White Blood Cell Variability and Clinical Outcomes in Hospitalized Patients With COVID-19

doi:10.21203/rs.3.rs-1289810/v1

Download PDF

Research

White Blood Cell Variability and Clinical Outcomes in Hospitalized Patients With COVID-19

https://doi.org/10.21203/rs.3.rs-1289810/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Inflammation parameters, such as white blood cell (WBC) counts, are closely related to the severity and mortality of Coronavirus disease 2019 (COVID-19). Nevertheless, the association between inflammation fluctuation and clinical outcomes in patients with COVID‐19 is not clear.

Methods: We performed a secondary analysis of the ORCHID trial. Inflammation fluctuation was assessed by WBC variability indices (SD [standard deviation], and CV [coefficient of variation]). Cox regression was used to compute the hazard ratio for WBC variability and associated hospitalization and death over 28 days.

Results: During the 28 days post-randomization, 46 (10.8%) patients died, and 345 (81.4%) patients were discharged. In multivariable analysis, a worse clinical outcome, including decreased risk of hospital discharge (SD: HR: 2.32; P<0.001; CV: 1.72; P=0.001) and increased mortality (SD: HR: 1.51; P=0.002; CV: 1.35; P=0.006), was found with increased WBC variability. Furthermore, WBC variability had moderate performances in predicting discharge (C-index, SD: 0.705; CV: 0.703), and all-cause death over 28 days (C-index, SD: 0.632, CV: 0.688). Sensitivity analyses that involved changing 28-day mortality to in-hospital mortality, recalculating the WBC indices by using the lowest WBC value, and using the competing risk model generated confirmatory results. Most of the interactions between sex, age (>65), hydroxychloroquine treatment group, and WBC variability indices were not significant (P>0.05).

Conclusion: In conclusion, our results showed that COVID-19 patients with higher WBC variability had lower discharge rates and higher mortality. Our study contributes to our understanding of the relationship between inflammation and COVID-19-related adverse outcomes.

Clinical trial registration: URL: https://clinicaltrials.gov. Unique identifier: NCT04332991.

Coronavirus disease-19

White blood cells

death

discharge

Coronavirus disease 2019 (COVID-19), the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is an acute infection of the respiratory tract. Since the first report of the disease in December 2019, there have been more than 191 million confirmed cases of COVID-19, with more than 4 million lives lost worldwide[1].

A body of studies has shown that COVID-19 is characterized by an aggregated immune response and cytokine storm. White blood cell (WBC) count or the subproportion of WBCs is a simple systemic immune-inflammatory parameter representing the immune status of patients and is reported to be a strong predictor of severity and increased mortality in patients with COVID-19 [2–5]. For example, baseline total WBC counts, such as a high neutrophil-to-lymphocyte ratio (NLR), were reported to be significantly linked to increased disease progression and in-hospital death of COVID-19 patients[6]. Furthermore, dynamic changes in WBCs or the subproportion of WBCs during the natural history of COVID-19 might also be a helpful predictor for the prognosis of COVID-19 patients[7]. However, the association between WBC variability and clinical outcomes in patients with COVID-19 is not clear. We hypothesized that WBC variability, as a reflection of the fluctuation in inflammation levels, might be closely related to clinical outcomes in patients with COVID‐19. Thus, we conducted an observational study to investigate whether WBC variability would be independently associated with clinical outcomes by analyzing data from the ORCHID trial (Outcomes Related to COVID-19 Treated With Hydroxychloroquine Among Inpatients With Symptomatic Disease; URL: https://www.clinicaltrials.gov; ID: NCT04332991).

Baseline characteristics

Table 1 shows the baseline characteristics of 424 hospitalized patients with COVID-19, with 42.2% being females and 60.6% being Hispanic or Latino. The mean ±SD age values of all patients were 57.1 ±16.1, and the mean BMI was 32.9±10.0. A total of 52.7% of patients had a history of hypertension, and 35.0% had diabetes as a comorbidity.

Table 1

Baseline characteristics of included hospitalized patients with COVID-19 stratified by white blood cell variability
	Overall	Standard Deviation (SD) of WBC			Coefficient of Variation (CV) of WBC
Characteristics		<1.84×10³/mm3	≥1.84×10³/mm3	P value	<17.6%	≥17.6%	P value
Number of patients	424	282	142	0.630	198	226	0.366
Randomization to hydroxychloroquine (%)	215(50.8)	141(50.0)	74 (52.5)	0.705	96(48.5)	119(52.9)	0.421
Demography
Age, years	57.1±16.1	57.0±15.6	57.4±17.1	0.819	56.6±15.9	57.60±16.4	0.539
Sex, male (%)	244(57.8)	157(55.9)	87(61.7)	0.299	119(60.1)	125(55.8)	0.428
BMI, kg/m2	32.9±10.0	32.9±10.6	33.1±8.5	0.821	33.20±10.9	32.78±9.1	0.679
Ethnic, Hispanic or Latino (%)	249(60.6)	174(63.0)	75(55.6)	0.177	112(58.0)	137(62.8)	0.371
Chronic health conditions, n(%)
Peripheral vascular disease	19 (4.5)	12(4.3)	7 (5.0)	0.934	8 (4.0)	11(4.9)	0.853
COPD	37 (8.7)	22(7.8)	15(10.6)	0.429	15 (7.6)	22(9.8)	0.53
Hypertension	223(52.7)	149(52.8)	74(52.5)	0.998	100(50.5)	123(54.7)	0.449
Coronary artery disease	39 (9.2)	27(9.6)	12 (8.5)	0.859	19 (9.6)	20(8.9)	0.934
Diabetes mellitus	148 (35.0)	99(35.1)	49(34.8)	0.997	66(33.3)	82(36.4)	0.571
Moderate to severe kidney disease ^a	38 (9.0)	26(9.2)	12 (8.5)	0.952	16 (8.1)	22(9.8)	0.661
Liver disease ^b				0.998			0.777
Mild	6 (1.4)	4(1.4)	2 (1.4)		3 (1.5)	3(1.3)
Moderate/severe	3 (0.7)	2(0.7)	1 (0.7)		2 (1.0)	1(0.4)
Tumor ^c				0.664			0.113
Present	10 (2.4)	8(2.8)	2 (1.4)		7 (3.5)	3(1.3)
With metastasis	6 (1.4)	4(1.4)	2 (1.4)		1 (0.5)	5(2.2)
Home medication, n(%)
Corticosteroids	44(10.4)	31(11.0)	13 (9.2)	0.693	20(10.1)	24(10.7)	0.976
ACE inhibitors	66(15.6)	38(13.5)	28(19.9)	0.118	24(12.1)	42(18.7)	0.086
ARB	37 (8.7)	23(8.2)	14 (9.9)	0.670	10 (5.1)	27(12.0)	0.019
Total SOFA score	3.3 ±2.8	2.8±2.1	4.3±3.7	<0.001	2.7±1.9	3.8±3.4	<0.001
Measurements
Systolic blood pressure, mmHg	110 (100, 123)	110 (101, 123)	109 (96, 123)	0.341	111 (102, 122)	109(98, 123)	0.347
Lowest SpO2, %	92 (90, 94)	92 (90, 94)	92 (90, 94)	0.680	92 (90, 94)	92 (90, 94)	0.267
Highest respiratory rate, breaths per minute	26±7.8	25 ±7.4	28 ±8.1	<0.001	25 ±7.8	27±7.8	0.027
White blood cell count, ×10³/mm³	6.0 (4.3, 8.0)	5.60(4.2, 7.16)	7.2 (4.4, 10.7)	<0.001	5.9 (4.5, 7.8)	6.0 (4.1, 8.4)	0.702
Hemoglobin, g/dl	12.8 (11.4, 14.2)	12.9 (11.5, 14.2)	12.8 (10.9, 14.3)	0.426	13.0 (11.5,14.3)	12.8 (11.2,14.1)	0.311
Sodium, mmol/L	136 (133, 138)	136 (134, 139)	136 (133, 138)	0.087	136.5 (134,138)	136 (133,139)	0.277
Potassium, mmol/L	3.8(3.55,4.2)	3.8 (3.5, 4.1)	3.9(3.6,4.3)	0.001	3.8 (3.5, 4.1)	3.9(3.6,4.2)	0.061
BUN, mmol/L	15 (11, 25)	14 (11, 23)	18 (11, 32)	0.012	14 (11, 24)	17 (11, 29)	0.014
Pre-medication up to randomization, n(%)
Hydroxychloroquine	4 (0.9)	4(1.4)	0 (0.0)	0.374	2 (1.0)	2(0.9)	0.994
Remdesivir	23 (5.4)	19(6.7)	4 (2.8)	0.150	14 (7.1)	9(4.0)	0.240
Corticosteroids	33 (7.8)	23(8.2)	10 (7.1)	0.848	18 (9.1)	15(6.7)	0.456
Tocilizumab	4 (0.9)	0(0.0)	4 (2.8)	0.021	0 (0.0)	4(1.8)	0.167
Interferon	1 (0.2)	1(0.4)	0 (0.0)	0.997	1 (0.5)	0(0.0)	0.949
Azithromycin	134 (31.7)	91(32.3)	43(30.5)	0.796	62(31.3)	72(32.0)	0.963
Medication between randomization and hospital discharge, n(%)
Corticosteroids	82 (19.4)	40(14.2)	42(29.8)	<0.001	31(15.7)	51(22.7)	0.091
Tocilizumab	27 (6.4)	13(4.6)	14 (9.9)	0.058	6 (3.0)	21(9.3)	0.014
Azithromycin	82 (19.4)	53(18.8)	29(20.6)	0.761	33(16.7)	49(21.8)	0.229
Immunomodulating medication	4 (0.9)	1(0.4)	3 (2.1)	0.214	1 (0.5)	3(1.3)	0.708
Notes: M(IQR) for nonnormally distributed data, M ± SD for normally distributed data, and n (%) for categoric variables. a Moderate to severe kidney disease was defined as Cr >3, ESRD, chart diagnosis of CKD stage 5 (eGFR <15 mL/min/1.73m²) not on dialysis ^bMild liver disease present (without portal hypertension, includes chronic hepatitis) or Moderate/severe liver disease present (cirrhosis with portal HTN or variceal bleeding) ^cSolid tumor present (exclude if > 5 years from diagnosis) or Solid tumor with metastasis present (ever) Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease; ACEi, angiotensin-converting enzyme inhibitors; ARB, Angiotensin Receptor Blockers; SOFA, Sequential Organ Failure Assessment; BUN, blood urea nitrogen; SD: standard deviation; CV: Coefficient of Variation; BUN, blood urea nitrogen.

The median values of WBC count within five days post-randomization are shown in Figure 1. Overall, the median WBC value was within the normal range within five days post-randomization, with an elevated trend with increasing days. Specifically, WBC cell counts at Day 4 and Day 5 were significantly higher than those at Day 1 (P<0.05). Moreover, there was no significant association between baseline WBC count and risk of hospital discharge and death over 28 days, either in unadjusted or adjusted analysis (data not shown). We then assessed the post-randomization WBC variabilities from the SD and CV. The baseline characteristics of the patients stratified by distributions of SD and CV are described in Table 1. The WBC value was elevated in the higher WBC SD groups compared with the low SD groups but not in the CV categories.

Association between WBC variability and hospital discharge and death

Unadjusted Analyses

The numbers of outcome events over 28 days are summarized in Table 2. During the 28 days post-randomization, 46 (10.8%) patients died, and 345 (81.4%) patients were discharged.

Table 2

Association between white blood cell variability and 28-days mortality, hospital discharge in hospitalized patients with COVID-19
Outcomes	28-days-discharge (N=345)				28-days death (N=46)
Outcomes	Hazard ratios	95%CI		P value	Hazard ratios	95%CI		P value
Crude models
SD
Continuous analysis (Per 1 ×10³/mm3)	0.98	0.94	1.03	0.430	1.03	1.00	1.06	0.051
Low SD (<1.84×10³/mm3)	Ref				Ref
High SD (≥1.84×10³/mm3	0.50	0.40	0.64	<0.001	3.79	2.08	6.91	<0.001
CV
Continuous analysis (Per 5%)	0.99	0.96	1.02	0.330	1.08	1.03	1.13	0.001
Low CV (<17.6%)	Ref				Ref
High CV (≥17.6%)	0.62	0.50	0.77	<0.001	3.41	1.69	6.88	0.001
Adjusted models
Model 1^#
SD
Continuous analysis (Per 1 ×10³/mm3)	0.93	0.87	0.99	0.041	1.07	1.04	1.10	<0.001
Low SD (<1.84×10³/mm3)	Ref				Ref
High SD(≥1.84×10³/mm3)	0.49	0.38	0.62	<0.001	4.29	2.32	7.92	<0.001
CV
Continuous analysis (Per 5%)	.982	.954	1.012	0.234	1.062	1.016	1.110	0.008
Low CV (<17.6%)	Ref				Ref
High CV (≥17.6%)	0.62	0.51	0.77	<0.001	3.49	1.72	7.08	0.001
Model 2^&
SD
Continuous analysis (Per 1×10³/mm3)	0.97	0.93	1.02	0.26	1.09	1.04	1.14	<0.001
Low SD (<1.84×10³/mm3)	Ref				Ref
High SD(≥1.84×10³/mm3)	0.71	0.53	0.94	0.017	3.07	1.51	6.28	0.002
CV
Continuous analysis (Per 5%)	0.99	0.96	1.02	0.580	1.05	0.99	1.11	0.085
Low CV (<17.6%)	Ref				Ref
High CV (≥17.6%)	0.76	0.61	0.94	0.014	2.87	1.35	6.10	0.006
^# Model 1: age, sex
&Model 2: model1+ body mass index, receiving corticosteroids between randomization and hospital discharge, history of diabetes mellitus, history of cerebrovascular disease, history of hypertension, moderate to severe kidney disease, baseline White blood cell, total SOFA score, taking ACE inhibitors as a home medication;
Abbreviations: DM, Diabetes mellitus; HR, hazard ratio.; SD: standard deviation; CV: Coefficient of Variation; SOFA: Sequential Organ Failure Assessment

A lower incidence of discharge and higher incidences of death were observed in the higher parameters of the WBC variability group (P for log-rank <0.001). (Supplemental Figure S1). Both parameters of WBC variability were associated with worse clinical outcomes, including lower discharge rates (SD: HR: 0.50; P<0.001; CV: 0.62; P<0.001) and increased mortality (SD: HR: 2.08; P<0.001; CV: 1.69; P<0.001) (Table 2).

Adjusted Analyses

Patients with a higher WBC variability index experienced a lower age-adjusted incidence of hospital discharge and increased mortality over 28 days (Figure 2, P<0.001). When the WBC variability index was analyzed as a categorical variable, higher WBC variability was independently associated with a worse clinical outcome over 28 days, including decreased risk of hospital discharge (SD: HR: 2.32; P<0.001; CV: 1.72; P=0.001) and increased mortality (SD: HR: 1.51; P=0.002; CV: 1.35; P=0.006) (Table 2) after adjusting for sex, age, body mass index, corticosteroid use between randomization and hospital discharge, history of diabetes mellitus, history of cerebrovascular disease, history of hypertension, moderate to severe kidney disease, baseline WBC, increased WBC level, and total Sequential Organ Failure Assessment score. The age-adjusted dose-response relationship of WBC SD and CV with hospital discharge and death is shown in Supplemental Figure S2. There was a positive linear dose-response association between WBC variability indices and hospital discharge and death (P _nonlinearity>0.05). (Figure 3)

Discriminatory performances of the WBC variability

The discriminatory performances of the WBC variability for predicting clinical outcomes were expressed as the C-indices. As shown in Figure 4, WBC variability had moderate performance in predicting discharge (SD: 0.705, 95% CI 0.642–0.768; CV: 0.703, 95% CI: 0.620–0.786) and all-cause death through 28 days (SD: 0.632, 95% CI: 0.566–0.692; CV: 0.688, 95% CI: 0.601–0.775).

Subgroup and sensitive analyses

The results of the subgroup analyses for endpoints are described in Supplemental Tables S1 to S3. There were no significant interactions in the all-subgroup analyses based on sex (females versus males), treatment arm (hydroxychloroquine versus placebo), or age (age ≥65 years versus age < 65 years) (all P _interaction > 0.05).

Sensitivity analysis was performed by changing the outcomes as in-hospital death or composite of ECMO or mortality, and the results were similar to those of the primary analyses (Supplemental Table S4). The sensitivity analyses using the lowest WBC value instead of the highest WBC value to recalculate the WBC variability remained confirmatory (Supplemental Table S5). Furthermore, the competing regression model confirmed a decreased discharge rate with WBC variability (data not shown).

This study developed two WBC variability indices to explore the potential prognostic role of early fluctuation in WBC counts in hospitalized patients with COVID-19. Our results showed that i) both parameters of WBC variability indices were independently associated with lower discharge and increased mortality, ii) early WBC variability indices have a moderate predictive ability for hospital discharge and death over 28 days, and iii) sensitivity analyses stratified by sex, age, and treatment arm generated confirmatory results. To our knowledge, this is the first study that attempts to examine the association between WBC variability and clinical outcomes in patients with COVID-19.

Although no study has explored the relationship between WBC variability and prognosis to date, the association between variability in clinical measures and adverse outcomes has been extensively studied. For example, variability in heart rate, blood glucose, blood pressure, and body mass index were shown to be associated with increased adverse outcomes in the general population or those with comorbidities in a longitudinal cohort. Previous studies reported that WBC count, a simple systemic immune-inflammatory parameter, is a strong predictor for severity and increased mortality in patients with COVID-19. For example, Zhu and his colleagues observed that COVID-19 patients with higher WBC counts at admission had higher mortality[8]. A meta-analysis also showed that COVID-19 patients with an increased WBC (>10 × 10⁹/L) had an ~3-fold higher risk of developing severe disease. Our results extended the previous study[9]. We showed that WBC variability, as determined by SD and CV, is an independent predictor of clinical outcomes, which was confirmed in clinically relevant subgroups. It should be emphasized that the aim of our study was not to elaborate a prognostic model for the need for discharge or death in COVID-19 but to point out the possible influence of inflammation fluctuation reflected by WBC variability in the natural history of COVID-19.

An early case report from China showed that at the early stage of admission, the WBC count was normal or decreased, steeply increased on the 8th day, and reached a peak on the 14th day after admission[10]. Consistently, Liu observed that nearly 80% of patients had normal or decreased WBC counts[11]; however, those with higher WBC levels were at a higher risk of death. In line with the observational study[9, 12], our results showed that WBC was not significantly increased on the first day of admission or the subsequent four days and that the value of baseline WBC was not associated with adverse outcomes. Nevertheless, a positive association between two parameters of WBC variability (SD and CV) and worse clinical outcomes persisted even after adjusting for baseline WBC or increased WBC levels, suggesting an independent role of WBC variability that helps predict clinical outcomes in patients with COVID-19. The underlying mechanism warrants further investigation. SARS-CoV-2 infection is characterized by a dysregulated immune response and cytokine-release syndrome[13]. Peripheral blood immune-inflammatory parameters, such as WBC, will change significantly with the progression of COVID-19. In the early stage of COVID-19, although the WBC count was under the normal range, the dynamic changes in the WBC count might be related to the prevailing SARS-CoV-2 infection, the direct invasion of the virus into hematopoietic cells, and the dysregulated immune response. This result was supported by the correlation between dynamic changes in WBC counts and inflammatory cytokine levels[14] (e.g., IL-6 and IL-10). Therefore, we proposed that dramatic dynamic WBC count changes might be a direct indicator of dysregulated systemic inflammation and eventually contribute to adverse outcomes.

Clinical implication

Predicting the risk factors for COVID-19 is a challenging and major issue of current research. Our study showed that patients with a higher WBC variability index experienced a significantly lower incidence of hospital discharge and increased mortality over 28 days. Notably, the included subjects were all patients within 48 hours of diagnosis, and a WBC count was carried out immediately after inclusion, reflecting the early course of the COVID-19 disease and the degree of inflammatory response in the body. The WBC count is a very routine clinical examination in the hospital that is obtained shortly after hospital admission. Therefore, higher WBC variability at an early stage of admission may be used as an early independent indicator for poor prognosis. Close clinical surveillance and effective intervention might be considered for those with higher WBC variability plus established risk factors (e.g., diabetes, hypertension, cardiac or liver injury).

Strengths and Limitations

The current study must be interpreted within the context of its strength and potential limitations. Our study results are strengthened by the prospective design, moderate sample size, multivariable-adjusted models, various sensitivity analyses, and subgroup analyses, which consistently yielded confirmatory results and reported the robustness of the results. However, our study inevitably has several limitations. First, the optimal number and frequency of WBC cell counts to assess WBC variability are uncertain. Second, the use of antibiotic medication, dose or type of medication, and medication compliance may strongly correlate with WBC counts; however, this information was not available in our research. Third, our observations are from the US. Thus, our findings may not be generalizable to COVID-19 patients from other regions. Fourth, whether WBC variability can be correlated with inflammatory response fluctuations and the correlations between WBC variability and CRP and other gold-standard indicators for evaluating inflammation have not been tested. Finally, a meta-analysis revealed that compared with non-severe COVID-19 patients, the severe patients had significantly higher levels of CRP, IL-6, IL-10, ferritin, SAA, and NLR[15]. However, data on these inflammation indicators were missing in our data; as a result, we cannot evaluate their influence on our research subjects. Fifth, there are several different types of white blood cells. The relationship between WBC subtype variability and the clinical outcome of COVID-19 patients warrants further investigation.

In conclusion, our results show that COVID-19 patients with higher WBC variability have lower discharge rates and higher mortality. Our study contributes to our understanding of the relationship between WBC variability and adverse outcomes associated with COVID-19.

This research conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement[16].

Aim

The association between WBC variability and clinical outcomes in patients with COVID-19 is not clear. Thus, we conducted an observational study aiming to investigate whether WBC variability would be independently associated with clinical outcomes by analyzing data from the ORCHID trial.

Study population

This study population was from the Outcomes Related to COVID-19 Treated With Hydroxychloroquine Among Inpatients With Symptomatic Disease (ORCHID) trial, a multicenter, blinded, randomized clinical trial across 43 hospitals in the US. The information on the study design and main results have been previously reported[17]. Briefly, the ORCHID trial included 479 hospitalized patients with COVID-19 (< 48 h) confirmed with laboratory-confirmed SARS-CoV-2 positivity, with symptoms of respiratory illness for less than ten days. Patients were randomly assigned to hydroxychloroquine or placebo to assess the effect of hydroxychloroquine in adults hospitalized with COVID-19. The primary outcome was the clinical status 14 days after randomization, which was evaluated with a 7-category ordinal scale (the COVID Outcomes Scale). The secondary outcomes were the COVID Outcomes Scale and the clinical outcomes (e.g., 14-day and 28-day death rate or discharge, the composite of death or receipt of extracorporeal membrane oxygenation (ECMO) over a 28-day period, or ICU admission). Patients were followed up for 28 days following hydroxychloroquine randomization using in-hospital records and telephone follow-up after discharge. ORCHID complied with the Declaration of Helsinki and received ethical clearance. The present study used a dataset obtained from the National Institutes of Heart, Lung, and Blood Institute’s Biologic Specimen and Data Repository Information Coordinating Center. The data set in this trial was obtained from the National Heart, Lung, and Blood Institute. The investigators of the ORCHID trial were not involved in this study.

White blood cell measurements and calculation of variability

At 1–5 days after randomization, WBC cell counts were tested in hospitalized patients with COVID-19 confirmed less than 48 h. We selected the highest cell count values on each day, as multiple tests might be performed within one day. We assessed post-randomization variability from the SD (standard deviation) and CV (coefficient of variation) across at least 2 WBC tests. The CV was calculated using the following formula: (SD/Mean) * 100%.

Outcomes

In this study, the primary outcomes were hospital discharge and all-cause death within 28 days after randomization. The detailed definitions of these outcomes were based on the previous descriptions[17]. A clinical endpoint committee adjudicated all events.

Clinical follow-up

Patients were followed up for death for 28 days following trial randomization. Patient follow-up included the use of in-hospital records and follow-up telephone interviews for discharged patients at 14 and 28 days after trial randomization.

Statistical analysis

The vital status was unavailable for two patients during the follow-up. Forty-two patients with an unavailable value of WBC variability were further excluded. Finally, 424 patients were included in this study.

Baseline characteristics for continuous variables with normal or nonmoral distributions are expressed as the means with standard deviations or medians with interquartile ranges, respectively. Comparisons between the groups were examined using unpaired Student’s t-test (normal distribution) or the Wilcoxon-Mann–Whitney tests (nonmoral distribution) for continuous variance. Categorical variables with a normal distribution or nonnormal distribution were compared using the χ2 test or the Kruskal–Wallis test. The best cutoff points of WBC variability (SD and CV) for endpoints were explored using regression tree analysis. We further plotted the crude or age-adjusted cumulative incidence of endpoints across categories of the indices of WBC variability using the log-rank test or Cox models, respectively. Cox proportional hazards models were used to compute the adjusted risk estimates (aka, hazard ratios [HRs] and their confidence intervals [CIs]). We developed two models (Model 1 and Model 2) of proportional hazards regression analyses. Model 1 was adjusted for age and sex. Model 2 included more confounding variables based on stepwise analysis (backward). We evaluated the discriminatory ability of WBC variability for predicting outcomes using the C-index, which was calculated through the time-dependent (at follow-up at 28 days) area under the receiver operating characteristic curve. All statistical analyses were performed using SPSS Statistics (IBM Corp., Armonk, NY, USA) and R software (R Project for Statistical Computing, Vienna, Austria). A two-sided P value less than 0.05 was considered statistically significant.

Supplementary analyses

We examined whether the association of clinical outcomes with indices of WBC variability differed by subgroup (randomized for treatments, sex, and age).

Sensitivity analyses were also performed; 1) by limiting the definition of the 28-day mortality to the in-hospital mortality; 2) further analyzing the composite endpoint, which was the in-hospital death rate or patients who received ECMO; 3) recalculating the WBC variability by changing the highest WBC values to the lowest WBC values within one day; 4) using a competing regression model to assess the association of WBC variability with 28-day discharge (Fine and Gray test) where all-cause death was defined as a competing event.

WBC: white blood cell

COVID-19: Coronavirus disease 2019

SD: standard deviation

CV: coefficient of variation

HR: hazard ratio

SARS-CoV-2: severe acute respiratory syndrome coronavirus 2

NLR: neutrophil-to-lymphocyte ratio

ORCHID: Outcomes Related to COVID-19 Treated With Hydroxychloroquine Among Inpatients With Symptomatic Disease

STROBE: Strengthening the Reporting of Observational Studies in Epidemiology

ECMO: extracorporeal membrane oxygenation

CI: confidence interval

Ethics approval and consent to participate

Not applicable

Consent for publication

Not applicable

Availability of data and materials

All data generated or analyzed during this study are included in this published article [and its supplementary information files].

Competing interests

The authors declare that they have no competing interests

Funding

Sources of funding: This work was supported in part by the Natural Science Foundation in Jiangxi Province grant (No.202002BAB216022 to J.Z., No.20192ACBL21037 and No.202004BCJL23049 to P.Y.), the National Natural Science Foundation of China (No. 82160371 to J.Z.; No. 82100869 to P.Y. and 82100347 to X.L). All fundings had no role in design, methods, subject recruitment, data collections, analysis, and preparation of the paper.

Authors’ contributions

X-L was responsible for the entire project and revised the draft. P-Y and PP-X performed the data extraction and drafted the first version of the manuscript. ZW-L, J-Z, and JY-M interpreted the data and revised the manuscript. MX-X, YC-Z, and YF-S performed the statistical analysis. All authors participated in the interpretation of the results, prepared the final version of the manuscript, and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated resolved.

Acknowledgments

We thank (ORCHID) Outcomes Related to COVID-19 Treated With Hydroxychloroquine Among Inpatients With Symptomatic Disease investigators for conducting this trial and making these data available.

We acknowledge all healthcare workers involved in the fight against COVID-19 worldwide.

Hooftman A, Angiari S, Hester S, Corcoran SE, Runtsch MC, Ling C, Ruzek MC, Slivka PF, McGettrick AF, Banahan K, et al: The Immunomodulatory Metabolite Itaconate Modifies NLRP3 and Inhibits Inflammasome Activation. Cell Metab 2020, 32:468-478.e467.
Chen R, Sang L, Jiang M, Yang Z, Jia N, Fu W, Xie J, Guan W, Liang W, Ni Z, et al: Longitudinal hematologic and immunologic variations associated with the progression of COVID-19 patients in China. The Journal of allergy and clinical immunology 2020, 146.
Qin C, Zhou L, Hu Z, Zhang S, Yang S, Tao Y, Xie C, Ma K, Shang K, Wang W, Tian D-S: Dysregulation of Immune Response in Patients With Coronavirus 2019 (COVID-19) in Wuhan, China. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2020, 71:762-768.
Lv Z, Cheng S, Le J, Huang J, Feng L, Zhang B, Li Y: Clinical characteristics and co-infections of 354 hospitalized patients with COVID-19 in Wuhan, China: a retrospective cohort study. Microbes and infection 2020, 22:195-199.
Wang Y, Lu X, Li Y, Chen H, Chen T, Su N, Huang F, Zhou J, Zhang B, Yan F, Wang J: Clinical Course and Outcomes of 344 Intensive Care Patients with COVID-19. American journal of respiratory and critical care medicine 2020, 201:1430-1434.
Yan X, Li F, Wang X, Yan J, Zhu F, Tang S, Deng Y, Wang H, Chen R, Yu Z, et al: Neutrophil to lymphocyte ratio as prognostic and predictive factor in patients with coronavirus disease 2019: A retrospective cross-sectional study. Journal of medical virology 2020, 92:2573-2581.
Lu G, Wang J: Dynamic changes in routine blood parameters of a severe COVID-19 case. Clinica chimica acta; international journal of clinical chemistry 2020, 508.
Zhu B, Feng X, Jiang C, Mi S, Yang L, Zhao Z, Zhang Y, Zhang L: Correlation between white blood cell count at admission and mortality in COVID-19 patients: a retrospective study. BMC infectious diseases 2021, 21:574.
Feng X, Li S, Sun Q, Zhu J, Chen B, Xiong M, Cao G: Immune-Inflammatory Parameters in COVID-19 Cases: A Systematic Review and Meta-Analysis. Front Med (Lausanne) 2020, 7:301.
Lu G, Wang J: Dynamic changes in routine blood parameters of a severe COVID-19 case. Clin Chim Acta 2020, 508:98-102.
Liu K, Fang YY, Deng Y, Liu W, Wang MF, Ma JP, Xiao W, Wang YN, Zhong MH, Li CH, et al: Clinical characteristics of novel coronavirus cases in tertiary hospitals in Hubei Province. Chin Med J (Engl) 2020, 133:1025-1031.
Wang D, Hu B, Hu C, Zhu F, Liu X, Zhang J, Wang B, Xiang H, Cheng Z, Xiong Y, et al: Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus-Infected Pneumonia in Wuhan, China. JAMA 2020.
Gupta A, Madhavan MV, Sehgal K, Nair N, Mahajan S, Sehrawat TS, Bikdeli B, Ahluwalia N, Ausiello JC, Wan EY, et al: Extrapulmonary manifestations of COVID-19. Nat Med 2020, 26:1017-1032.
Liu J, Li S, Liu J, Liang B, Wang X, Wang H, Li W, Tong Q, Yi J, Zhao L, et al: Longitudinal characteristics of lymphocyte responses and cytokine profiles in the peripheral blood of SARS-CoV-2 infected patients. EBioMedicine 2020, 55:102763.
Mahat RK, Panda S, Rathore V, Swain S, Yadav L, Sah SP: The dynamics of inflammatory markers in coronavirus disease-2019 (COVID-19) patients: A systematic review and meta-analysis. Clinical epidemiology and global health 2021, 11:100727.
Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Annals of internal medicine 2007, 147:573-577.
Self WH, Semler MW, Leither LM, Casey JD, Angus DC, Brower RG, Chang SY, Collins SP, Eppensteiner JC, Filbin MR, et al: Effect of Hydroxychloroquine on Clinical Status at 14 Days in Hospitalized Patients With COVID-19: A Randomized Clinical Trial. JAMA 2020, 324:2165-2176.

Supplementalmaterial.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

White Blood Cell Variability and Clinical Outcomes in Hospitalized Patients With COVID-19

Status:

Version 1

Abstract

Figures

Background

Results

Baseline characteristics

Association between WBC variability and hospital discharge and death

Unadjusted Analyses

Adjusted Analyses

Discriminatory performances of the WBC variability

Subgroup and sensitive analyses

Discussion

Clinical implication

Strengths and Limitations

Conclusion

Methods

Aim

Study population

White blood cell measurements and calculation of variability

Outcomes

Clinical follow-up

Statistical analysis

Supplementary analyses

Abbreviations

Declarations

References

Supplementary Files

Status:

Version 1