Epidemiological profile of imported malaria in Anhui Province, 2012–2019
There were 941 cases of imported malaria reported in Anhui province from 2012 to 2019. Cross-checks of the patient identifiers revealed that 6 patients had 2 clinical episodes due to different malarial parasites after returning to China. These patients were reported twice due to different dates of onset. PCR confirmed they were mixed infections (all were P. falciparum co-infected with P. ovale). For the purpose of the study, the six cases were recorded as mixed infections, according to the date of the second clinical episode, and the number of imported malaria cases was updated to 935. P. falciparum was the dominant species, accounting for 733 cases (78.4%), followed by P. ovale (115 cases, 12.30%), P. malariae (38 cases, 4.06%), P. vivax (35 cases, 3.74%) and mixed infection (14 cases, 1.50%). Among 14 cases of mixed infection, there were 12 cases of P. falciparum co-infected with P. ovale, one of P. falciparum with P. malariae, and one of P. malariae with P. ovale. The proportion of imported cases due to P. ovale spp. reached a peak in 2015 (19.69%) and 2018 (19.35%), and increased from 2012 to 2018 (χ2 = 9.626, p = 0.002, excluded 2019) (Fig. 1).
Origin of imported cases of P. ovale spp.
All 128 patients infected with P. ovale spp., including mixed infections, were imported from 16 countries in Africa. A total of 127 (99.22%) P. ovale spp. cases were successfully confirmed by PCR; one patient did not have enough blood available for PCR and was only confirmed as P. ovale spp. by microscopy. The top four countries of origin for these infections were Equatorial Guinea (24, 18.75%), Angola (22, 17.19%), Nigeria (18, 14.06%) and Cameroon (12, 9.38%) (Table 1). One case, infected in Angola, was found to be P. o. curtisi co-infected with P. o. wallikeri. Therefore, 129 P. ovale isolates were included in the analysis. Except for Ethiopia and Uganda (only one case reported), P. o. curtisi and P. o. wallikeri were simultaneously detected in all countries. The proportion of P. o. curtisi ranged from 39.13% to 66.67% (Fig. 2).
Table 1
Origin of imported P. ovale spp. in Anhui, 2012–2019.
Country | P. ovale curtisi | | P. ovale wallikeri |
N | % | | N | % |
Equatorial Guinea | 13 | 19.12 | | 11 | 18.33 |
Nigeria | 11 | 16.18 | | 7 | 11.67 |
Angolaa | 9 | 13.24 | | 13 | 21.67 |
Cameroon | 6 | 8.82 | | 6 | 10.00 |
Ghana | 5 | 7.35 | | 3 | 5.00 |
Gabon | 4 | 5.88 | | 3 | 5.00 |
Zambia | 3 | 4.41 | | 2 | 3.33 |
Mozambique | 3 | 4.41 | | 2 | 3.33 |
Ivory Coast | 3 | 4.41 | | 4 | 6.67 |
Congo(Brazzaville) | 3 | 4.41 | | 2 | 3.33 |
Malawi | 2 | 2.94 | | 1 | 1.67 |
Guinea | 2 | 2.94 | | 0 | 0.00 |
Congo(Kinshasa) | 2 | 2.94 | | 3 | 5.00 |
Uganda | 1 | 1.47 | | 0 | 0.00 |
Ethiopia | 1 | 1.47 | | 0 | 0.00 |
Benin | 0 | 0.00 | | 3 | 5.00 |
Total | 68 | 100.00 | | 60 | 100.00 |
a One case, infected in Angola, did not have enough blood available to discriminate subspecies of P. ovale spp. |
The incidence of P. ovale spp. in the main countries of origin
Because the number of people returning from countries of infectious origin were variable, a high proportion of P. ovale spp. did not necessarily reflect high incidence. In this study, we utilized the number of returnees (Anhui Statistical Yearbook, http://tjj. ah.gov.cn/ssah/qwfbjd/tjnj/index.html) from countries of infectious origin to estimate P. ovale spp. and P. falciparum incidence rates in the four main countries (Cameroon, Angola, Equatorial Guinea and Nigeria). Using this method, the estimated average annual incidence rates of P. ovale spp. in Cameroon, Angola, Equatorial Guinea and Nigeria were 2.48%, 0.30%, 1.87%, and 1.70%, respectively. The average annual incidence rates of P. falciparum in Cameroon, Angola, Equatorial Guinea and Nigeria were 9.90%, 2.86%, 8.86%, and 8.11%, respectively (Table 2).
Table 2
The incidence rates of P. ovale spp. and P. falciparum in the population returning from Africa to Anhui province
Year | P. ovale spp. | P. falciparum |
Cameroon (%, n/Na) | Angola (%, n/N) | Equatorial Guinea(%, n/N) | Nigeria (%, n/N) | Cameroon (%, n/N) | Angola (%, n/N) | Equatorial Guinea(%, n/N) | Nigeria (%, n/N) |
2012 | 0.00(0/64) | 0.00(0/3648) | 0.50(2/397) | 0.00(0/208) | 3.13(2/64) | 0.52(19/3648) | 4.03(16/397) | 2.88(6/208) |
2013 | 2.3(1/44) | 0.14(2/1427) | 1.53(5/326) | 3.85(3/78) | 25.00(11/44) | 3.22(46/1427) | 10.74(35/326) | 23.08(18/78) |
2014 | 0.00(0/60) | 0.51(3/593) | 2.65(7/265) | 2.65(4/176) | 11.67(7/60) | 4.72(28/593) | 12.08(32/265) | 7.39(13/176) |
2015 | 2.15(2/93) | 1.02(5/489) | 5.38(6/112) | 0.00(0/205) | 1.08(1/93) | 7.16(35/489) | 10.71(12/112) | 5.37(11/205) |
2016 | 4.73(4/85) | 1.21(3/249) | 0.00(0/74) | 3.41(5/147) | 12.94(11/85) | 16.47(41/249) | 5.41(4/74) | 7.48(11/147) |
2017 | 4.76(2/42) | 1.00(1/101) | 3.88(2/52) | 5.00(5/100) | 11.90(5/42) | 12.87(13/101) | 11.54(6/52) | 10.00(10/100) |
2018 | 5.71(1/18) | 3.69(6/163) | 18.18(1/6) | 0.00(0/87) | 16.67(3/18) | 5.52(9/163) | 66.67(4/6) | 13.79(12/87) |
Average annual incidence (%) | 2.48 | 0.30 | 1.87 | 1.70 | 9.90 | 2.86 | 8.86 | 8.11 |
a n was the number of reported imported cases, and N was the number of people returning from the infectious origin countries in the same year. |
Epidemiological characteristics of P. o. curtisi and P. o. wallikeri
Of 128 patients infected with P. ovale, 113 were single-species infections, as determined by PCR confirmation. Sixty-two (62/113, 54.87%) cases were P. o. curtisi and the remaining fifty-one were P. o. wallikeri (51/113, 45.13%). Among the other 15 cases, there were 13 cases of mixed infections, 1 case with co-infection of P. o. curtisi and P. o. wallikeri, and 1 case without subspecies confirmation. The median latency period for P. o. curtisi (59.50 days, IQR: 23.0–192.75) was longer than P. o. wallikeri (34 days, IQR: 12–112.50), but the difference was not significant (P = 0.070) (Fig. 3, Table 3). There were also no significant differences between the P. o. curtisi and P. o. wallikeri groups in terms of sex, age, occupation or previous malaria (Table 3). In the present study, we detected 13 cases of P. ovale spp. co-infecting with other species, predominantly with P. falciparum (12 cases). One case involved co-infection of P. ovale and P. malariae. The rate of mixed infection was 10.16% (13/128). Of all 12 mixed P. ovale spp. and P. falciparum cases, six had only one clinical episode and six had two clinical attacks whereby the first attack was due to P. falciparum and the second was due to P. ovale spp.. The intervals between the 2 clinical episodes were 33, 56, 127, 204, 295, and 1279 days, for the six patients. Among patients with one clinical episode, 4 were infected with P. o. wallikeri and two were infected with P. o. curtisi; among patients with two clinical episodes, 3 cases were of P. o. wallikeri and 3 were P. o. curtisi.
Table 3
Epidemiologic characteristics of imported two P. ovale subspecies.
Variables | P. ovale curtisi(n = 62) N (%) | P. ovale wallikeri(n = 51) N (%) | P value |
Sexa | | | |
Male | 61 (98.4) | 51 (100.0) | 1.000 |
Female | 1 (1.6) | 0 (0.0) | |
Age(years), mean(SD) | 43.84 ± 9.21 | 43.53 ± 7.82 | 0.850 |
Occupationa | | | |
Worker | 51 (82.3) | 46 (90.2) | 0.463 |
Waiter | 7 (11.3) | 3 (5.9) | |
Other | 4 (6.5) | 2 (3.9) | |
Previous malaria | | | |
Yes | 44 (71.0) | 41 (80.4) | 0.248 |
No | 18 (29.0) | 10 (19.6) | |
Duration of stay overseasa,b | | | |
≤ 30 | 0 (0.0) | 1 (2.3) | 0.489 |
≤ 180 | 6 (11.8) | 7 (16.3) | |
≤ 365 | 10 (19.6) | 11 (25.6) | |
> 365 | 35 (68.6) | 24 (55.8) | |
Latency period (days)c, median (IQR) | 59.50(23.0–192.75) | 34 (12–112.50) | 0.070 |
aDifferences in proportions were tested by Fisher’s exact test. |
b94 cases (83.19%; 94/113) with available information of duration of stay overseas. |
cThe time elapsed, in days, between arrival in China and onset of disease was defined as the latency period. Cases showing onset of symptoms before arrival were not included in the analysis. A case with a record of 1533 days is excluded, as the authenticity is questionable. |
Diagnosis of imported P. ovale spp.
To analyze the diagnostic data on imported P. ovale, P. falciparum was used as the control group in the study. The median interval of P. ovale from arrival in China to symptom onset was 49 (IQR: 16.5 − 169.5) days; from onset to the first medical visit was 1 (IQR: 0 − 3) day; and from the first medical visit to diagnosis was 1 (IQR: 0 − 3) day. The median interval of P. falciparum from arrival in China to symptom onset was 6 (IQR: 2 − 10) days; from onset to the first medical visit was 1 (IQR: 0 − 2) day; and from the first medical visit to diagnosis was 1 (IQR: 0–2) day. Differences in median intervals between the two species showed that P. ovale spp. had significantly longer interval from return to onset than P. falciparum (Z = -12.947, p < 0.001).
The parasite detection rates for P. ovale spp. and P. falciparum were comparable at 93.04% and 93.59%, respectively. However, the rates of species identification were significantly different between P. ovale spp. and P. falciparum (χ2 = 255.841, p < 0.001). Of the 115 P. ovale cases, only 20% of the cases (23/115) had an accurate species identification (Table 4), while 32.17% (37/115) were misdiagnosed as P. vivax and 6.96% (8/115) as P. falciparum, by the reporting units. In the remaining cases, species identification was not attempted. Among 733 P. falciparum cases, 86.90% of the cases (637/733) had accurate species identification, while 11.87% (87/733) could not provide the results of species identification. Only 9 cases were misdiagnosed as other malaria species or mix-infections.
Table 4
The diagnosis of imported P. ovale and P. falciparum in Anhui, 2012–2019.
Variables | P. ovale (n = 115) N (%) | P. falciparum (n = 733) N (%) | P value |
Time from return to the onset (days), Median (IQR)a | 49 (16.5, 169.5) | 6(2,10) | ༜ 0.001 |
Time from onset to first medical visit (days), Median (IQR) | 1(0,3) | 1(0,2) | 0.479 |
Time from first medical visit to diagnosis (days), Median (IQR) | 1(0,3) | 1(0,2) | 0.091 |
Parasite detectionb,c | 107(93.04) | 686(93.59) | 0.826 |
Species identificationc,d | 23 (20.00) | 636 (86.77) | ༜ 0.001 |
aCases with onset of symptoms before arrival were not included in the analysis. |
bThe results of diagnosis came from the reporting units. |
cParasite detection: calculated according to whether or not a parasite was detected [13]. |
dSpecies identification: calculated based on the number of correctly identified species [13]. |