Background
The COVID-19 pandemic, due to SARS-CoV2 infection, was characterized in France by
a first wave of patients in the spring of 2020, followed by the current second wave
beginning in October 2020. The clinical spectrum of SARS-CoV2 infection is broad,
ranging from asymptomatic disease to severe pneumonia with respiratory failure, leading
to intensive care unit (ICU) requirement in 14 % of the cases and to an overall estimated
death rate around 2 % [1]. Some studies focused on mortality risk factors in patients
with COVID-19 and identified age, comorbidities and biological parameters [2]. The
relevance of common biological parameters is still to determine, and a scoring system
relying on biological features at admittance to predict disease evolution would be
highly useful. We present here the result of a monocentric study based on two cohorts,
each corresponding to the epidemic waves of spring and fall/winter respectively. The
first retrospective cohort of 154 patients, admitted in conventional or intensive
care units, identified 3 biological relevant risk factors of increased mortality among
the 36 analyzed: plasmatic sodium, potassium levels, and prothrombin time. This score
was then prospectively validated in an independent cohort of 81 patients, showing
a strong reliability to identify patients at risk of severe COVID-19 evolution. This
biological score, relying on broadly available parameters, is an easy and usable tool
to early discriminate high-risk patients that are most likely to benefit from intensive
care treatment.
Material and methods
Study design and clinical features
The first retrospective derivation cohort included all consecutive adult patients
(≥ 18 years old) admitted from February 21 st, 2020 to March 30th, 2020 in University
Hospital (CHU) of Amiens and diagnosed with COVID-19 according to the viral detection
of SARS-CoV2 (PCR). The second prospective validation cohort included all consecutive
adult patients admitted from October 19th, 2020 to November 17th, 2020 in the same
hospital and diagnosed with COVID-19 with a similar test. Clinical and biological
data were extracted from electronic medical records. The analyzed biological parameters
were collected at hospital admittance (day 0).
This study was approved by the institutional review board of Amiens University Hospital
(number PI2020_843_0031, 30th March), and are in accordance with French legislation
of non-interventional studies.
Statistical analysis
Student's t-test, receiver operating characteristic (ROC) curve analysis and Kaplan-Meier
analysis were used as indicated. Univariate and multivariate analyses were performed
using a step-by-step backward Cox regression. P < 0.05 was considered as statistically
significant.
Results
Patient demographics and baseline characteristics
In the derivation cohort, 154 patients were admitted with confirmed SARS-CoV2 infection
in the Amiens hospital. These patients were hospitalized in conventional units (n = 111,
72 %) or ICU (n = 43, 28 %), and 18 of them were transferred from conventional to
ICU. The median age at admission was 77 years (range 23–100), 56 % of the patients
were male and 44 % female. The median length of stay was 12 days both in conventional
care units and ICU. At the time of analysis, 122 patients were alive (79 %) and 32
died (21 %) from SARS-CoV2 infection. The second validation cohort prospectively included
81 SARS-CoV2 patients meeting the same criteria, whose characteristics were not statistically
different from the first cohort (data not shown).
Biological markers to identify high risk patients
We then performed survival analyses to determine if biological parameters at diagnosis
could predict clinical outcomes. OS was defined as time to death or latest news. To
consider the risk of severe clinical worsening, EFS was defined as time to ICU transfer,
death in conventional care units, or latest news. At median follow up, EFS and OS
were close to 85 %. Univariate analysis for EFS identified lymphopenia (p = 0.048),
hyponatremia (p = 0.003), high uremia levels (p = 0.02) and hypocalcemia (p = 0.01)
as risk factors for severe clinical worsening, whereas age had no significant impact
(p = 0.53). In multivariate analysis, hyponatremia (HR = 11.7 (95 %CI:3.1–44.2), p < 0.001)
and low bicarbonate levels (HR = 5.4 (95 %CI:1.4–21), p = 0.02) negatively affected
EFS.
Regarding OS, hyponatremia (p = 0.038), hyperkaliemia (p = 0.005), high creatinine
levels (p = 0.049), hyperphosphoremia (p = 0.006), and low O2 saturation (p = 0.01)
significantly predicted shorter survival in univariate analysis. There was a trend
for prolonged prothrombin time > 16,8 s (p = 0.1). Age was also an adverse prognostic
factor for OS (p = 0.008, HR = 1.04 (1.01–1.07)). Then, in multivariate analysis,
age (HR = 1.13 (95 %CI:1.04-1.22), p = 0.002), prothrombin time > 16,8 s (HR = 4.62
(95 %CI:1.19-17.94), p = 0.03), hyponatremia (HR = 6.99 (95 %CI:1.83-26.72), p = 0.005),
and hyperkaliemia (HR = 12.1 (1.66–87.75), p = 0.01) were independent factors predicting
shorter OS.
Easy usable score to predict early survival in COVID-19 patients: Biovid-19
Based on the results of the multivariate analysis presented above, we proposed a simple
biological prognostic score including sodium, potassium levels and prothrombin time.
0, 1 or 2 points are assigned to each of the 3 parameters, according to the weight
of their respective HR, resulting in a score ranging from 0 to 6 (Table 1
). Using ROC analysis, we proposed that a threshold ≥2 predicted poor prognosis with
a sensitivity of 80,7 % and a specificity of 93,3 %. Therefore, we independently validated
in our second prospective cohort an improved OS with a score <2 (Fig. 1
A), showing a strong reliability of Biovid-19 score to predict the risk of death in
time-different waves. Interestingly, we also validated that Biovid-19 score predicts
clinical worsening, as it prospectively highlighted a significantly lower cumulative
incidence of both secondary transfer to ICU and development of Acute Respiratory Distress
Syndrome (ARDS) in patients with a score <2 (Fig. 1B and C). Moreover, our score was
significantly lower in non-hospitalized patients and patients hospitalized in conventional
units compared to patients hospitalized in ICU (Fig. 1D), reflecting the association
with severe disease presentation.
Table 1
Biovid-19 score.
Table 1
Parameter
Range
Score point
Sodium (mmol/L)
≤130
2
]130-135]
1
>135
0
Potassium (mmol/L)
≥5,5
2
]5,5-5]
1
<5
0
PT (s)
>20
2
]20-16,8]
1
≤16,8
0
Interpretation of the proposal score: Patient at high risk of mortality if score ≥2.
Fig. 1
Overall survival according to the proposal score by the Kaplan Meier method.
Fig. 1
Taken together, these data suggest that few common and broadly available biological
parameters at admittance can easily and reliably identify patients at risk of severe
Covid-19 evolution. We assume that the Biovid-19 prognostic score is useful to early
identify high-risk patients requiring close monitoring, and could help guidance for
early ICU admittance.
Declarations
Ethics approval and consent to participate : This study was approved by the institutional
review board of Amiens University Hospital (number PI2020_843_0031, 30th March).
Consent for publication
Not applicable.
Availability of data and materials
Not applicable.
Funding
Not applicable.
Authors’contribution
TB and CS designed the research study. TB, CS, GC, ML, OE and AC collected and analysed
the data. CF, EB and SC set up the virological diagnosis. RN, JLS, CA, JM managed
patients and provided clinical data. TB, CS, GC, ML, AC and LG wrote the paper, which
was approved by all the authors.
Declaration of Competing Interest
The authors declare that they have no competing interests.