During the study period spanning 8 influenza seasons from January 1st, 2008 till June 30th, 2016, a total of 11288 patients presented to AUBMC with a respiratory illness for which they were tested for influenza infection. Out of all patients tested, 1829 patients (16%) had a positive test result for influenza; 542 influenza cases (29.6%) were identified during the 2009 influenza pandemic. Of the 1829 positive cases, 1529 (84%) were caused by influenza A, 242 cases (13%) by influenza B, and 55 cases (3%) had co-infection with both types (type was not reported in three cases). During the 2009 pandemic, 97.6% of the infections were caused by influenza A and 2% were co-infected. However, subtyping was done for only 35 cases during the pandemic, of which 28 cases were caused by influenza A (H1N1)pdm09.
Demographic and clinical characteristics of the patients
Of the 1829 cases, 916 (50.1%) cases were female and 72 cases were pregnant; 24% of female patients in the childbearing age (Table 1). Most of the patients resided in Beirut and Mount Lebanon governorates (73%) reflecting the population served by our center. Patients were divided into 7 age groups: the highest proportion (22.5%) was for adults between 19 and 50 years old, while only 9.5% were below 2 years old. At least one underlying comorbidity was present in 45.8% of cases reviewed and 31% had two or more comorbidities. The most common underlying chronic disease among adults was cardiovascular diseases (33%), malignancy (15%) and diabetes mellitus (14.8%). However, in the younger age groups (< 19 years) the most common underlying chronic diseases were asthma (16%) and malignancy (10.4%). Data on the immunization status against influenza was available for 456 cases; 104 of these patients (22.8%) had received influenza vaccine in the past 12 months at the time of diagnosis with influenza infection.
The most common presenting symptoms were respiratory (96%) with cough constituting 91.5 % of these symptoms. Fever was present in 89% of cases; 78% of febrile patients with data on highest temperature recorded had high-grade fever ≥ 39°C. Gastrointestinal symptoms were present in 33.7% including vomiting (47%), diarrhea (32%) and abdominal pain (29.6%). Myalgia was reported in 31 % of patients. Other less common symptoms (13.6%) included conjunctivitis, otitis, neurological events, confusion and dizziness (Table S1).
Table 1. Demographic and clinical characteristics of laboratory-confirmed influenza patients
|
Number (%)
|
Gender (n=1829)
|
|
Male
|
913 (49.9)
|
Female
|
916 (50.1)
|
Age groups (n=1829)
|
|
[0-2 years
|
173 (9.5)
|
[2- 5 years
|
240 (13.1)
|
[5-10 years
|
338 (18.5)
|
[10-19 years
|
328 (17.9)
|
[19- 50 years
|
411 (22.5)
|
[50-65 years
|
164 (9.0)
|
≥ 65 years
|
175 (9.6)
|
Presence of underlying comorbidities (n=1303)
|
597 (45.8)
|
Number of underlying comorbidities (n=1303)
|
|
None
|
706 (54.2)
|
One
|
412 (31.6)
|
Multiple comorbidities
|
185 (14.2)
|
Underlying comorbidities (n=1303)
|
|
Cardiovascular disease
|
222 (17.0)
|
Chronic pulmonary disease
|
54 (4.1)
|
Asthma
|
160 (12.3)
|
Immunosuppressive state (other than malignancy)
|
19 (1.5)
|
Diabetes Mellitus
|
96 (7.4)
|
Renal disease
|
52 (4.0)
|
Neuromuscular disorders
|
12 (0.9)
|
Neurological disorders
|
58 (4.5)
|
Malignancy
|
164 (12.6)
|
Hemoglobinopathy
|
30 (2.3)
|
Time interval between symptom onset and speicmen collection (n=1368)
|
|
0-2 days
|
882 (64.5)
|
3-4 days
|
338 (24.7)
|
>4 days
|
148 (10.8)
|
Influenza Vaccine (n=456)
|
104 (22.8)
|
Pregnant among childbearing age females (N=322)*
|
72 (22.4)
|
* This denominator represents the number of females in childbearing age for whom data on pregnancy was available.
Complications secondary to influenza infection were captured when available. Almost 24% (295/1245) of patients developed at least one complication. Pneumonia, confirmed by the presence of infiltrates or consolidation on radiological examination, was the most common complication (241/295;81.6%). Chest X-ray was done in 233 of these cases and detected the presence of pneumonia in 218 cases (90.4%); the remaining 15 patients (6.6%) had a clear chest X-ray but had abnormal findings on computed tomography (CT). In eight patients, only a CT was done initially and detected the presence of pneumonia. However, the treating physician did not request an imaging study in 1095 subjects (60 %). When imaging was requested, pneumonia was detected in 32.6% of cases (241/739). The highest rate of pneumonia cases was in the older age groups: ≥65 years old (27.4%) and 50-65 years (19.1%) (P-value<0.001, using Pearson's Chi-square test). Other reported complications were croup (3 cases), shock (20 cases) of which 20% were below 2 years and 25% were above 65 years of age, sepsis (23 cases) of which 39% were ≥65 years, myocardial infarction (4 cases), renal failure (12 cases, all adults), acute respiratory distress syndrome (16 cases, 25% were < 2 years and 37.5% were between 19 and 50 years old), encephalopathy (6 cases), hemorrhagic pneumonia (1 case), stroke (2 cases), seizures (7 cases) and liver failure (5 cases; all adults).
Moderate to severe influenza was encountered in 591 cases (42%) with a significant P-value (<0.001) between different age groups. Almost half of the patients below 5 years (50.5%) and above 50 years (51.6%) had moderate to severe disease while around a third (28%) of those 10 to 50 years did (Table 2). When comparing percentages of moderate to severe disease within each type of influenza detected, influenza B infection was associated with a slightly higher incidence (49.8%) of moderate to severe disease compared to influenza A infection (40.9%) (P-value = 0.054).
Table 2. Severity parameters of laboratory-confirmed influenza cases according to different age groups
|
Total
n (%)
|
[0-2[
n (%)
|
[2-5[
n (%)
|
[5-10[
n (%)
|
[10-19[
n (%)
|
[19-50[
n (%)
|
[50-65[
n (%)
|
≥65
n (%)
|
p-value
|
Moderate to severe disease (N=1399)
|
591 (42.2)
|
80 (51.9)
|
103 (52.0)
|
95 (45.5)
|
60 (28.0)
|
91 (29.3)
|
73 (49.7)
|
89 (53.6)
|
<0.001
|
Hospitalization rate (N=1787)
|
475 (26.6)
|
57 (34.5)
|
37 (16.2)
|
24 (7.3)
|
24 (7.5)
|
103 (25.5)
|
99 (60.4)
|
131 (74.9)
|
<0.001
|
ICU admission (N=1767)
|
70 (4.0)
|
13 (8.0)
|
6 (2.6)
|
1 (0.3)
|
2 (0.6)
|
14 (3.5)
|
11 (6.7)
|
23 (13.2)
|
<0.001
|
Oxygen supplementation (N=1613)
|
121 (7.5)
|
13 (9.4)
|
8 (4.1)
|
6 (2.1)
|
3 (1.1)
|
17 (4.4)
|
25 (15.5)
|
49 (28.3)
|
<0.001
|
Mechanical ventilation (N=1767)
|
35 (2.0)
|
8 (4.9)
|
5 (2.2)
|
1 (0.3)
|
2 (0.6)
|
8 (2.0)
|
4 (2.4)
|
7 (4.0)
|
0.002†
|
Complication rate (N=1245)
|
295 (23.7)
|
45 (34.4)
|
36 (22.1)
|
20 (11.6)
|
16 (9.6)
|
51 (16.9)
|
52 (35.6)
|
75 (45.5)
|
<0.001
|
Mortality rate in hospitalized patients (N=469)
|
18 (3.8)
|
2 (3.6)
|
2 (5.7)
|
0 (0.0)
|
0 (0.0)
|
3 (2.9)
|
5 (5.1)
|
6 (4.6)
|
0.909†
|
Duration of ICU stay in days, mean (±SD) (N=70)
|
9.2 (±8.9)
|
8.2 (±5.2)
|
8.7 (8.8)
|
22*
|
5.5 (±6.4)
|
12.6 (±12.3)
|
7.7 (±6.6)
|
8.2 (9.3)
|
0.533
|
Duration of hospital stay in days, mean (±SD) (N=473)**
|
7.0 (±8.8)
|
8.5 (±10.3)
|
8.1 (±8.7)
|
7.2 (13.1)
|
6.6 (±5.0)
|
6.5 (±8.7)
|
6.4 (±6.8)
|
7.0 (±8.8)
|
0.788
|
Pearson's Chi-Square test was used (no expected count less than 5).
†Fisher’s exact test was used when expected count was less than 5
*Only one patient was admitted to ICU in this age group
** Two patients were excluded; one stayed in 113 days and the other stayed for 262 days related to their underlying medical condition (Neuromuscular disease and cerebral palsy)
To note that a different denominator was used to calculate the rate of each variable. The denominator depended on the number of cases with available data on the studied variable.
Influenza type A, B and Co-infection
Influenza detection was done using RDT or RT-PCR except for 2 cases where diagnosis was made by serology because the clinical picture was highly suggestive of influenza infection despite negative RDT and PCR. These 2 cases had an Influenza A IgG dilution titers of 1:5120 (negative <1;10). Most of the cases were diagnosed by RDT (1546 samples, 84.5 %) while 144 cases (7.9%) were detected by PCR. In 137 cases (7.5%), both RDT and PCR were used at the same time for detection. In these cases, PCR was likely requested for subtyping. Subtyping was done for 245 cases only, of which 138 cases were H1N1pdm09, 7 cases were H1N1 Brisbane, 61 cases were H3N2 and 3 cases were co-infected with H1N1pdm09 and H3N2.As for influenza B cases, 20 were Victoria lineage, 14 were Yamagata and 2 were Shanghai lineage.
Characteristics of patients with laboratory-confirmed influenza by virus type are illustrated in (Table 3). There was a significant variation in the incidence of each type of influenza virus through the different influenza seasons and among different age groups. Mean hospital stay duration was significantly higher in those who were co-infected with both types of influenza (17 days vs 7 days). Co-infection with both A and B viruses was significantly associated with higher incidence of ICU admission. Antibiotic and antiviral prescription was similar for both types and for co-infection. Death was equally reported for both types of influenza viruses (P-value=1).
Table 3. Characteristics of laboratory confirmed influenza cases by virus type
Characteristics
|
Influenza A, N=1529
|
Influenza B, N=242
|
Co-infection A&B, N=55
|
p-value
|
|
n (%)
|
n (%)
|
n (%)
|
|
Season (N=1826)
|
|
|
|
<0.001¥
|
Jan 2008-Jun 2008
|
12 (66.7)
|
6 (33.3)
|
0 (0.0)
|
|
2008- 2009
|
47 (78.3)
|
6 (10.0)
|
7 (11.7)
|
|
2009- 2010
|
529 (97.6)
|
2 (0.4)
|
11 (2.0)
|
|
2010- 2011
|
43 (58.9)
|
25 (34.2)
|
5 (6.8)
|
|
2011- 2012
|
92 (84.4)
|
2 (1.8)
|
15 (13.8)
|
|
2012- 2013
|
93 (73.8)
|
27 (21.4)
|
6 (4.8)
|
|
2013- 2014
|
183 (89.3)
|
17 (8.3)
|
5 (2.4)
|
|
2014- 2015
|
153 (64.3)
|
82 (34.5)
|
3 (1.3)
|
|
2015- 2016
|
377 (82.9)
|
75 (16.5)
|
3 (0.7)
|
|
Age group (N=1826)
|
|
|
|
|
[0-2 years [
|
134 (8.8)
|
28 (11.6)
|
11 (20.0)
|
0.039¥
|
[2- 5 years [
|
205 (13.4)
|
25 (10.3)
|
9 (16.4)
|
|
[5-10 years [
|
289 (18.9)
|
44 (18.2)
|
5 (9.1)
|
|
[10-19 years [
|
283 (18.5)
|
37 (15.3)
|
7 (12.7)
|
|
[19- 50 years [
|
340 (22.2)
|
56 (23.1)
|
14 (25.5)
|
|
[50-65 years [
|
138 (9.0)
|
25 (10.3)
|
1 (1.8)
|
|
≥ 65 years
|
140 (9.2)
|
27 (11.2)
|
8 (14.5)
|
|
Gender (N=1826)
|
|
|
|
0.992
|
Male
|
764 (50.0)
|
121 (50.0)
|
27 (49.1)
|
|
Female
|
765 (50.0)
|
121 (50.0)
|
28 (50.9)
|
|
Underlying medical conditions (N=1300)
|
|
|
|
|
Any
|
511 (47.8)
|
89 (45.2)
|
15 (44.1)
|
0.74
|
Cardiovascular disease
|
179 (16.7)
|
35 (17.8)
|
8 (23.5)
|
0.563
|
Chronic pulmonary disease
|
51 (4.8)
|
3 (1.5)
|
0 (0.0)
|
0.054
|
Asthma
|
135 (12.6)
|
20 (10.2)
|
5 (14.7)
|
0.543
|
Immunosuppressive state (other than malignancy)
|
16 (1.5)
|
3 (1.5)
|
0 (0.0)
|
1.000*
|
Diabetes Mellitus
|
81 (7.6)
|
13 (6.6)
|
2 (5.9)
|
0.938*
|
Renal disease
|
43 (4.0)
|
8 (4.1)
|
1 (2.9)
|
1.000*
|
Neuromuscular disorders
|
8 (0.7)
|
4 (2.0)
|
0 (0.0)
|
0.204*
|
Neurological disorders
|
47 (4.4)
|
8 (4.1)
|
3 (8.8)
|
0.395*
|
Malignancy
|
136 (12.7)
|
27 (13.7)
|
1 (2.9)
|
0.213*
|
Hemoglobinopathy
|
27 (2.5)
|
3 (1.5)
|
0 (0.0)
|
0.749*
|
Antibiotic prescription (N=1201)
|
395 (40.0)
|
87 (47.8)
|
12 (37.5)
|
0.134
|
Antiviral prescription (N=1191)
|
684 (69.8)
|
119 (65.7)
|
17 (56.7)
|
0.192
|
Radiologically confirmed pneumonia (N=739)
|
189 (31.5)
|
45 (39.1)
|
7 (29.2)
|
0.260
|
Hospitalization rate (N=1784)
|
381 (25.4)
|
84 (36.2)
|
10 (18.5)
|
0.001
|
Characteristics
|
Influenza A,
N=381
|
Influenza B,
N=84
|
Co-infection A&B, N=10
|
p-value
|
|
n (%)
|
n (%)
|
n (%)
|
|
Progression of illness and in-hospital outcome
|
|
|
|
|
Moderate to severe disease (N=470)
|
224 (59.4)
|
59 (71.1)
|
7 (70.0)
|
0.129*
|
ICU admission (N=475)
|
51 (13.4)
|
15 (17.9)
|
4 (40.0)
|
0.042*
|
Oxygen therapy (N=470)
|
87 (23.0)
|
20 (24.4)
|
4 (40.0)
|
0.399*
|
Mortality (N=470)
|
15 (4.0)
|
3 (3.7)
|
0 (0.0)
|
1.000*
|
|
Mean (± SD)
|
Mean (± SD)
|
Mean (± SD)
|
p-value
|
Hospital stay duration (N=473)π
|
6.8 (±8.4)
|
6.8 (±6.7)
|
16.7 (±23.8)
|
0.02
|
|
|
|
|
|
|
Percentages are expressed within each influenza type, except for the season variable in which percentages are expressed within each season
Pearson's Chi-Square test was used (no expected count less than 5).
Independent Sample-t test is used to compare means.
* Fisher’s exact test was used when expected count was less than 5.
¥ Monte Carlo estimate was used when the data set was too large to compute exact significance
π Two patients were excluded; one stayed in 113 days and the other stayed for 262 days related to their underlying medical condition (Neuromuscular disease and cerebral palsy)
Figure (1) shows the monthly distribution of influenza A and B through the various seasons. Influenza A was mostly detected during the months of January (37.4%) and February (34.8%). Another peak, however, is seen in November (51.2%) mostly during the 2009 pandemic. Influenza B tended to peak later during the month of March (32.1%). Over the different seasons, influenza infections started by the month of November and peaked by January except during the 2009 pandemic when a first wave was seen during the summer months and peaked in August and then another significant wave that started in October and peaked in November.
Progression of illness and in-hospital outcome of laboratory-confirmed influenza patients
A total of 475 cases (26.6%, N= 1787) required hospital admission for management. Table (2) shows a comparison of the course of hospital stay among the different age groups with laboratory-confirmed influenza. The highest hospitalization rates were among the elderly ≥ 65 years (74.3%), older adults aged 50-65 years (60.4%), followed by children <2 years old (34.5%). The mean duration of hospital stay was 7 days and there was no statistically significant difference between the different age groups (P-value = 0.788). In total, 70 (4.0%) patients required ICU admission for management of influenza-related complications. The highest rates of these admissions were at both age extremes; 13.2 % of the elderly ≥ 65 years and 8.0% of children < 2 years with a significant P-value <0.001 between the different age groups. However, there was no significant difference in the mean duration of ICU stay for all age groups (P-value = 0.533). Oxygen supplementation was required for 121 patients (7.5%) of which 35 cases ultimately required mechanical ventilation. There was a significant statistical difference among different age groups concerning oxygen supplementation and mechanical ventilation with the highest percentages for oxygen supplementation requirement among the elderly ≥ 65 years (28.3%) while the highest percentages for mechanical ventilation was among children < 2 years (4.7%).
Since follow up was not documented for outpatients, outcome was only reported in 469 hospitalized patients, out of which 18 cases (3.8%) died during their hospital stay. Influenza type A was the detected type in 15 out of the 18 mortality cases, while Influenza type B was detected in the remaining 3 cases. Eleven cases (69%) were older than 50 years and three cases were young adults. Only 4 cases were in the pediatric age group (<5 years old) with influenza A detected in all four cases. All cases with fatal outcome had at least one comorbid condition except for one case who was one month old. This case was also co-infected with respiratory syncytial virus. Another 7 cases also had laboratory proven viral co-infections (Cytomegalovirus (CMV), Epstein Barr Virus (EBV) and Herpes Simplex Virus Type-1 (HSV-1); five of whom had an underlying malignancy, one had an underlying neurological disease and one was a healthy pregnant female.
Table (4) shows risk factors for increased mortality among hospitalized patients due to influenza infection. A statistically significant P-value was found for moderate to severe disease, radiologically-confirmed pneumonia, and presence of viral or bacterial co-infection. Other statistically significant risk factors included neurological disorder (0.002) and malignancy (0.043). There was no statistically significant difference in mortality among different age groups or types of influenza infection detected. However, when multivariable logistic regression was done, only the presence of a proven viral co-infection, bacterial co-infection and underlying neurological diseases were found to be independent risk factors for increased mortality (Table S2).
Table 4. Risk factors for mortality in cases of laboratory proven influenza in hospitalized patients
Characteristic
|
Recovery,
n (%) N=451
|
Death,
n (%) N=18
|
p-value
|
Age group
|
|
|
0.909*
|
[0-2 years [
|
54 (12.0)
|
2 (11.1)
|
|
[2- 5 years [
|
33 (7.3)
|
2 (11.1)
|
|
[5-10 years [
|
24 (5.3)
|
0 (0.0)
|
|
[10-19 years [
|
24 (5.3)
|
0 (0.0)
|
|
[19- 50 years [
|
99 (22.0)
|
3 (16.7)
|
|
[50-65 years [
|
93 (20.6)
|
5 (27.8)
|
|
≥ 65 years
|
124 (27.5)
|
6 (33.3)
|
|
Gender
|
|
|
0.611
|
Male
|
228 (50.6)
|
8 (44.4)
|
|
Female
|
223 (49.4)
|
10 (55.6)
|
|
Presence of underlying conditions
|
318 (70.5)
|
16 (88.9)
|
0.091
|
Underlying medical conditions
|
|
|
|
Cardiovascular disease
|
161 (35.7)
|
8 (44.4)
|
0.449
|
Chronic pulmonary disease
|
43 (9.5)
|
1 (5.6)
|
1.000*
|
Asthma
|
39 (8.6)
|
2 (11.1)
|
0.665*
|
Immunosuppressive state (other than malignancy)
|
15 (3.3)
|
1 (5.6)
|
0.471*
|
Diabetes Mellitus
|
75 (16.6)
|
3 (16.7)
|
1.000*
|
Renal disease
|
45 (10.0)
|
3 (16.7)
|
0.414*
|
Neuromuscular disorders
|
9 (2.0)
|
1 (5.6)
|
0.326*
|
Neurological disorders
|
34 (7.5)
|
6 (33.3)
|
0.002*
|
Malignancy
|
101 (22.4)
|
8 (44.4)
|
0.043*
|
Hemoglobinopathy
|
15 (3.3)
|
0 (0.0)
|
1.000*
|
Type of influenza
|
|
|
1.000*
|
Influenza A
|
363 (80.5)
|
15 (83.3)
|
|
Influenza B
|
79 (17.5)
|
3 (16.7)
|
|
Co-infection A&B
|
9 (2.0)
|
0 (0.0)
|
|
Moderate to Severe disease (N=466)
|
269 (60.0) †
|
18 (100.0)
|
0.001
|
Radiologically confirmed pneumonia (N=418)
|
173 (43.3)a
|
17 (94.4)
|
<0.001
|
Proven viral co-infection b
|
13 (2.9)
|
7 (38.9)
|
<0.001*
|
Proven bacterial co-infection c
|
94 (20.8)
|
11 (61.1)
|
0.001*
|
Pearson's Chi-Square test was used (no expected count less than 5).
* Fisher’s exact test was used when expected count was less than 5
†3 patients in this group had missing data
a 51 patients had no imaging results
b Proven viral co-infection was defined by a positive PCR for other viruses whether from respiratory or other specimens.
c Proven bacterial co-infection was defined by the presence of a positive culture.
Antibiotic and antiviral use
Antibiotics were prescribed for 494 patients (41% of those with available data N=1202) with laboratory-confirmed influenza infection. Only 102 (20.6%) of those who received antibiotics had a proven bacterial co-infection. A proven bacterial co-infection was considered in the presence of a positive culture. In fact, pneumonia cases accounted for 49% of antibiotic prescription though not all of these cases were proven to be bacterial.
Antivirals were prescribed for 821 patients (68.8% of those with available data N=1193) with laboratory-confirmed influenza. Oseltamivir was prescribed in 807 subjects while zanamivir was prescribed in the remaining 14 patients. Table (S3) describes the different characteristics of the patients who received antiviral treatment. The majority of elderly patients and children below 2 years of age received antiviral therapy. There was a statistically significant higher percentage of antiviral prescription among inpatients, pregnant females and those with comorbid conditions.
Influenza hospitalization rates
Overall influenza-associated hospitalization rate reached 26.6%. The annual influenza-associated hospitalization rate was lowest during the 2009 pandemic season (6.8%) and trended up gradually to reach the highest during 2016 (41%) (Figure 2). In addition, the monthly influenza-associated hospitalization rate was as low as 3% in July and reached almost 37% during January.