |
Arjo | Education status | | | | |
≥ Secondary | 0.11 | 5 | 588 | 1 |
Primary | 0.49 | 22 | 1382 | 1.89(0.71-5.00) |
Preschool age€ | 0.40 | 18 | 896 | 2.39(0.88–6.47) |
Never attended school | 0.96 | 43 | 1598 | 3.22(1.27–8.18)** |
Gambella | Education status | | | | |
≥ Secondary | 1.88 | 41 | 477 | 1 |
Primary | 1.79 | 39 | 838 | 0.74(0.46–1.19) |
Preschool age€ | 0.28 | 6 | 305 | 0.19(0.07–0.53)*** |
Never attended school | 2.16 | 47 | 556 | 1.73(1.06–2.83)** |
Duration of stay in the area | | | | |
> 3 years | 5.19 | 113 | 1975 | 1 |
1–3 years | 0.28 | 6 | 87 | 2.16(0.85–5.52) |
7–12 months | 0.28 | 6 | 66 | 5.55(1.90-16.18)*** |
< 6 months | 0.37 | 8 | 48 | 4.72(1.83–12.14)*** |
Migrant worker | | | | |
No | 3.91 | 85 | 1927 | 1 |
Yes | 2.21 | 48 | 249 | 4.66(3.02–7.18)*** |
p < 0.05, * :p < = 0.01,**: p < 0.001***, €: children less than 7 years old not yet start formal education |
In Arjo, household risk factors such as irrigation status, family size, roof material, floor material, wall materials, number of sleeping rooms in the house, number of LLINs in the house, and IRS were not associated with infection. In contrast, season and LLINs utilization were risk factors for malaria infection. Wet season [OR: 5.99, 95%CI (2.33, 15.41), p = 0.0001] and not utilizing LLIN [OR: 8.55, 95%CI (3.00, 24.34), p = 0.0001] were significantly associated with infection (Table 5.). Stepwise backward elimination of independent variables with the highest p-value resulted in the final model, in which season and LLIN utilization significant predictors of infection with full model AIC = 317 and in the final model AIC = 300 (Table 6.).
Table 5
Univariate analysis of household risk factor by microscopy, Ethiopia (March 2019 and October 2019)
Household risk factor | Arjo | Gambella |
Prevalence | Case | | Prevalence | Case | |
(%) | n | N | OR (95%CI) | (%) | n | N | OR (95%CI) |
Irrigation Status | Non-irrigated | 1.16 | 21 | 822 | 1 | 3.95 | 27 | 335 | 1 |
Irrigated | 0.94 | 17 | 986 | 0.67(0.35,1.28) | 8.92 | 61 | 349 | 2.42(1.49,3.91)*** |
Season | Dry | 0.28 | 5 | 847 | 1 | 7.16 | 49 | 389 | 1 |
Wet | 1.83 | 33 | 961 | 5.99(2.33,15.41)*** | 5.70 | 39 | 295 | 1.06(0.67,1.66) |
Family size | 1_5 | 1.27 | 23 | 1226 | 1 | 9.21 | 63 | 528 | 1 |
> 5 | 0.83 | 15 | 582 | 1.38(0.72,2.67) | 3.65 | 25 | 156 | 1.41(0.85,2.33) |
Roof material | Corrugated iron | 1.44 | 26 | 1215 | 1 | 8.77 | 60 | 417 | 1 |
Grass (thatch) | 0.66 | 12 | 593 | 0.94(0.47,1.89) | 4.09 | 28 | 267 | 0.70(0.43,1.12) |
Floor material | Cement | 0.11 | 2 | 94 | 1 | 5.41 | 37 | 208 | 1 |
Mud/earth | 1.99 | 36 | 1714 | 0.99(0.23,4.16) | 7.46 | 51 | 476 | 0.55(0.35,0.88)** |
Wall material | Mud | 1.77 | 32 | 1446 | 1 | 7.31 | 50 | 467 | 1 |
Corrugated iron | 0.33 | 6 | 362 | 0.74(0.31,1.79) | 5.56 | 38 | 217 | 1.77(1.12,2.79)** |
Number of sleeping rooms | ≥ Three | 0.28 | 5 | 230 | 1 | 0.73 | 5 | 47 | 1 |
Two | 1.11 | 20 | 799 | 1.16(0.43,3.11) | 2.19 | 15 | 139 | 1.02(0.35,2.96) |
One | 0.72 | 13 | 779 | 0.76(0.27,2.17) | 9.94 | 68 | 498 | 1.33(0.51,3.48) |
Number of LLIN per household | ≥Three | 0.28 | 5 | 271 | 1 | 0.73 | 5 | 47 | 1 |
Two | 0.50 | 9 | 531 | 0.92(0.30,2.76) | 1.17 | 8 | 156 | 0.45(0.14,1.46) |
One | 0.39 | 7 | 523 | 0.72(0.23,2.29) | 5.26 | 36 | 265 | 1.32(0.49,3.56) |
Not available | 0.94 | 17 | 483 | 1.94(0.71,5.32) | 5.70 | 39 | 216 | 1.85(0.69,4.98) |
LLIN usage per HH | Always | 0.22 | 4 | 773 | 1 | 6.29 | 43 | 438 | 1 |
Sometime | 0.17 | 3 | 307 | 1.90(0.42,8.53) | 0.44 | 3 | 12 | 3.06(0.80,11.74) |
Never | 1.71 | 31 | 728 | 8.55(3.00,24.34)*** | 6.14 | 42 | 234 | 2.01(1.27,3.18)** |
IRS sprayed the past 12 months | Yes | 0.61 | 11 | 680 | 1 | 5.85 | 40 | 291 | 1 |
No | 1.49 | 27 | 1128 | 1.49(0.73,3.03) | 7.02 | 48 | 393 | 0.87(0.56,1.37) |
p < 0.05, * : p < = 0.01,**: p < 0.001*** |
Table 6
Multivariate logistic regression of household risk factors between irrigated and non-irrigated clusters by microscopy in Arjo and Gambella, Ethiopia (March 2019 and October 2019)
Household risk factor | Prevalence | Case | |
(%) | n | N | AOR (95%CI) |
Arjo | Season | | | | |
Dry | 0.28 | 5 | 847 | 1 |
Wet | 1.83 | 33 | 961 | 15.90(6.01,42.04)*** |
LLIN Usage per household | | | | |
Always | 0.22 | 4 | 773 | 1 |
Sometime | 0.17 | 3 | 307 | 10.15(2.15,47.88)*** |
Never | 1.71 | 31 | 728 | 22.32(7.74,64.34)*** |
Gambella | Irrigation Status | | | | |
Non-irrigated | 3.95 | 27 | 335 | 1 |
Irrigated | 8.92 | 61 | 349 | 2.43(1.45, 4.07)*** |
Family size | | | | |
1_5 | 9.21 | 63 | 528 | 1 |
> 5 | 3.65 | 25 | 156 | 2.30(1.30, 4.09)*** |
Number of LLIN per household | | |
≥Three | 0.73 | 5 | 47 | 1 |
Two | 1.17 | 8 | 156 | 0.49(0.15, 1.61) |
One | 5.26 | 36 | 265 | 1.53(0.53, 4.45) |
Not available | 5.70 | 39 | 216 | 1.88(0.65, 5.39) |
p < 0.05, * : p < = 0.01,**: p < 0.001*** | | | |
In Gambella, being a rice plantation worker [OR: 3.88, 95%CI: (1.60–9.40), p = 0.0026], age group < 15 [OR: 0.41, 95%CI: (0.20–0.83), p = 0.013], primary educational status [OR: 0.52, 95%CI: (0.33–0.82), p = 0.0046], being a preschool child [OR = 0.21, 95%CI: (0.09–0.51), p = 0.0005], < 6 months [OR: 3.29, 95%CI: (1.51–7.21), p = 0.0028], being a migrant worker [OR: 5.18, 95%CI: (3.53–7.59), p = 0.0001] were associated with Plasmodium infection (Table 3.). Sex was not significantly associated with Plasmodium infection. In the final model, level of education, duration of stay in the area and migrant worker significant predictors of Plasmodium infection with full model AIC = 925.3 and in the final model AIC = 920.5 (Table 4.)
In Gambella, household risk factors such as family size, roof material, number of sleeping rooms in the house, number of LLIN per household and a house that was treated with IRS in the preceding 12 months and season were not associated with Plasmodium infection in the univariate analysis. Whereas, irrigation [OR: 2.42, 95%CI, (1.49–3.91), p = 0.0003], mud/earth floor material [OR: 0.55, 95%CI, (0.35–0.88), p = 0.0118], corrugated iron wall material [OR: 1.77, 95%CI (1.12–2.79), p = 0.0142] and no LLIN utilization [OR: 2.10, 95%CI, (1.27–3.18), p = 0.0029] were significantly associated with infection (Table 5.). Stepwise backward elimination of independent variables with the highest p-value resulted in the final model with full model AIC = 514.7 and in the final model AIC = 505.1. Households located in irrigated clusters and HH with a family size of > 5 were significant predictors of Plasmodium infection. Even if, number of LLIN per HH in the final model it was not significantly associated to HH Plasmodium infection (Table 6.).