Patient’s Characteristics
The study population included 1003 positive for SARS-COV-2 samples corresponding to 1003 patients (Fig. 1). Patient characteristics were summarized in Table 2. Patients were classified according to increasing severity. The "ambulatory" group include symptomatic patients who were seen at the adult emergency department and discharged without hospitalization. The "conventional care unit" group include symptomatic patients requiring hospitalization without intensive care. The "intensive care unit" group included patients requiring intensive care, mainly for high flow oxygen therapy or intubation due to respiratory distress. The mean age of our population was 66 years with extremes of age from 16 to 104 years. The proportion of males and females in the total population yielded a sex ratio of 1.22 (Table 2).
We first analyzed the global distribution of different variables scored in this study among the three groups of patients as defined in the Method section. The details of the comparative statistics on the referral group are summarized in Table 2. There was a significant difference in the survival criterion between the three groups (p < 0.0001), more marked in the outpatient group than in the inpatient and intensive care groups, with a survival rate of 100% in the outpatient group (ambulatory group), 78.8% in the inpatient group (CCU group) and 76.5% in the intensive care group (ICU group) respectively. There was also a significant difference on the "sex" criterion for all three groups (p < 0.001), more pronounced with a multiple Fisher test between the CCU and ICU groups (p < 0.001), and between the ambulatory and ICU groups (< 0.05), but no significant difference between ambulatory patients and those hospitalized in conventional care. The use of oxygen therapy technique was also significantly different between the three observed groups (p < 0.0001).
Table 2
Population characteristics and comparative statistics between orientation groups
Patients
N = 1003 (%)
|
Total
|
Ambulatory
204 (20.4)
|
CCU
557 (55.5)
|
ICU
242 (24.1)
|
p-value
|
Median age [IIQ]
|
68 [52–83]
|
47 [30.75-66]
|
79 [65–87]
|
62 [51.25-70]
|
< 0.0001
|
Maximum age
|
104
|
104
|
100
|
85
|
-
|
Minimum age
|
16
|
16
|
17
|
21
|
-
|
Survival
|
828 (82.6)
|
204 (100)
|
439 (78.8)
|
185 (76.5)
|
< 0.0001
|
No oxygen therapy
|
352 (35)
|
203 (99.5)
|
146 (26.2)
|
3 (1.2)
|
< 0.0001
|
Oxygen therapy
|
426 (42.5)
|
1 (0.5)
|
411 (73.8)
|
14 (5.8)
|
High flow oxygen therapy
|
96 (9.6)
|
0
|
0
|
96 (39.7)
|
Tracheal intubation
|
129 (12.9)
|
0
|
0
|
129 (53.3)
|
Male
|
552 (55)
|
110 (53.9)
|
284 (51)
|
158 (65.3)
|
< 0.001
|
Haemophilus influenzae
|
|
7 (3.4)
|
27 (4.8)
|
16(6.6)
|
0.3
|
Neisseria meningitidis
|
|
0 (0)
|
6 (1.1)
|
6 (2.5)
|
0.05
|
Streptococcus pneumoniae
|
|
5 (2.5)
|
21 (3.8)
|
11 (4.5)
|
0.5
|
Influenzae A
|
|
0
|
0
|
0
|
-
|
Pcr-based Detection Of Microbial Agents
PCR-based detection of Hi using hpd and ompP2 genes allowed the detection of 50 Hi-positive samples with an average Ct value of 32 accounting for 5.0% of all the 1003 tested samples. All the 50 samples were negative for bexD gene suggesting that they corresponded to non-capsulated (non-typeable) isolates.
PCR-based detection of Nm was performed on the basis of the detection of sodC gene that yielded 12 samples positive for the presence of N. meningitidis, with an average Ct value of 32 and corresponding to 1.2% of all the 1003 tested samples PCR analysis using ctrA gene identified 6 ctrA positive samples suggesting that they correspond to groupable meningococci but complete analysis could not be performed due to lack of sufficient extract.
PCR-based detection of S. pneumoniae used with the lytA gene revealed 37 positive samples with an average Ct value of 33 and corresponding to 3.7% of all the 1003 tested samples.
Finally, no sample was positive for the presence of influenza A virus. Double bacterial carriage was observed for 16 samples. No sample was positive for the presence of the 3 bacterial agents tested.
A temporal distribution of bacterial carriage is shown in Fig. 2. The prevalence curves as a function of time overlap for H. influenzae and S. pneumoniae carriage. They appear to follow the trend in the frequency of COVID-19 infection. A concomitant increase of COVID-19 cases as well as the detection of the three bacterial agents was observed since November 2021 after easing from the two-periods of lockdowns in France that were implemented during the study period. The first period, from October 30 to December 15, 2020, was partially included in our study, and the second period was from April 3 to May 3, 2021 (Fig. 2). There was no significant difference between the referral groups on the carriage of H. influenzae, N. meningitidis or S. pneumoniae.
Association Of The Presence Of The Three Bacterial Agents With The Groups Of Patients
We next compared the distribution of the variables scored in this study between one group called "carriage" group, which includes patients with samples positive for the presence of at least one of the three bacteria in nasopharyngeal carriage, and a "no carriage" group corresponding to patients with samples for whom none of the 3 bacteria tested was detected (Table 3). The "carriage" group is younger (p < 0.01) than the “non-carriage” and includes more males (p < 0.01).
Table 3
Comparative statistics between patients with and without bacterial carriage
Patients
N = 1003
|
Bacterial carriage
N = 83 (8,3%)
|
No carriage
N = 920 (91,7%)
|
|
p-value
|
|
Median
|
[IIQ]
|
Median
|
[IIQ]
|
|
|
Age
|
64 [46–72]
|
69 [53–83]
|
|
< 0,01
|
|
Number
|
(%)
|
Number
|
(%)
|
OR
|
[IC 95%]
|
|
Sex (male)
|
59
|
(71,1)
|
493
|
(53,6)
|
2.12
|
[1, 3–3, 6]
|
< 0,01
|
Survival
|
74
|
(89,2)
|
754
|
(82)
|
0,55
|
[0,2 − 1,1]
|
0,13
|
No oxygen therapy
|
20
|
(24,1)
|
332
|
(36,1)
|
0,56
|
[0,3 − 0,9]
|
0,03
|
Oxygen therapy
|
38
|
(45,8)
|
388
|
(42,2)
|
1,16
|
[0,7 − 1,9]
|
0,56
|
HF oxygen therapy
|
6
|
(7,2)
|
90
|
(9,8)
|
0,72
|
[0,2 − 1,7]
|
0,56
|
Tracheal ntubation
|
19
|
(22,9)
|
110
|
(12)
|
2,18
|
[1, 2–3, 8]
|
< 0,01
|
All oxygen therapy cases
|
63
|
(76)
|
588
|
(64)
|
1.77
|
[1,0–3,2]
|
0.04
|
Amubulatory
|
10
|
(12)
|
194
|
(21,)
|
0,51
|
[0,2 − 1,0]
|
0,06
|
Hospitalization
|
47
|
(56,6)
|
510
|
(55,4)
|
1,04
|
[0,6 − 1,7]
|
0,90
|
Intensive care
|
26
|
(31,3)
|
216
|
(23,5)
|
1,48
|
[0,9 − 2,5]
|
0,11
|
All hospitalization cases
|
73
|
(88)
|
726
|
(79)
|
1.94
|
[1,0–4,3]
|
0.07
|
There was a significant difference in the need for oxygen therapy, with the "non-carriage" group seeming to have less need for oxygen therapy (p < 0.05). Patients requiring oxygen therapy were significantly more frequent among carriage (p = 0.04) and this was mainly observed among patient with intubation (p < 0.01). High flow and simple oxygen therapy was not significantly different.
Concerning the orientation of the patient, there was no significant difference between the two groups.
Among the 19 patients intubated in the "carriage" group, 16 samples were positive for H. influenzae, 11 for S. pneumoniae and 6 for N. meningitidis. In this sub-population, 8 patients presented X-ray and bacteriological findings suggesting a pulmonary bacterial superinfection with H. influenzae, such as pneumopathy may have been acquired under mechanical ventilation, with a possible passage of the bacterial carriage during intubation. All these patients were treated with beta-lactams.
Prediction Of Susceptibility To Beta-lactam For Hi-positive Samples
The use of beta lactams to treat patients suffering from a suspected pulmonary bacterial superinfection mainly with H. influenzae led us to focus on genetic typing of beta lactam susceptibility of H. influenzae positive samples. We therefore analyzed the sequence of the ftsI gene to predict beta-lactamase independent resistance to amoxicillin and third generation cephalosporins (3GC). The sequences of fstI gene were successfully obtained from 621 bp ftsI PCR-amplified fragments from all the 50 Hi positive samples. These DNA sequences allowed the identification of 29 different ftsI alleles and were named ftsIn where n is the allele number. Alleles 312, 313, and 314 were new alleles identified in our study and were added to the PUBMLST database. The 29 alleles were grouped into three groups 1 to 3 according to the already described scheme based on to the presence/absence of mutations in ftsI that were suggested to be involved in beta-lactam resistance (15) (Table 4). The allele frequencies among samples varied from 1 to 6 samples with 7 alleles that were represented by more than one sample (ftsI4, ftsI8, ftsI10, ftsI27, ftsI50, ftsI193 and ftsI314). Concerning the frequency of the groups of beta-lactam resistance, we note a prevalence for alleles of group 1, that do not harbor any mutations in the critical region of PBP3. This group was represented by 10 alleles found in 25 samples (50% of all the 50 Hi-positive samples). Alleles of Group 2, that do not harbor mutations associated with resistance to beta lactams but present other polymorphic sites in the critical region of PBP3, were represented by 7 alleles that were found in 12 samples (24% of all the 50 Hi-positive samples). Alleles of Group 3, that harbor mutations associated with resistance to beta lactams in the critical region of PBP3, were represented by 12 alleles that were found in 13 samples (26% of all the 50 Hi-positive samples). The alleles are correlated to beta-lactamase negative ampicillin resistance (15) (Table 4). No allele belonging to group 4, correlated to the resistance to 3GC and harboring four key mutations (D350N, S357N, M377I and S385T) was detected (15). The phylogenetic tree drawn by SplitsTree4 based on the CLUSTALW alignment of amino acid sequences deduced from the 29 ftsI alleles (Fig. 3) clearly showed the clustering of the 29 deduced amino sequences into the 3 groups. One allele in group 2 (ftsI302) was distantly separated from the other group 2 alleles and seems to be also shared by Haemophilus parainfluenzae (on the data base pubmlst.org/organisms/haemophilus-influenzae access on the 08/11/2022) (15).
Table 4
Mutations in the ftsI gene
Group*
|
Alleles (n° of isolates if > 1)
|
Mutation associate with amoxicillin resistance (N526K)*
|
Mutations associate with third generation cephalosporine resistance (D350N, S357, M377I, S385T)*
|
Other mutations
|
1
|
4 (2), 8(5), 10(5), 15, 27(6), 37, 63, 70, 156, 193 (2)
|
No
|
No
|
No
|
2
|
39, 50(5), 55, 97, 116, 302 and 314 (2)
|
No
|
No
|
Yes
|
3
|
1, 5, 20, 21, 48(2), 122, 142, 153, 266, 272, 312, 313
|
Yes
|
No
|
Yes
|
*The groups and key mutations for resistance were according to Deghmane et al (15). Alleles (55 and 39); (5, 122) (20, 142) and (153, 313) showed identical amino-acid sequences |