Medical records of total 383 COVID-19 positive patients admitted in GNRC Medical from June, 2020 to September, 2021 were scrutinized. Out of these 383 medical records, it was observed that most common clinical presentation was cough (47.5%) followed by fever (38.4%), both fever and cough (37.1%), respiratory distress (23%), gastrointestinal symptoms (16.7%), general weakness (12.8%), urinary symptoms (10.2%) and sore throat (2.6%). Rest (12.8%) were asymptomatic, hospitalised for some other ailment and diagnosed accidentally to be SARS-CoV-2 positive. Though 23% of the patients were admitted with respiratory distress, but only 15.4% of the COVID-19 positive patients showed positive radiological findings like opacities, consolidations, pleural and septal thickening, cavitary lesions, fibrosis and pleural effusions. Out of 383 SARS-CoV-2 positive participants, 110 (28.7%) presented with co-morbidities such as diabetes 19.1% (21/110), hypertension 18.2% (20/110), both diabetic and hypertensive 10% (11/110). Though other co-morbidities like renal disease (5.5%), liver disease (6.4%), chronic lung disease (3.6%) heart disease (6.4%) and anaemia (5.5%) were less frequent, but multi-organ involvement was found in 30.9% (34/110) of the studied COVID-19 positive participants. Common treatment protocol followed were antibiotic (94.8%, 363/383), steroid (50%, 192/383), antiviral (28.2%, 108/383), and oxygen support (23%, 88/383). In 29.5% patients, oxygen saturation was found to be < 95%. Pulse rate (> 100), respiratory rate (> 24) and blood pressure (> 140/90) were recorded high in respectively, 19.3%, 5.2% and 15.4% of the studied COVID-19 patients (Table S1, Supplementary file).
For convenience of data analysis, the COVID-19 positive participants were divided into four categories like, dysnatremic (aberrant serum sodium status, i.e. serum sodium level > 146 mmol/L or < 134 mmol/L), hypernatremic (high serum sodium status, i.e. > 146 mmol/L), hyponatremic group (low serum sodium status, i.e. < 134 mmol/L) and eunatremic (normal serum sodium status, i.e. serum sodium in between 134 to 146 mmol/L). Along with the COVID-19 positive patients (test group), serum sodium status of 90 hospitalised patients (control group) who demonstrated a negative test result of PCR/RAT test for SARS-CoV-2 were also observed. There was no significant difference (p = 0.9) in the mean serum sodium level between the test (141 ± 5.9 mmol/L) and the control group (142 ± 6.5 mmol/L) ranging from 114–182 mmol/L in test group and 123–163 mmol/L in the control group. The age of the COVID-19 positive participants of this study ranges from 10 months to 90 years with a mean age of 52 ± 16.1 years, among which 262 (68.4%) were male and rest 121 (31.6%) were female while the mean age of the control group was 53 ± 15.7 years, ranging from 17 to 81 years with 56.7% (51/90) male and 43.3% (39/90) female. Age and gender distribution of the studied groups did not display any significant variation.
Hypernatremia was observed in 27.7% (106/383) of the included COVID-19 patients against only 13.3% (12/90) of the COVID-19 negative patients, which is a significant difference with a p value of 0.02. (Fig. 1). Hyponatremia was witnessed in 12.3% (47/383) of the included COVID-19 patients compared to 6.7% (06/90) of the COVID-19 negative patients, although the difference is not statistically significant (p = 0.1).
Altogether, 145 out of 383 (37.9%) of the included COVID-19 participants exhibited dysnatremia compared to only 18 out of 90 (20%) of the patients from the control group, demonstrating a significant difference with a p value of 0.02.
Incidence of hypernatremia was relatively high compared to hyponatremia among COVID-19 positive (27.7% vs 12.3%) as well as control group (13.3% vs 6.7%) and the incidence was reported to be significantly high (p = 0.01) in the COVID-19 positive participants. It also turned out to be relatively severe with significantly high (p = 0.01) mortality i.e. 41.5% (44 of 106) among hypernatremic compared to only 21.3% (10 of 47) among hyponatremic COVID-19 participants. (Table 1).
However, 2.1% (08/383) of the included COVID-19 participants developed both hyper and hyponatremia during their hospital stay without any history of intravenous administration. The length of hospitalisation of the participants from the control group (2 ± 1.1 days) was significantly less (p = 0.03) compared to that of the test group (9 ± 6.2 days). The difference in the occurrence of dysnatremia between deceased and surviving test and control group did not reach statistical significance, which may be due to insufficient data in the control group for proper statistical comparison. On several occasions, i.e. 11.7% (17/145) in case of COVID-19 positive participants and 5.5% (1/18) of COVID-19 negative participants, the serum sodium concentration reached critical values on either side of the normal (high up to 182 mmol/L and low up to 114 mmol/L in COVID − 19 positive group and up to 163 mmol/L in control group). Though there was a noticeable difference in incidence of critical reporting of serum sodium value between the test (11.7%; 17/145) and the control (5.5%: 01/18) group but it failed to attain statistical significance with a p value of 0.1 (Table 1). Because of the limited period of hospital stay and inadequate laboratory data, we could not compare other clinical characteristics of the COVID-19 positive participants with the control group.
Number of COVID − 19 positive participants presenting aberration in their serum sodium level at the time of admission was more compared to the number participants that had developed later and the difference was statistically significant in case of hyponatremic patients (p = 0.005). For example, 50.9% (54/106) of hypernatremia and 70.2% (33/47) of hyponatremia was detected during admission whereas 49.1% (52/106) of hypernatremia and 29.8% (14/47) of hyponatremia developed later during their hospital stay. But whether the aberrant serum sodium was presented during admission or developed later, it did not affect the mortality rate among them (p > 0.05).
Table 2
Serum sodium status with statistical analysis (Chi-square test) of the Covid-19 positive participants
Observed criteria | Dysnatremia (n = 145) | Hypernatremia (n = 106) | Hyponatremia (n = 47) | Both (n = 08) |
On admission Developed later p value (X2) Mortality when aberration presented on admission Mortality when aberration developed later p value Mean Age (yrs) p value (compared with eunatremic group, mean age = 50.5) Male Female p value Mortality among male Mortality among female p value | 87 (60%) 66 (45.5%) 0.08 31 23 0.2 54.8 0.6 104(71.7%) 41(28.3%) < 0.0001 37 12 0.0003 | 54(50.9%) 52(49.1%) 0.8 23 21 0.7 54.1 0.7 75(70.8%) 31(29.2%) < 0.0001 32 12 0.002 | 33(70.2%) 14(29.8%) 0.005 06 02 0.06 55.6 0.6 36(76.6%) 11(23.4%) 0.0002 09 01 0.01 | - - 53.1 0.7 07(87.5%) 01(12.5%) 0.03 4 1 0.1 |
Observed criteria | n (%) | p value |
1. Total critical hypernatremic value 2. Total critical hyponatremic value 3. Critical value (hyper) detected on admission a. Non-Survivors 4. Critical value (hyper) detected later a. Non-Survivors 5. Critical value (hypo) detected on admission a. Non-Survivors b. Critical value (hypo) detected on later | 16/106 (15.1%) 01/47 (2.1%) 06/54 (11.1%) 06/06 (100%) 10/52 (19.2%) 04/10 (40%) 01/33 (3%) 0/01 (0%) 0/14 (0%) | p1 = 0.001 p2 = 0.1 p3 < 0.0001 |
*p1 (1–2); p2 (3–4); p3 (3a-4a) |
The occurrence of dysnatremia (p = < 0.0001) including both hyper (p < 0.0001) and hyponatremia (p = 0.0002) was observed to be significantly high among male patients than the female patients. Furthermore, the condition was proved to be more critical with significantly high mortality rate, among male than the female COVID-19 positive participants but did not show any significant correlation with age.
On several occasions, the serum sodium concentration of the COVID − 19 positive participants reached critical values on either side of the normal. Total 6 out of 54 (11.1%) hypernatremia detected on admission, showed critical sodium value with 100% (6/6) mortality. Whereas 10 out of 52 (19.2%) of the critical sodium level was detected later with 40% mortality. It was observed that there was a significantly high mortality among the COVID-19 positive participants when admitted with a critically high sodium value (p < 0.0001) (Table 2).
The association of sodium levels with adverse outcome was analysed by the number of the non-surviving COVID-19 participants and their length of average hospital stay. As depicted in Fig. 2 and Table 3, mortality among the dysnatremic group including both hyper and hyponatremia was significantly high compared to that of the eunatremic COVID-19 participants.
Though the rate of casualty among both the hyper and hyponatremic group of participants were more compared to the eunatremic participants, hypernatremia was proved to be graver than the hyponatremia as illustrated by the significantly less mortality rate in the hyponatremic group compared to that of hypernatremic group of COVID-19 participants. The average hospital stay of the dysnatremic (10 days) including both hypernatremic (9 days) and hyponatremic (10 days) did not show any significant difference with that of the eunatremic participants (8 days).
Table 3. Statistical analysis (Fisher's exact test) showing the association of sodium levels (on admission) with adverse outcome
Adverse Outcome
|
Dysnatremic
(n = 145)
|
Hypernatremic
(n = 106)
|
Hyponatremic
(n = 47)
|
Eunatremic
(n = 238)
|
p value
|
Non-surviving: n(%)
OR
Hospital stay (days)
|
49 (33.8%)
0.236
10
|
44 (41.5%)
0.1923
09
|
10 (21.3%)
0.375
10
|
19 (8%)
08
|
p1<0.0001
p2<0.0001
p3 = 0.02
p4= 0.01
p1 = 0.6
p2 = 0.8
p3 = 0.6
|
*p1 (Dysnatremic-Eunatremic); p2 (Hypernatremic-Eunatremic);
p3 (Hyponatremic-Eunatremic); p4 (Hypernatremic-Hyponatremic).
To evaluate the clinical impact of dysnatremia and its correlation with other laboratory parameters, we retrospectively evaluated the data from all the included COVID-19 positive patients. While correlating the other laboratory findings with the patients’ serum sodium status, it was observed that, all the scrutinised parameters were not analysed in each and every participant of this study. For example, out of 383 human participants, TLC/TC was done in 377(98.4%) patients, likewise DLC in 376(98.2%), haemoglobin (Hb) in 349(91.1%), platelet in 232(60.6%), random blood sugar (RBS) in 193(50.4%), glycosylated haemoglobin (A1C) in 51(13.3%), urea and creatinine in 382(99.7%), liver enzymes like alanine transaminase (ALT), aspartate transaminase (AST) and alkaline phosphatase (ALP) in 364(95%), total proteins (TP) and fractions in 362(94.5%), hs-CRP in 202(52.7%), ferritin in 97(25.3%), PT in 141(36.8%), LDH in 91(23.8%), FIB in 93(24.3%), PCT in 93(24.3%), IL-6 in 93(24.3%) and D-Dimer in 104(27.2%) participants of the study (Table S2, Supplementary file).
Magnitude of dysnatremic patients (including both hyper and hypo) showing aberration in the circulatory level of the other laboratory parameters e.g. TC, polymorph (poly), monocyte (mono), urea, creatinine, ferritin, PT, LDH, FIB, PCT, IL-6, D-Dimer were significantly high to that of eunatremic COVID-19 positive participants. Except for A1C, there was no significant difference in magnitude in the above mentioned observations between hypernatremic and hyponatremic group of COVID-19 positive participants. Number of hyponatremic COVID-19 patients with high A1C value (92.3%) was significantly high (p = 0.04) than the hypernatremic COVID-19 patients with high A1C (66.7%). Out of the total 383 participants, 116 (30.3%) participants required intensive care unit (ICU) facilities of which 36 (31%) succumbed to death. Requirement of ICU facility was significantly high in dysnatremic (67/145, 46.2%; p = 0.001) and also in hypernatremic group of patients (58/106, 54.7%; p < 0.0001) than the eunatremic group (49/238, 20.6%). ICU requirement in hyponatremic group (16/47, 34%) did not reach significant level (p = 0.06) and the difference between hypernatremic and hyponatremic was significant showing a p value of 0.02. It was observed that, dysnatremia, especially hypernatremia when associated with aberrant circulatory levels of the scrutinized parameters except, mono and A1C ended up with a significantly high (p < 0.05) mortality rate compared to that of the eunatremic group of COVID-19 positive participants. Almost similar observation was also noticed in hyponatremia except when associated with aberrant TC/poly/creatinine/ TP/FIB and PCT.
Fatality rate was significantly more among hypernatremic with aberrant laboratory parameters than that of hyponatremic except when associated with high ALP, hs-CRP, PT, FIB, PCT, IL-6, and D-Dimer, where it failed to attain the significant level. Similarly, dysnatremic including both hyper and hyponatremic participants demanding ICU facilities demonstrated significantly high mortality rate (p < 0.0001) compared to eunatremic group requiring ICU facilities. Though the magnitude of non-survivors among hypernatremic requiring ICU admission (51.7%; 30/58) was more than that of the hyponatremic (37.5%; 06/16) but the difference did not reach the significant level (p = 0.1). (Table S2, Supplementary file).
The severity parameters which showed statistically significant raised mean values in the dysnatremic group (including both hyper and hyponatremic participants) were TC, poly, urea, creatinine, hs-CRP (with an exception in hyponatremic group, where the elevation was not statistically significant), ferritin, PT, LDH, PCT and D-Dimer. Conversely, parameters showing significantly decreased mean values in the dysnatremic group (including both hyper and hyponatremic participants) compared to the eunatremic group were lymphocytes (lympho), mono (with an exception in hyponatremic group, where the reduction was not statistically significant), TP and albumin. The mean values of all the scrutinised parameters did not demonstrate significant difference between hypernatremic and hyponatremic group with the exception of urea and IL-6 where mean urea and mean IL-6 was found to be significantly high in hypernatremic group than that of the hyponatremic group (Table S3, Supplementary file).
Severity of failed organs can be numerically quantified by the SOFA score (a method for risk stratification and prognosis of patients with severe sepsis) based on PaO2/FiO2, the Glasgow Coma Scale, mechanical ventilation (yes/no), platelets, bilirubin, mean arterial pressure or administration of vasoactive agents, and creatinine.24,30 As PaO2 was not captured for most of patients, the peripheral arterial oxygen saturation (SpO2) to FiO2 ratio (SaO2/ FIO2) was used. This alternative option has been previously validated with high correlation.30,31 The CURB-65 score is another mortality risk score for community acquired pneumonia (CAP), based on confusion, blood urea nitrogen, respiratory rate, blood pressure, and age, attributing 1 point for each item. It has been extensively validated to predict 30-day mortality in CAP.32
We employed both the severity scores but due to non-availability of all the required parameters we could calculate the SOFA score of only 28.7% (110/383) and CURB 65 score of only 32.4% (124/383) of the COVID-19 positive participants. It was observed that dysnatremic group (p < 0.0001) of participants, including both hyper (p < 0.0001) and hyponatremia (p = 0.0005) demonstrated significantly high SOFA score compared to eunatremic group of participants. But the difference between hypernatremic and hyponatremic was not significant (p = 0.7). Mortality risk based on CURB 65 score was found to be significantly high in hypernatremic group (p = 0.001) but failed to reach the statistical significance in hyponatremic group (p = 0.1) compared to eunatremic group (Table 4).
Table 4. Calculated SOFA score and CURB 65 score with statistical comparison (Student’s t test) among different study groups.
Sodium Status
|
SOFA Score
|
Survivors
|
Non Survivors
|
p value
|
Total
|
p value
|
n
|
Mean ± SD
|
n
|
Mean ± SD
|
n
|
Mean ± SD
|
Dysnatremic
Hypernatremic
Hyponatremic
Eunatremic
|
36
18
18
46
|
4.5 ± 2.2
4.1 ± 2.22
4.9 ± 2.17
2.7 ± 1.09
|
20
18
02
08
|
5.1 ± 2.1
5.2 ± 2.15
4.5 ± 2.12
3.6 ± 1.3
|
0.3
0.1
0.8
0.09
|
56
36
20
54
|
4.71 ± 2.16
4.6 ± 2.22
4.9 ± 2.11
2.9 ± 1.16
|
p1<0.0001
p2<0.0001
p3=0.0005
p4= 0.7
|
|
CURB 65 Score
|
Dysnatremic
Hypernatremic
Hyponatremic
Eunatremic
|
72
44
28
90
|
1.44 ± 0.63
1.5 ± 0.59
1.36 ± 0.68
1.26 ± 0.49
|
52
42
10
13
|
1.6 ± 0.77
1.57 ± 0.74
1.7 ± 0.95
1.31 ± 0.48
|
0.2
0.6
0.3
0.7
|
124
86
38
103
|
1.51 ± 0.69
1.53 ± 0.66 1.45 ± 0.76
1.26 ± 0.48
|
p1=0.001
p2=0.001
p3=0.1
p4= 0.5
|
*Note: p1 (dysnatremic‒eunatremic); p2 (hypernatremic‒eunatremic);
p3 (hyponatremic‒eunatremic); p4 (hypernatremic‒hyponatremic).
However, the difference in the occurrence of both the high SOFA scores and high CURB 65 scores between deceased and surviving dysnatremic (including both hyper and hyponatremia) participants did not reach statistical significance (Table 4).
Probability of death based on serum sodium values 24hrs after admission was assessed with the help of Kaplan-Meier curve. It was observed that probability of death was 2.35 fold increased in hypernatremic patients compared to eunatremic patients demonstrating a significant (Bonferroni corrected) p value of 0.002 (Fig. 3). On the contrary, the probability of death in the hyponatremic group of patients failed to attain the statistical significance (p = 0.3) with a 0.7 fold increase.