Socio-demographic and clinical presentations of the study participants
Samples were received for 133 patients during the three outbreaks. However, case report forms were received for 112 (84.2%) and therefore used in the statistical analysis. The average age of all subjects was 19.2±3.7 years. More than half (62, 55.3%) of the students were below 20 years. The percentage of females was slightly higher than that of males (51.8% vs. 48.2%). No history of animal contact or exposure to swine was reported by the subjects, however, contact with suspected or confirmed infected people was reported (44, 38.6%). Table 1 gives descriptive information on the study participants and study variables.
Table 1. Basic demographics of study participants and study variables
Variable
|
Description
|
Total number of participants
|
112
|
Age
|
|
Mean (SD)
|
19.2 (3.7)
|
Gender
|
|
Female
|
58 (51.8)
|
Male
|
54 (48.2)
|
Contact with infected person
|
|
No
|
70 (61.4)
|
Yes
|
44 (38.6)
|
Temperature (oC)
|
|
Mean (SD)
|
37.8 (1)
|
Fever during last 24 hours
|
|
No
|
54 (51.9)
|
Yes
|
50 (48.1)
|
Sepsis suspected
|
|
No
|
91 (87.5)
|
Yes
|
13 (12.5)
|
Malaria suspected
|
|
No
|
87 (83.7)
|
Yes
|
17 (16.3)
|
Diarrhoea
|
|
No
|
98 (94.2)
|
Yes
|
6 (5.8)
|
Headaches
|
|
No
|
31 (29.8)
|
Yes
|
73 (70.2)
|
Epidemiological picture during the three time points
For the three outbreaks that occurred, twenty (20) and 40 samples were collected during outbreak 1 and 2 respectively, while 52 samples were collected during outbreak 3.
The majority (77.6%) of influenza positive cases were recorded between November to mid-December. Accordingly, outbreaks 1 and 3 which occurred in December had higher proportion of influenza A positive cases than the second outbreak in May. Figure 2 shows the daily distribution of cases during the three outbreak periods.
Distribution of influenza by site
The majority of the participants were recruited from KUMACA (57, 50.9%) followed by KATH (40, 35.7%), whilst KNUST recorded the least number of participants (15, 13.4%). The proportion of influenza positive students in each outbreak varied, with 71.9% in KUMACA, 46.7% in KNUST, and 25.0% in KATH.
Virus detection
Overall, 58 individuals tested positive for influenza A, giving an overall prevalence of 51.8% (95% CI = 42.1-61.3) during the three outbreaks. Of these, 11 (19.0%) were confirmed to be H1N1 (pdm09) and 47 (81%) were H3N2 strains. All the 11 H1N1 (pdm09) were identified during the first outbreak as KUMACA students. The H3N2 strains were identified during the second and third outbreaks in May 2019 and December 2019 respectively.
Of the 58 subjects who tested positive for influenza A, 10 individuals who were exposed to infection with influenza A were younger (17 years; IQR = 15.2-20.2) than those not infected (21 years; IQR = 17.8-22.0). Similarly, a higher proportion of positive subjects presented with fever during the last 24 hours of infection (38; 67.9%) as compared to those without fever (18; 32.1). Headache was more common in individuals positive for Influenza A (51; 91.1%) than those negative for Influenza A (5; 8.9%).
A comparison of the various variables collected by exposure to influenza A showed significant variations (p-value <0.01) in age, contact with infected subjects, temperature, fever within the last 24 hours and headaches. Table 2 details the detection of viruses for the different variables considered under the study during the three outbreaks.
Table 2. Association between variables and influenza A detection during the three outbreaks
Variable
|
Negative
|
Positive
|
Test stat.
|
p-value
|
Total
|
54
|
58
|
|
|
Age
|
|
|
Ranksum test
|
< 0.001
|
Median (IQR)
|
21 (17.8,22)
|
17 (15, 20.2)
|
|
|
|
|
|
|
|
Gender
|
|
|
Chisq. (1 df) = 0.92
|
0.337
|
Female
|
31 (57.4)
|
27 (46.6)
|
|
|
Male
|
23 (42.6)
|
31 (53.4)
|
|
|
|
|
|
|
|
Contact with infected person
|
|
|
Chisq. (1 df) = 22.63
|
< 0.001
|
No
|
20 (37)
|
48 (82.8)
|
|
|
Yes
|
34 (63)
|
10 (17.2)
|
|
|
|
|
|
|
|
Temperature
|
|
|
Ranksum test
|
< 0.001
|
Median (IQR)
|
37.2 (37.1,37.5)
|
38.2 (37.2,38.8)
|
|
|
|
|
|
|
|
Fever during last 24 hours
|
|
|
Chisq. (1 df) = 17.34
|
< 0.001
|
No
|
36 (75)
|
18 (32.1)
|
|
|
Yes
|
12 (25)
|
38 (67.9)
|
|
|
|
|
|
|
|
Sepsis suspected
|
|
|
Chisq. (1 df) = 0.00
|
1
|
No
|
42 (87.5)
|
49 (87.5)
|
|
|
Yes
|
6 (12.5)
|
7 (12.5)
|
|
|
|
|
|
|
|
Malaria suspected
|
|
|
Chisq. (1 df) = 0.51
|
0.474
|
No
|
42 (87.5)
|
45 (80.4)
|
|
|
Yes
|
6 (12.5)
|
11 (19.6)
|
|
|
|
|
|
|
|
Diarrhoea
|
|
|
|
1
|
No
|
45 (93.8)
|
53 (94.6)
|
|
|
Yes
|
3 (6.2)
|
3 (5.4)
|
|
|
|
|
|
|
|
Headaches
|
|
|
|
< 0.001
|
No
|
26 (54.2)
|
5 (8.9)
|
|
|
Yes
|
22 (45.8)
|
51 (91.1)
|
|
|
Predictors of Influenza Positivity
In order to determine the factors and symptoms that could predict influenza A positivity, we performed a logistic regression model using variables that were significant at the univariate level of analysis. The variables inserted in the model were having contact with an infected person, age, fever and headaches. All variables were not significant except for having headache. Subjects presenting with headaches have about 5 times odds (95% CI=1.4–17.7) of being infected with influenza A infection compared to subjects without headache (p=0.014).