Table 2.

Characteristics of Index Case and Household Contacts Associated With Household Secondary Infection Risk (SIR) Analysis Including Individuals Testing Influenza Positive in the First 8 Days, Klerksdorp and Pietermaritzburg, South Africa, 2013–2014

Characteristic	SIR,a n/N (%)	Univariate, OR (95% CI)	P	Multivariable, OR (95% CI)	P
Index case characteristics
Age group, y
<1	0/6 (0)	Not estimated		Not estimated
1–4	24/62 (39)	2.5 (1.2–5.6)	.019	3.6 (1.2–11.3)	.026
5–14	17/86 (20)	Reference	Reference	Reference	Reference
15–24	11/56 (20)	1.0 (0.4–2.7)	.987	1.9 (0.6–6.4)	.280
25–44	21/85 (25)	1.3 (0.6–3.2)	.520	8.0 (1.8–36.7)	.007
45–64	4/38 (11)	0.5 (0.1–3.0)	.426	2.6(0.3–22.6)	.395
≥65	0/0 (0)	Not included		Not included
Sex
Male	30/122 (25)	1.1 (0.6–1.9)	.706
Female	47/211 (22)	Reference	Reference
Site
Klerksdorp	49/177 (28)	1.8 (0.9–3.5)	.111
Edendale	28/156 (18)	Reference	Reference
Year
2013	19/108 (18)	Reference	.149
2014	58/225 (26)	1.6 (0.9–3.2)
HIV status
Uninfected	59/220 (27)	Reference	Reference	Reference	Reference
Infected	18/113 (16)	0.5 (0.2–1.3)	.158	0.2 (0.1–0.6)	.007
HIV status
Uninfected	59/215 (27)	Reference	Reference
Infected, receiving	13/77 (17)	0.5 (0.2–1.7)	.304
ART	5/36 (14)
Infected, not receiving ART		0.4 (0.2–0.9)	.034
Non-HIV underlying illnessd
Absent	77/326 (24)	Not estimated
Present	0/7 (0)
Currently taking treatment for tuberculosis
No	61/281 (22)	Reference	.002
Yes	9/15 (60)	5.4 (2.4–12.2)
Alcohol use
No	24/89 (27)	2.8 (0.6–8.0)	.171
Yes	5/43 (12)	Reference
Smoking
No	25/108 (34)	1.5 (0.5–7.0)	.591
Yes	4/24 (17)	Reference
Influenza type
A	71/294 (24)	Reference	.579
B	12/39 (31)	1.4 (0.4–4.7)
Influenza type and subtype
H3N2	52/219 (24)	1.4 (0.7–2.9)	.405
H1N1pdm09	12/65 (18)	Reference	Reference
A (Unsubtyped)	2/10 (20)	1.1 (0.2–7.9)	.921
B	11/39 (28)	1.7 (0.7–6.4)	.404
Influenza Ct value
<30	64/243 (26)	2.3 (1.1–4.8)	.036	2.7 (1.2–6.1)	.020
≥30	12/88 (14)	Reference	Reference	Reference	Reference
Pneumococcal colonization
Yes	37/131 (28)	0.9 (0.4–2.4)	.909
No	22/75 (29)	Reference
Viral coinfectionb
Yes	15/64 (23)	1.0 (0.5–2.2)	.947
No	62/269 (23)	Reference
Received influenza vaccine
Yes	0/5 (0)	Not estimated
No	69/298 (23)
Sore throat
Yes	21/135 (16)	0.4 (0.2–0.9)	.032
No	50/173 (29)	Reference	Reference
Myalgia
Yes	9/64 (14)	0.5 (0.2–1.4)	.176
No	62/244 (25)	Reference
Household contact characteristics
Age group, y
<1	3/6 (50)	3.3 (0.6–13.6)	.095	1.6 (0.3–5.8)	.466
1–4	13/28 (46)	2.8 (1.1–76)	.035	3.5 (1.2–10.3)	.021
5–14	19/82 (23)	Reference	Reference	Reference	Reference
15–24	11/76 (14)	0.6 (0.2–1.3)	.169	0.5 (0.2–1.2)	.115
25–44	16/84 (19)	0.8 (0.4–1.7)	.528	0.8 (0.3–1.7)	.535
45–64	9/40 (23)	1.0 (0.3–2.7)	.941	0.9 (0.3–3.0)	.832
≥65	6/17 (35)	1.8 (0.6–5.5)	.294	2.4 (0.6–9.3)	.216
Sex
Male	26/109 (24)	1.0 (0.6–1.8)	.865
Female	46/200 (23)	Reference
HIV status
Uninfected	38/140 (27)	Reference	.454
Infected	5/26 (19)	0.6 (0.2–2.1)
Non-HIV underlying illnessd
Absent	75/321 (23)	Reference	.573
Present	2/12 (17)	0.5 (0.1–2.9)
Alcohol usec
No	30/145 (21)	Reference	Reference
Yes	10/57 (18)	0.8 (0.3–1.8)	.647
Smokingc
No	33/169 (21)	Reference	Reference
Yes	7/33 (21)	1.1 (0.4–2.8)	.825
Received influenza vaccine
Yes	2/12 (17)	Reference	.734
No	54/259 (21)	1.3 (0.3–6.6)
Share a bed with the index case
Yes	29/86 (34)	2.0 (1.1–3.8)	0.021	2.7 (1.3–5.5)	0.009
No	45/226 (20)	Reference		Reference	Reference
Number of people in household
2–3	12/85 (14)	Reference	Reference	Reference	Reference
4–6	43/195 (23)	1.7 (0.7–4.0)	.213	2.3 (0.7–7.6)	.180
7+	22/53 (42)	4.3 (1.4–13.3)	.012	6.1 (1.3–28.4)	.022
People per room
<3	52/225 (23)	Reference	.358
≥3	30/108 (28)	1.3 (0.8–2.2)
Number of children aged <5 years in house
≤2	64/315 (20)	Reference	<.001
>2	13/18 (72)	10.2 (7.1–14.7)

Missing data: Out of 87 index cases included in the analysis, data were missing for alcohol use (n = 4), smoking (n = 4), current tuberculosis treatment (n = 8), received influenza vaccine (n = 8), influenza cycle threshold value (n = 1), pneumococcal colonization (n = 34), symptoms (sore throat, runny nose, headache and myalgia) (n = 7). Out of 333 household contacts included in the analysis, data were missing for contact sex (n = 24), contact HIV status (n = 167), contact alcohol use (n = 15), contact smoking (n = 15), contact influenza vaccine receipt (n = 62), contact sharing a bed with index (n = 21). There were 70 (21%) household contacts missing 1 follow-up sample, 20 (6%) missing 2 samples, and 31 (9%) missing 3 samples.

OR and P values for all variables included in the multivariable models are displayed in the 2 right hand columns of the table. No pairwise interactions were included in the multivariable models as no interaction terms were found significant. Additional factors evaluated but not found to be associated with SIR were sharing eating utensils with the index case and avoiding contact with the sick household member.

Abbreviations: ART, antiretroviral therapy; CI, confidence interval; Ct, cycle threshold; HIV, human immunodeficiency virus; OR, odds ratio.

^aSecondary infection risk (SIR) = Number of secondary cases in the household/number of exposed household contacts.

^bTest positive on polymerase chain reaction for at least 1 of parainfluenza virus types 1–3, respiratory syncytial virus, enterovirus, human metapneumovirus, adenovirus or human rhinovirus.

^cAmong individuals aged >15 years.

^dAsthma, other chronic lung disease, chronic heart disease, liver disease, renal disease, diabetes mellitus, immunocompromising conditions excluding HIV infection, neurological disease, or pregnancy. Comorbidities were considered absent in cases for which the medical records stated that the patient had no underlying medical condition or when there was no direct reference to that condition.