STRENGTHS AND LIMITATIONS OF THIS STUDY

The study assessed COVID-19 risk using real-world data comparing non-severe and severe patients with COVID-19.

The main limitation of this study was that it was a single-centre retrospective study with insufficient sample size and possible bias.

Further clinical studies were needed to determine the applicability of albumin-neutrophil count to lymphocyte count ratio score in pregnant patients and those under 18 years old.

Introduction

The novel COVID-19 pandemic poses significant challenges for society, medicine and science. The disease can quickly progress from mild influenza-like symptoms to various complications and even life-threatening respiratory failure.1 2 COVID-19 has caused millions of confirmed cases and deaths worldwide, especially in severe COVID-19 with high mortality rates. Therefore, it is important to identify patients who might progress to severe disease as early as possible to carry out timely and accurate interventions, help provide appropriate care and optimise the use of limited resources.3 Studies have suggested that advanced age and underlying medical conditions (including hypertension, diabetes and cardiovascular disease) are risk factors.4 Relevant biomarkers, such as serum albumin and neutrophil count to lymphocyte count ratio (NLR), can reflect the severity of COVID-19 to a certain extent. However, a single biomarker may have limitations and produce inconsistent results. Researchers have used a combination of multiple biomarkers.

Previous studies have observed that serum albumin is a biomarker of malnutrition-inflammatory syndrome and that there is a complex relationship between serum albumin and inflammation under a variety of physiological and pathological conditions.5 Serum albumin is a negative acute-phase reactant that reflects a patient’s nutritional status, surgery, burns and inflammation through reduced plasma levels.6 Kheir et al reported that serum albumin levels are associated with the prognosis of patients with COVID-19.7 The decreased serum albumin levels are common in patients with COVID-19.8 9 Studies have shown that serum albumin is associated with COVID-19 mortality.10–12

Inflammatory state biomarkers are important prognostic indicators in patients with COVID-19. The NLR is related to systemic inflammatory status and disease activity such as cardiovascular, autoimmune, tumour and infectious diseases.13–15 NLR is an independent risk factor for patients with COVID-19.16–18

Although serum albumin level and NLR are common biomarkers, they are easy to assess in clinical practice. However, an increasing number of studies have shown that the combination of serum albumin and NLR, as a serum albumin-NLR score (ANS), could have good prognostic value for coronary artery disease,19 cancer20–22 and postoperative complications.23 24 Whether ANS could be used to measure the prognosis of COVID-19 is unclear, and this study hypothesised that ANS was a simple prognostic tool that could be used to the severity of patients with COVID-19.

Materials and methodsOverview of the study

Between December 2022 and February 2023, 370 newly diagnosed patients with novel coronavirus pneumonia were diagnosed at the Affiliated Hospital of Jiangsu University (Zhenjiang, China). All patients with novel coronavirus pneumonia were newly diagnosed and untreated. Two independent researchers extracted and analysed the electronic medical records of each patient using standardised data collection forms. Early in the morning before treatment, a peripheral blood sample was collected on an empty stomach. We included data available during the hospital stay (laboratory tests, imaging findings, underlying disease and disease severity). We adopted the WHO diagnostic criteria to classify the severity of COVID-19. Severe and critical diseases were uniformly divided into severe diseases, whereas the rest were classified as non-severe diseases.25

Inclusion and exclusion criteria

All selected patients met the following criteria: (1) nasal and throat swab samples were positive for the new coronavirus nucleic acid test using RT-PCR; (2) age ≥18 years; (3) chest CT revealed COVID-19-related changes, such as small patch opacities and interstitial changes, multiple ground-glass opacities, infiltrative opacities or pulmonary consolidation; (4) not received treatment for a novel coronavirus infection before the visit; (5) clinical information collected before treatment. All enrolled patients were excluded for the following reasons: (1) acute infection at other sites (acute pancreatitis, acute cholecystitis and liver abscess); (2) lung infection caused by other factors; (3) recently, antiplatelet and anticoagulant drugs, immunosuppressants and other diseases that the investigators believed may have affected the results of the study; (4) pregnancy; (5) Case data were missed.

Detection metrics

Chest imaging was performed using CT (SOMATOM Definition, Germany). Routine blood tests used an SYSMEXXN3000 automated blood analyser (Sysmex Corporation, Japan). Serum albumin was measured using a BEKMAN AU5800 automated biochemistry analyser (Beckman Coulter). Normal ranges for all indicators were recorded according to the manufacturer’s instructions. The NLR and ANS were calculated as follows: the NLR was defined as a neutrophil count (109/L)/lymphocyte count (109/L). The ANS combined serum albumin and NLR and was 1 point below the optimal cut-off for serum albumin (as determined by the Youden index) or above the best NLR cut-off. The ANS scores ranged from 0 to 2 for each patient.

Statistical analysis

SPSS software (V.22.0) was used for statistical analysis. Based on the data expressed as mean±SD (x±s) or median (IQR) (M (P25, P75)). Categorical variables were expressed as the number and percentage of patients (%). The optimal cut-off value was determined by maximising the Youden index. A student’s t-test or Mann-Whitney U test was used to compare differences between two groups, whereas a one-way analysis of variance (for normally distributed variables) or the Kruskal-Wallis test (for non-normally distributed variables) was used for comparisons among three groups. The χ2 test was used for categorical variables. Logistic regression analysis was performed to explore the association of NLR, serum albumin and ANS with severe COVID-19. Subgroup analysis was conducted by stratifying age (<60/≥60 years), sex (female/male), diabetes (yes/no) and hypertension (yes/no). Statistical significance was set at a two-sided p value<0.05.

Patient and public involvement

The patients and the public were not involved in the research design, conduct, reporting or dissemination plans.

ResultsBaseline characteristics of patients

In total, 370 patients with positive nucleic acid test results for novel coronaviruses were included. Of these, 100 (27%) had severe disease, and 270 (73%) had non-severe disease. The average age of patients in the severe group was 78.60±11.16 years; the non-severe group was 73.43±13.49 years. The age of the patients in the severe group was significantly higher than that in the non-severe group (p<0.05). Moreover, there were 76 (76%) and 174 men (64.4 %) in the severe and non-severe groups, respectively. The proportion of male patients in the severe group was significantly higher than that in the non-severe group (p<0.05). The respiratory rate (RR) of patients in the non-severe group was 20.10±2.43 bpm, while the severe group was 23.51±4.97 bpm. The RR of patients in the severe group was significantly higher than that of patients in the non-severe group (p<0.05). Oxygen saturation was 96.53%±1.59% in the non-severe group and 86.69%±6.18% in the severe group. The oxygen saturation of patients in the severe group was significantly lower than that of patients in the non-severe group (p<0.05). Neutrophil and NLR levels were significantly higher in severe patients than those in non-severe patients. Lymphocyte and serum albumin levels were significantly lower in severe patients than those in non-severe patients (both p<0.05). No significant differences were observed in hypertension or diabetes between the two groups (p>0.05) (table 1).

Table 1

Comparison of baseline characteristics of severe and non-severe COVID-19

Cut-off values for the maximum potential effectiveness of serum albumin and NLR

The Youden index was used to determine the cut-off values representing the maximum potential effectiveness of serum albumin and NLR. ANS was calculated using the following data: the cut-offs for serum albumin and NLR were 33.5 g/L (area under the curve (AUC) for serum albumin, 0.643) and 8.25 (AUC for NLR, 0.698), respectively, as defined by receiver operating characteristic curve analysis. For the ANS assessment, patients with serum albumin levels ≥33.5 g/L and NLR ≤8.25 were assigned a score of 0; patients with low serum albuminemia (serum albumin level, <33.5 g/L) or elevated NLR levels (>8.25) were assigned a score of 1; patients with low serum albuminemia (serum albumin level, <33.5 g/L) and elevated NLR levels (>8.25) were assigned a score of 2; ANS scores ranged from 0 to 2 for each patient (table 2, figures 1 and 2).

Table 2

Cut-off values for maximum potential validity of serum albumin and NLR

Figure 1

The AUC-ROC analysis of NLR. AUC, area under the curve; NLR, neutrophil count to lymphocyte count ratio; ROC, receiver operating characteristic curve.

Figure 2

The AUC-ROC analysis of serum albumin. AUC, area under the curve; ROC, receiver operating characteristic curve.

ANS-based baseline characteristics

Table 3 shows that 98, 152 and 120 patients were in the ANS 0, 1 and 2 groups, respectively. The age of patients in the ANS 0, 1 and 2 groups was 69.51±14.63, 76.74±11.79 and 76.74±12.13 years, respectively. The age of the patients in the ANS 2 group was significantly higher than that in the other two groups (p<0.05). Moreover, there were 45 men (45.9%) in the ANS 0 group, 48 men (31.6%) in the ANS 1 group and 27 men (22.5%) in the ANS 2 group. The proportion of men in the ANS 2 group was significantly lower than that in the other two groups (p<0.05). The average RR of patients in the ANS 0, 1 and 2 groups was 20.04±2.45, 20.61±2.89 and 22.33±4.77 bpm, respectively. The RR of patients in the ANS 2 group was significantly higher than that in the ANS 0 and 1 groups (p<0.05). The average oxygen saturation of patients in the ANS 0, 1 and 2 groups was 95.60±3.55%, 94.58±4.92% and 91.56±6.86%, respectively. Patients’ oxygen saturation in the ANS 2 group was significantly lower than that in the other two groups (p<0.05). According to the current analysis, the neutrophil and NLR levels in the ANS 2 group were significantly higher than those in the ANS 0 and 1 groups, whereas the lymphocyte and serum albumin levels of the ANS 2 group showed the opposite trend, with a statistically significant difference (p<0.05). No significant differences were observed in the incidence of hypertension or diabetes among the three ANS groups (p>0.05). As the ANS index increased, the number of non-severe patients decreased, whereas the number of severe patients increased.

Table 3

Baseline characteristics of the study population according to ANS

Association between serum albumin, NLR, ANS and severe COVID-19

The relative risk of severe COVID-19 in patients with low serum albumin in the unadjusted group was 2.93 (95% CI 1.68 to 5.11, p<0.05), high NLR was 4.22 (95% CI 2.60 to 6.88, p<0.05) and ANS 2 was 7.65 (95% CI 3.72 to 15.75, p<0.05). To exclude the effects of age, sex, hypertension, diabetes, RR and oxygen saturation, the data were corrected for the analysis. In adjustment I, we adjusted for age and sex. The relative risk of severe COVID-19 in patients with low serum albumin was 2.42 (95% CI 1.37 to 4.29, p<0.05), high NLR was 4.04 (95% CI 2.45 to 6.69, p<0.05) and ANS 2 was 6.28 (95% CI 2.99 to 13.16, p<0.05). In the adjusted II, we adjusted for age, sex, hypertension, diabetes mellitus, RR and oxygen saturation confounders. The relative risk of severe COVID-19 in patients with low serum albumin was 1.52 (95% CI 0.80 to 2.88, p>0.05), which was not statistically significant (p>0.05). The relative risk of severe COVID-19 in patients with high NLR was 2.74 (95% CI 1.57 to 4.79, p<0.05), and ANS 2 was 3.56 (95% CI 1.54 to 8.24, p<0.05) (table 4).

Table 4

Association between serum albumin, NLR, ANS and severe COVID-19

Subgroup analysis

Studies showed that older age and underlying medical conditions (including hypertension, diabetes and cardiovascular disease) were risk factors for COVID-19. Age, sex, hypertension and diabetes were included in the subgroup analysis (figure 3). No interaction was observed between age, sex, hypertension, diabetes and COVID-19 severity, and there was no significant difference in the severity of COVID-19 in these subgroups (p>0.05).

Figure 3

Subgroup analysis of age, sex, hypertension and diabetes mellitus.

Discussion

This study revealed that the serum albumin level, NLR and ANS were significant prognostic factors of COVID-19 severity. Patients in the ANS 2 group (lower serum albumin levels and higher NLR) were more likely to become critically ill. The findings of this study shed new light on the ANS as a biomarker of severity in patients with COVID-19. The ANS can aid clinicians in promptly identifying COVID-19 severity and determining appropriate treatment objectives.

Studies have shown that multiple serum albumin-related rates are strongly associated with COVID-19. These included the neutrophil count to serum albumin ratio,26 C reactive protein to serum albumin ratio,27–29 D-dimer to serum albumin and fibrinogen to serum albumin ratios.28 Neutrophils play an important role in mediating inflammatory responses.30 The body’s immune response to viral infections primarily relies on lymphocytes.31 NLR, which comprises central neutrophils and lymphocytes, is an important biomarker of COVID-19.32 The ANS used in this study is a comprehensive evaluation index with high prognostic value.23 Additionally, it reacted more comprehensively to the immune and inflammatory states of the body. Therefore, we considered the ANS as a biomarker for the severity of COVID-19.

Several studies have shown that the ANS can be used as an important predictor of disease and surgery.33 These include pancreaticoduodenectomy,23 vascularised fibular flap reconstruction,24 coronary artery disease, subclinical coronary artery disease,19 neck cancer21 and rectal cancer.20 Based on the current research, there is no indication that the ANS is linked to COVID-19. ANS was divided into three groups: ANS 0 (high serum albumin and low NLR), ANS 1 (low serum albumin or high NLR) and ANS 2 (low serum albumin and high NLR). Serum albumin plays a major role in maintaining the plasma colloidal osmotic pressure and serves as a transporter.34 As one of the important biomarkers of COVID-19, a low serum albumin level is an important biomarker of COVID-19.8 Increased NLR levels in patients could further cause an increase in systemic proinflammatory cytokines,35 which are independent risk factors for COVID-19. Therefore, compared with serum albumin level and NLR alone, the ANS could provide more reliable outcomes.23 In this study, we observed that the ANS is associated with COVID-19 severity. The ANS is convenient for calculations in clinical practice and is beneficial for clinical patients. In summary, ANS can be used as a prognostic indicator of COVID-19 severity.

To the best of our knowledge, this study demonstrated, for the first time, the prognostic value of the ANS for COVID-19 severity. However, the limitations of this study warrant further investigation. The review was retrospective; data were obtained from a single clinical site rather than multiple clinical sites; the sample size was insufficient and there may have been a few biases.

Conclusion

Our study showed that serum albumin, NLR and ANS levels were related to the severity of infection. The conclusions of this study supported the use of the ANS as a prognostic indicator of COVID-19 severity.

Data availability statement

Data are available upon reasonable request. The data involved in this study are all patients in our centre and are not routinely disclosed.

Ethics statementsPatient consent for publication

Not applicable.

Ethics approval

This study involves human participants and was approved by Scientific Research Ethics Committee of the Affiliated Hospital of Jiangsu University KY2023K1005. Participants gave informed consent to participate in the study before taking part.