The usefulness of D-dimer as a predictive marker for mortality in patients with COVID-19 hospitalized during the first wave in Italy

Shermarke Hassan; Barbara Ferrari; Raffaella Rossio; Vincenzo la Mura; Andrea Artoni; Roberta Gualtierotti; Ida Martinelli; Alessandro Nobili; Alessandra Bandera; Andrea Gori; Francesco Blasi; Valter Monzani; Giorgio Costantino; Sergio Harari; Frits Richard Rosendaal; Flora Peyvandi; on behalf of the COVID-19 Network working group

doi:10.1371/journal.pone.0264106

. 2022 Jul 22;17(7):e0264106. doi: 10.1371/journal.pone.0264106

The usefulness of D-dimer as a predictive marker for mortality in patients with COVID-19 hospitalized during the first wave in Italy

Shermarke Hassan ^1,², Barbara Ferrari ³, Raffaella Rossio ³, Vincenzo la Mura ^1,³, Andrea Artoni ⁴, Roberta Gualtierotti ^1,⁴, Ida Martinelli ⁴, Alessandro Nobili ⁵, Alessandra Bandera ^1,⁶, Andrea Gori ^1,⁶, Francesco Blasi ^1,⁷, Valter Monzani ⁸, Giorgio Costantino ^9,¹⁰, Sergio Harari ^9,¹¹, Frits Richard Rosendaal ², Flora Peyvandi ^1,^3,^*; on behalf of the COVID-19 Network working group^¶

Editor: Massimo Filippi¹²

¹Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy

²Department of Clinical Epidemiology, Leiden University Medical Center, Leiden, the Netherlands

³U.O.C. Medicina Generale Emostasi e Trombosi, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy

⁴Angelo Bianchi Bonomi Hemophilia and Thrombosis Centre, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy

⁵Department of Neuroscience, Istituto di Ricerche Farmacologiche Mario Negri IRCCS, Milan, Italy

⁶Infectious Disease Unit, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy

⁷Respiratory Unit and Cystic Fibrosis Adult Center, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy

⁸Acute Medical Unit, Department of Medicine, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy

⁹Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy

¹⁰Emergency Department, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy

¹¹UO di Pneumologia e Terapia Semi-Intensiva Respiratoria, Servizio di Fisiopatologia Respiratoria ed Emodinamica Polmonare, MultiMedica IRCCS, Milan, Italy

¹²IRCCS San Raffaele Scientific Research Institute, ITALY

Competing Interests: B. Ferrari has received consulting fees and travel support from Sanofi Genzyme. R. Gualtierotti reports participation in advisory boards for Biomarin, Pfizer, Bayer and Takeda as well as participation at educational seminars sponsored by Pfizer, Sobi and Roche. I. Martinelli reports personal and non-financial support from Bayer, Roche, Rovi and Novo Nordisk outside of the submitted work. A. Gori has received grants for research support, honoraria, consultation fees, and travel support from Gilead, Janssen, MSD, Pfizer, Angelini, Menarini, ViiV. F. Peyvandi has received honoraria for participating as a speaker at educational meetings, symposia and advisory boards of Roche, Sobi, Sanofi, Grifols and Takeda. All other authors have no conflicts of interest to disclose. This does not alter our adherence to PLOS ONE policies on sharing data and materials.

¶ The complete membership of the author group can be found in the Supplement

^✉

* E-mail: flora.peyvandi@unimi.it

Roles

Shermarke Hassan: Data curation, Formal analysis, Methodology, Writing – original draft, Writing – review & editing

Barbara Ferrari: Conceptualization, Data curation, Writing – review & editing

Raffaella Rossio: Conceptualization, Data curation, Writing – review & editing

Vincenzo la Mura: Conceptualization, Data curation, Writing – review & editing

Andrea Artoni: Conceptualization, Data curation, Writing – review & editing

Roberta Gualtierotti: Conceptualization, Data curation, Writing – review & editing

Ida Martinelli: Conceptualization, Data curation, Writing – review & editing

Alessandro Nobili: Conceptualization, Data curation, Writing – review & editing

Alessandra Bandera: Conceptualization, Data curation, Writing – review & editing

Andrea Gori: Conceptualization, Data curation, Writing – review & editing

Francesco Blasi: Conceptualization, Data curation, Writing – review & editing

Valter Monzani: Conceptualization, Data curation, Writing – review & editing

Giorgio Costantino: Conceptualization, Data curation, Writing – review & editing

Sergio Harari: Conceptualization, Data curation, Writing – review & editing

Frits Richard Rosendaal: Conceptualization, Data curation, Writing – review & editing

Flora Peyvandi: Conceptualization, Data curation, Writing – review & editing

Massimo Filippi: Editor

PMCID: PMC9307169 PMID: 35867647

Abstract

Background

The coronavirus disease 2019 (COVID-19) presents an urgent threat to global health. Identification of predictors of poor outcomes will assist medical staff in treatment and allocating limited healthcare resources.

Aims

The primary aim was to study the value of D-dimer as a predictive marker for in-hospital mortality.

Methods

This was a cohort study. The study population consisted of hospitalized patients (age >18 years), who were diagnosed with COVID-19 based on real-time PCR at 9 hospitals during the first COVID-19 wave in Lombardy, Italy (Feb-May 2020). The primary endpoint was in-hospital mortality. Information was obtained from patient records. Statistical analyses were performed using a Fine-Gray competing risk survival model. Model discrimination was assessed using Harrell’s C-index and model calibration was assessed using a calibration plot.

Results

Out of 1049 patients, 507 patients (46%) had evaluable data. Of these 507 patients, 96 died within 30 days. The cumulative incidence of in-hospital mortality within 30 days was 19% (95CI: 16%-23%), and the majority of deaths occurred within the first 10 days. A prediction model containing D-dimer as the only predictor had a C-index of 0.66 (95%CI: 0.61–0.71). Overall calibration of the model was very poor. The addition of D-dimer to a model containing age, sex and co-morbidities as predictors did not lead to any meaningful improvement in either the C-index or the calibration plot.

Conclusion

The predictive value of D-dimer alone was moderate, and the addition of D-dimer to a simple model containing basic clinical characteristics did not lead to any improvement in model performance.

Introduction

The coronavirus disease 2019 (COVID-19) is an urgent threat to global health that has severely strained the healthcare system of many countries [1–3]. Since the outbreak in early December 2019, the number of patients confirmed to have the disease has exceeded 521,563,472 and the number of people infected is probably much higher. More than 6,264,178 people have died from COVID-19 infection (up to May 16^th 2022) [4].

Due to a large number of COVID-19 patients overwhelming the Italian healthcare system during the first COVID-19 wave in Lombardy, Italy (Feb-May 2020), it was important to look for biomarkers measured at admission that could predict mortality and other in-hospital adverse outcomes, in order to better triage patients.

D‐dimer is a fibrin degradation product, which originates from the formation and lysis of cross‐linked fibrin and reflects activation of coagulation and fibrinolysis. Among the clinical and biochemical parameters associated with poor prognosis, increased D-dimer levels seemed to be predictive for acute respiratory distress syndrome (ARDS), the need for admission to an intensive care unit (ICU) or death [5,6]. Furthermore, several studies have reported an increased incidence of thromboembolic events in hospitalized COVID-19 patients [7].

Taken together, these early studies indicate that D-dimer values at admission might be used to determine which patients would require hospitalization. Patients predicted to have a low enough mortality risk wouldn’t need to be hospitalized, thereby decreasing the burden on the healthcare system.

Therefore, the primary aim of this paper was to study the predictive value of D-dimer levels at admission for in-hospital mortality. The secondary aim to assess if there was any causal relationship between D-dimer levels and in-hospital mortality.

Methods

Study design and population

This was an observational cohort study. The study population consisted of patients aged > 18 years who were hospitalized and who were positive for COVID-19 based on real-time PCR at 9 Italian hospitals, during the first COVID-19 wave in Lombardy, Italy (Feb-May 2020). Patients that were directly admitted to the ICU were excluded. Patients in this study were followed-up for 30 days from hospital admission.

This observational study was approved by the Medical Ethics Committee of the Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico. The need to obtain informed consent was waived by the Medical Ethics Committee in cases where it was not possible to obtain informed consent, due to severe illness or death. In all other cases, written informed consent was obtained.

Data collection and definition of variables

All information was obtained from electronic patient records, using a standardized case report form. The exposure of interest, D-dimer levels (expressed as ng/mL) was used as either a continuous variable or a categorical variable depending on the analysis of interest. When showing descriptive statistics and estimating the association between D-dimer levels and in-hospital mortality, D-dimer was converted to a categorical variable with 4 levels that correspond to the 1^st, 2^nd, 3^rd and 4^th quartile of D-dimer levels. This was done to make the results easier to interpret for the reader. D-dimer was analyzed as a continuous variable when assessing the predictive value of D-dimer levels at admission for in-hospital mortality. This is because categorizing a variable always leads to some loss in predictive power. The outcome used for all analyses was in-hospital mortality.

The following patient- and treatment characteristics were obtained: age (continuous variable), sex (dichotomous variable: male, female), the use of anticoagulant therapy during the study (dichotomous variable: yes, no) and the number of days between symptom onset and hospital admission (continuous variable). Lastly, the number of comorbidities (continuous variable) that each patient had at admission was calculated. Comorbidities to be assessed were selected based on their usage in the Charlson comorbidity index [8], a well-known risk score used to predict 10-year survival in patients with several comorbidities. To this list of comorbidities, we also added obesity (as defined by the clinician). The final list of comorbidities used was as follows: cardiovascular disease, chronic obstructive pulmonary disease, chronic kidney disease, diabetes mellitus, cancer, liver disease, dementia, connective tissue disease, acquired immunodeficiency syndrome (AIDS) and clinician-defined obesity.

Statistical analysis: General approach

Descriptive analyses were reported as mean/standard deviation (SD), median/interquartile range (IQR), or as proportions. The cumulative incidence of in-hospital mortality and the relationship between D-dimer levels and in-hospital mortality were assessed using survival analysis methods. Discharge within 30 days with a good prognosis served as a competing outcome, in that it almost completely precludes the occurrence of the main outcome of interest (in-hospital mortality). Therefore, it was decided to model the relationship between D-dimer and in-hospital mortality using the Fine-Gray competing risk survival model, which accounts for the presence of competing events. In the multivariable analyses, we adjusted for age, sex and the number of comorbidities. For similar reasons, we did not use the Kaplan Meier function to estimate the cumulative incidence of mortality. Instead, we used the cumulative incidence function, which correctly accounts for competing events. A complete case analysis was performed, meaning that patients with missing values for exposure, outcome or confounders were removed.

Statistical analysis: Causal relationship between D-dimer levels and in-hospital mortality

The relationship between D-dimer levels (modeled as a categorical variable as described above) and in-hospital mortality was estimated using the aforementioned Fine-Gray competing risk survival model. We adjusted for age, sex and comorbidities, as well as anticoagulant therapy during hospitalization and the time between symptom onset and hospital admission.

Statistical analysis: The predictive value of D-dimer for in-hospital mortality

D-dimer was modeled as a continuous variable for all analyses regarding the predictive value of D-dimer for in-hospital mortality. Three prediction models were tested: a model containing only D-dimer as a predictor, a model containing age, sex and comorbidities, and a model containing D-dimer, age, sex and comorbidities. For all three models, model discrimination and model calibration were assessed.

Model discrimination refers to how well a model can discriminate between patients with and without the outcome. Model discrimination was measured by calculating a modified version of Harrel’s C-index [9], which is a measure of how well the model can discriminate between patients with and without the outcome. In the presence of competing risks, Harrel’s C-index is biased [10]. We calculated a modified version of Harrel’s C-index by setting the follow-up time of patients who experience a competing event to larger than our prediction horizon (which is 30 days), instead of censoring these patients, as was proposed by Wolbers et al. [10].

Model calibration refers to the degree to which the risk of mortality predicted by a model and the actual observed mortality rate in a group of patients are similar. For example, if the predicted mortality risk for a group of 100 patients is equal to 12% than the observed proportion of patients that died should also be around 12%. If the predicted mortality risk is much higher or lower than observed mortality rate, than the model is miscalibrated. To assess model calibration, we first fitted a model to the data, and then calculated the individual predicted mortality risk for each patient using this model. We then divided the patient population into ten groups (deciles) based on their predicted mortality risk. For each group, the predicted mortality risk for the whole group was compared with the observed mortality rate in that same group. This was done visually in a scatterplot that shows the predicted mortality risk on the X-axis and the observed mortality on the Y-axis for each decile. Furthermore, to examine calibration across the whole range, we also fitted a LOWESS (Locally Weighted Scatterplot Smoothing) line to the data.

Sample size calculation

A formal sample size calculation for the development of a prediction model was not performed. However, 96 patients died during follow-up (see Results section) and the number of predictors used in the prediction models ranged from 1 (for the model containing only D-dimer as a predictor) to 4 (for the model containing D-dimer, age, sex and comorbidities as predictors). Accordingly, the number of events per predictor ranged from 96 to 24, well above the minimum of 10 events per variable needed to accurately estimate the model coefficients [11]. Therefore, we deemed the sample size sufficient for these analyses.

Results

Baseline characteristics

Out of 1094 patients, 506 patients had missing D-dimer levels, 27 patients had incomplete follow-up data, 13 patients were excluded due to immediately being admitted to the ICU after admission and 41 patients were excluded due to missing data for one or more confounding factors. Finally, 507 patients (46%) had evaluable data. Of these, 96 patients died within 30 days after admission. Patients were enrolled between March 6^th 2020 and September 20^th 2020. Almost all (98%) patients were enrolled before May 31^st 2020, which roughly corresponds to the first three months of the initial COVID-19 epidemic in Italy [12]. D-dimer values were associated with advanced age and the number of comorbidities at admission (Table 1).

Table 1. Baseline characteristics of COVID-19 patients hospitalized in the region of Lombardy, Italy, during the first COVID-19 wave (Feb-May 2020).

Variables	D-dimer, < 538 ng/mL (N = 128)	D-dimer, 538–957 ng/mL (N = 125)	D-dimer, 957–1764 ng/mL (N = 127)	D-dimer, > 1764 ng/mL (N = 127)	Overall (N = 507)
mean age (SD)	57.3 (16.4)	62.3 (14.2)	67.5 (14.1)	69.3 (13.6)	64.1 (15.3)
sex
female	46 (35.9%)	31 (24.8%)	47 (37.0%)	52 (40.9%)	176 (34.7%)
male	82 (64.1%)	94 (75.2%)	80 (63.0%)	75 (59.1%)	331 (65.3%)
charlson comorbidity index
no comorbidities	68 (53.1%)	70 (56.0%)	64 (50.4%)	52 (40.9%)	254 (50.1%)
1 comorbidity	41 (32.0%)	24 (19.2%)	34 (26.8%)	33 (26.0%)	132 (26.0%)
2 comorbidities	15 (11.7%)	21 (16.8%)	14 (11.0%)	28 (22.0%)	78 (15.4%)
3 or more comorbidities	4 (3.1%)	10 (8.0%)	15 (11.8%)	14 (11.0%)	43 (8.5%)
anticoagulant therapy during hospitalization
no	28 (21.9%)	25 (20.0%)	18 (14.2%)	25 (19.7%)	96 (18.9%)
yes	100 (78.1%)	100 (80.0%)	109 (85.8%)	102 (80.3%)	411 (81.1%)
mean number of days between first symptoms and admission (SD)	8.9 (6.5)	10.6 (12.3)	11.7 (12.6)	11.2 (10.5)	10.6 (10.8)

Open in a new tab

SD: Standard deviation.

From March 6^th 2020 to September 20^th 2020, the cumulative incidence of in-hospital mortality within 30 days was 19% (95CI: 16%-23%). Mortality was higher in the early phase of the epidemic and slightly decreased over time. (S1 Table) The cumulative incidence of discharge because of a good prognosis within 30 days was 72% (95CI: 68%-75%) (Fig 1) and 75% of deaths occurred in the first 10 days. After this period the death rate slowed down, as evidence by the flattening of the survival curve (Fig 1).

With increasing D-dimer levels, the absolute risk of mortality also increased strongly, from 4% (95CI:2%-9%) in patients with D-dimer levels in the lowest quartile to 28% (95CI: 20%-36%) in patients with D-dimer levels in the highest quartile (Table 2).

Table 2. Association between D-dimer values and in-hospital mortality.

	N	n	Observed incidence after 30 days	Univariate model	Multivariable model 1^a	Multivariable model 2^b
D-dimer
< 538 ng/mL	128	5	0.04 (0.02–0.09)	ref	ref	ref
538–957 ng/mL	125	21	0.17 (0.11–0.24)	4.5 (1.7–11.8)	3.9 (1.5–9.9)	4.0 (1.6–10.2)
957–1764 ng/mL	127	35	0.28 (0.21–0.36)	8.2 (3.2–20.8)	5.2 (2.0–13.1)	5.4 (2.1–13.8)
> 1764 ng/mL	127	35	0.28 (0.20–0.36)	7.8 (3.1–19.6)	4.3 (1.7–10.7)	4.5 (1.8–11.5)

Open in a new tab

^a Model, corrected for age, sex and Charlson comorbidity index score.

^b Model, corrected for age, sex, Charlson comorbidity index score, anticoagulant therapy during hospitalization, and the time between symptom onset and hospital admission.

Causal relationship between D-dimer levels and in-hospital mortality

Compared with patients in the lowest quartile of D-dimer blood concentration, the unadjusted hazard ratio for in-hospital mortality in patients in the 2^nd, 3^rd and 4^th quartile was 4.5 (95CI: 1.7–11.8), 8.2 (95CI: 3.2–20.8) and 7.8 (95CI: 3.1–19.6) respectively. (Table 2) After adjusting for age, sex, comorbidities, anticoagulant therapy during hospitalization and the time between symptom onset and hospital admission, the hazard ratio for patients in the 2^nd, 3^rd and 4^th quartile was 4.0 (95CI: 1.6–10.2), 5.4 (95CI: 2.1–13.8), and 4.5 (95CI: 1.8–11.5) respectively (Table 2).

The predictive value of D-dimer for in-hospital mortality

The predictive model containing D-dimer as the only predictor had a C-index of 0.66 (95%CI: 0.61–0.71). Overall calibration of the model was very poor (Fig 2A). Next, the predictive model containing age, sex and comorbidities as predictors had a C-index of 0.83 (95%CI: 0.79–0.87). Overall calibration of the model was acceptable (Fig 2B). Lastly, the predictive model containing D-dimer, age, sex and comorbidities as predictors had a C-index of 0.83 (95%CI: 0.80–0.87). Overall calibration of this model was acceptable (Fig 2C).

Discussion

Our results show that despite a strong correlation between D-dimer levels and mortality, the predictive value of D-dimer as a single biomarker was unclear. Model discrimination was moderate (C-index: 0.66) while model calibration was very poor. Furthermore, the addition of D-dimer to a simple model containing only basic clinical characteristics (age, sex and co-morbidities) did not lead to any meaningful improvement in either the C-index or the calibration plot.

D-dimer is a breakdown product, generated after a fibrin clot is degraded by fibrinolysis. It is a recognized valid lab biomarker that is widely used as part of the diagnostic workup of patients with a suspected venous thromboembolism or disseminated intravascular coagulation and is predictive of poor outcomes and thromboembolic events [13]. Changes in D-dimer levels are seen in most patients that are hospitalized with COVID-19 [14]. Changes in other hemostatic parameters, such as a slightly elongated PT, elongated aPTT, or mild thrombocytopenia are less common [15]. Furthermore, in addition to the increase in D-dimer (which is also an acute-phase protein that rises with general inflammation) an increase in inflammatory biomarkers such as CRP, particularly in COVID-19 patients with a more severe disease phenotype, is also seen [16].

Mechanisms underlying this COVID-19-induced coagulopathy may, in part, be explained by the same general mechanisms that also underlie other cases of bacteria-induced septic coagulopathy such as overproduction of pro-inflammatory cytokines by monocytes [17]. Furthermore, direct activation of coagulation by monocytes via tissue-factor and phosphatidylserine (which are expressed on the cell surface of monocytes) also play a role [16]. Furthermore, studies have reported endothelial dysfunction in patients with COVID-19 induced coagulopathy, which is probably mediated by the production of pro-inflammatory cytokines as well as activation of the complement cascade [18,19].

A strong correlation between D-dimer and mortality was also reported by other studies. A meta-analysis of six studies containing 1355 hospitalized patients found that D-dimer levels were higher in deceased patients (standardized mean difference: 3.59 mcg/L, 95%CI 2.79–4.40) [20]. This meta-analysis did not calculate a pooled C-index to assess the overall predictive performance of D-dimer.

A later meta-analysis reporting on 16 studies containing 4468 COVID-19 patients reported a pooled C-index of 0.86 (95CI: 83–89) for predicting all-cause mortality [21]. However, these results were most likely strongly influenced by publication bias. In addition, it is somewhat unclear how the pooled C-index was calculated, as many studies did not report the C-index directly.

A retrospective study by Zhang et al. evaluated D-dimer levels and mortality in 343 patients [22]. D-dimer levels were measured within the first 24 hours, and hospitalized patients were followed until death or discharge. The study showed a very strong correlation between D-dimer levels over 2.0 mcg/mL and mortality. (HR: 51.5, 95%CI 12.9–206.7). However, the study did not adjust for any confounders. It is therefore unclear how different confounders could have affected the reported results. The predictive value of D-dimer was also very high (C-index: 0.89) There was no information about anticoagulant use during the study follow-up.

These strong results were not confirmed by a later study by Naymagon et al. [23] that followed 1062 COVID-19 patients during hospitalization. Each 1 μg/ml increase in D-dimer levels (measured within 3 days from admission) was associated with a hazard ratio of death of 1.05 (95%CI: 1.04–1.07). The association did not change after adjustment for age, smoking, Charlson comorbidity index and anticoagulant use at admission. However, discriminative performance of D-dimer levels was moderate (C-index: 0.694). At baseline, 9.1% of patients were on anticoagulants and no information was given about thromboprophylaxis during the study.

Overall, it seems that early studies reported that D-dimer was strongly predictive of mortality, although this effect was not as strong in the larger studies. Furthermore, all aforementioned studies only assessed the discriminative performance of D-dimer, but not model calibration.

Our study has several strengths. Firstly, our study had a large sample size with a sufficient number of events. Secondly, we applied a competing risk survival model to analyze the relationship between D-dimer and poor outcomes to avoid bias. Not taking competing risks into account could lead to misleading results, as shown in a recent simulation study on competing risks in COVID-19 research [24]. In addition, we evaluated both discrimination (as was done in earlier studies) and model calibration (which was not reported in any of the aforementioned studies). This is important because models may show good discrimination but could still be poorly calibrated [25].

Our study also has some limitations. The main limitation is that values for D-dimer levels were not available for 506 out of 1094 patients. D-dimer tests are most commonly ordered if a patient has some symptoms or medical history which are indicative of a thromboembolic event. Therefore, patients that were excluded from the study due to missing information on D-dimer were most likely patients with a low a priori likelihood of having a venous thromboembolic event. Also, D-dimer assays vary widely in their set-up. This lack of standardization makes comparison of different study results somewhat difficult [26,27].

Due to the rapid pace of change in the treatment of patients with COVID-19, the predictive value of D-dimer (and therefore, it’s clinical usefulness) will most likely have diminished over time. For example, prophylactic anticoagulation for hospitalized COVID-19 patients became much more common over time. Furthermore, as the outbreak went on, patients with milder symptoms were also being hospitalized. Due to these treatment changes, we can speculate that patients hospitalized after the first COVID-19 wave will have had lower D-dimer levels at admission, when compared to patients admitted in the first COVID-19 wave (Feb-May 2020). Furthermore, D-dimer levels would have been less strongly associated with mortality in these patients, when compared to patients admitted in the first COVID-19 wave (Feb-May 2020).

As shown before, a part of COVID-19 related mortality is due to an underlying coagulopathy. (which might manifest as a venous thromboembolic event, as disseminated intravascular coagulation or as thrombotic microangiopathy) Consequently, some studies have suggested that D-dimer levels could be used to stratify patients with COVID-19, and to individualize treatment [22]. However, our analyses show that, despite a strong correlation between D-dimer levels and mortality, the predictive value of D-dimer alone was not sufficient. However, that was to be expected as COVID-19 is not a coagulation disorder but a multi-systemic (although mainly respiratory) disease that influences health through multiple pathways, one of which is the coagulation system. However, D-dimer also showed little added value when added to simple risk prediction model containing only age, sex and comorbidities as predictors.

Conclusions

The predictive value of D-dimer alone was moderate, and the addition of D-dimer to a simple model containing basic clinical characteristics did not lead to any improvement in model performance.

Supporting information

S1 Table. Cumulative incidence of death, per time-period.

(DOCX)

Click here for additional data file.^{(12.1KB, docx)}

S1 List. List of participants in COVID-19 Network working group.

(DOCX)

Click here for additional data file.^{(14.8KB, docx)}

S1 File. File containing the data and code used to generate the results.

(ZIP)

Click here for additional data file.^{(20.8KB, zip)}

Acknowledgments

We thank the COVID-19 Network working group (contact: dr. Alessandro Nobili, alessandro.nobili@marionegri.it) for their help with patient recruitment and data collection. The full member list can be found in the Supplement.

Data Availability

All relevant data are within the paper and its Supporting Information files.

Funding Statement

The authors received no specific funding for this work.

References

1.Arabi YM, Murthy S, Webb S. COVID-19: a novel coronavirus and a novel challenge for critical care. Intensive Care Med. 2020/03/04. 2020;46: 833–836. doi: 10.1007/s00134-020-05955-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Grasselli G, Pesenti A, Cecconi M. Critical Care Utilization for the COVID-19 Outbreak in Lombardy, Italy: Early Experience and Forecast During an Emergency Response. Jama. 2020/03/14. 2020. doi: 10.1001/jama.2020.4031 [DOI] [PubMed] [Google Scholar]
3.Xie J, Tong Z, Guan X, Du B, Qiu H, Slutsky AS. Critical care crisis and some recommendations during the COVID-19 epidemic in China. Intensive Care Med. 2020/03/04. 2020;46: 837–840. doi: 10.1007/s00134-020-05979-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Dong E, Du H, Gardner L. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect Dis. 2020/02/23. 2020;20: 533–534. doi: 10.1016/S1473-3099(20)30120-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Wu Z, McGoogan JM. Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for Disease Control and Prevention. Jama. 2020/02/25. 2020. doi: 10.1001/jama.2020.2648 [DOI] [PubMed] [Google Scholar]
6.Tang N, Li D, Wang X, Sun Z. Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia. J Thromb Haemost. 2020/02/20. 2020;18: 844–847. doi: 10.1111/jth.14768 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Al-Ani F, Chehade S, Lazo-Langner A. Thrombosis risk associated with COVID-19 infection. A scoping review. Thromb Res. 2020;192: 152–160. doi: 10.1016/j.thromres.2020.05.039 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987;40: 373–383. doi: 10.1016/0021-9681(87)90171-8 [DOI] [PubMed] [Google Scholar]
9.Harrell FE, Lee KL, Califf RM, Pryor DB, Rosati RA. Regression modelling strategies for improved prognostic prediction. Stat Med. 1984;3: 143–152. doi: 10.1002/sim.4780030207 [DOI] [PubMed] [Google Scholar]
10.Wolbers M, Koller MT, Witteman JCM, Steyerberg EW. Prognostic models with competing risks methods and application to coronary risk prediction. Epidemiology. 2009;20: 555–561. doi: 10.1097/EDE.0b013e3181a39056 [DOI] [PubMed] [Google Scholar]
11.Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996/12/01. 1996;49: 1373–1379. doi: 10.1016/s0895-4356(96)00236-3 [DOI] [PubMed] [Google Scholar]
12.Alicandro G, Remuzzi G, La Vecchia C. Italy’s first wave of the COVID-19 pandemic has ended: no excess mortality in May, 2020. The Lancet. Lancet Publishing Group; 2020. pp. e27–e28. doi: 10.1016/S0140-6736(20)31865-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Weitz JI, Fredenburgh JC, Eikelboom JW. A Test in Context: D-Dimer. Journal of the American College of Cardiology. Elsevier USA; 2017. pp. 2411–2420. doi: 10.1016/j.jacc.2017.09.024 [DOI] [PubMed] [Google Scholar]
14.Panigada M, Bottino N, Tagliabue P, Grasselli G, Novembrino C, Chantarangkul V, et al. Hypercoagulability of COVID-19 patients in intensive care unit: A report of thromboelastography findings and other parameters of hemostasis. J Thromb Haemost. 2020;18: 1738–1742. doi: 10.1111/jth.14850 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Peyvandi F, Artoni A, Novembrino C, Aliberti S, Panigada M, Boscarino M, et al. Hemostatic alterations in COVID-19. Haematologica. 2020;105: haematol.2020.262634. doi: 10.3324/haematol.2020.262634 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Iba T, Levy JH, Levi M, Thachil J. Coagulopathy in COVID-19. Journal of Thrombosis and Haemostasis. Blackwell Publishing Ltd; 2020. pp. 2103–2109. doi: 10.1111/jth.14975 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Simmons J, Pittet JF. The coagulopathy of acute sepsis. Curr Opin Anaesthesiol. 2015;28: 227–236. doi: 10.1097/ACO.0000000000000163 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Cugno M, Meroni PL, Gualtierotti R, Griffini S, Grovetti E, Torri A, et al. Complement activation and endothelial perturbation parallel COVID-19 severity and activity. J Autoimmun. 2021;116: 102560. doi: 10.1016/j.jaut.2020.102560 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Cugno M, Meroni PL, Gualtierotti R, Griffini S, Grovetti E, Torri A, et al. Complement activation in patients with COVID-19: A novel therapeutic target. J Allergy Clin Immunol. 2020;146: 215–217. doi: 10.1016/j.jaci.2020.05.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Sakka M, Connors JM, Hékimian G, Martin-Toutain I, Crichi B, Colmegna I, et al. Association between D-Dimer levels and mortality in patients with coronavirus disease 2019 (COVID-19): a systematic review and pooled analysis. JMV-Journal de Medecine Vasculaire. Elsevier Masson SAS; 2020. pp. 268–274. doi: 10.1016/j.jdmv.2020.05.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Zhan H, Chen H, Liu C, Cheng L, Yan S, Li H, et al. Diagnostic Value of D-Dimer in COVID-19: A Meta-Analysis and Meta-Regression. Clin Appl Thromb. 2021;27. doi: 10.1177/10760296211010976 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Zhang L, Yan X, Fan Q, Liu H, Liu X, Liu Z, et al. D-dimer levels on admission to predict in-hospital mortality in patients with Covid-19. J Thromb Haemost. 2020;18: 1324–1329. doi: 10.1111/jth.14859 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Naymagon L, Zubizarreta N, Feld J, van Gerwen M, Alsen M, Thibaud S, et al. Admission D-dimer levels, D-dimer trends, and outcomes in COVID-19. Thromb Res. 2020;196: 99–105. doi: 10.1016/j.thromres.2020.08.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Oulhaj A, Ahmed LA, Prattes J, Suliman A, Alsuwaidi AR, Al-Rifai RH, et al. The competing risk between in-hospital mortality and recovery: A pitfall in COVID-19 survival analysis research. medRxiv. 2020; 2020.07.11.20151472. doi: 10.1101/2020.07.11.20151472 [DOI] [Google Scholar]
25.Van Calster B, McLernon DJ, Van Smeden M, Wynants L, Steyerberg EW, Bossuyt P, et al. Calibration: The Achilles heel of predictive analytics. BMC Med. 2019;17. doi: 10.1186/s12916-019-1466-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Thachil J, Longstaff C, Favaloro EJ, Lippi G, Urano T, Kim PY. The need for accurate D‐dimer reporting in COVID‐19: Communication from the ISTH SSC on fibrinolysis. J Thromb Haemost. 2020;18: 2408–2411. doi: 10.1111/jth.14956 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Linkins LA, Takach Lapner S. Review of D-dimer testing: Good, Bad, and Ugly. International Journal of Laboratory Hematology. Blackwell Publishing Ltd; 2017. pp. 98–103. doi: 10.1111/ijlh.12665 [DOI] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0264106.r001

Decision Letter 0

Massimo Filippi

8 Apr 2022

PONE-D-22-03309The usefulness of D-dimer as a predictive marker for mortality in patients with COVID-19 hospitalized during the first wave in Italy.PLOS ONE

Dear Dr. Hassan,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by May 19 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Massimo Filippi

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for including your consent statement: "Written informed consent was obtained from patients before data collection. In cases where it was not possible to obtain informed consent, due to severe illness or death. Data collection was still performed assuming the patient’s consent."

Please state in the ethics statement in the Methods and online submission form whether the ethics committee that approved your study waived the requirement for written informed consent from all patients.

3. Thank you for stating the following in the Competing Interests section:

(B. Ferrari has received consulting fees and travel support from Sanofi Genzyme. R. Gualtierotti reports participation in advisory boards for Biomarin, Pfizer, Bayer and Takeda as well as participation at educational seminars sponsored by Pfizer, Sobi and Roche. I. Martinelli reports personal and non-financial support from Bayer, Roche, Rovi and Novo Nordisk outside of the submitted work. A. Gori has received grants for research support, honoraria, consultation fees, and travel support from Gilead, Janssen, MSD, Pfizer, Angelini, Menarini, ViiV. F. Peyvandi has received honoraria for participating as a speaker at educational meetings, symposia and advisory boards of Roche, Sobi, Sanofi, Grifols and Takeda. All other authors have no conflicts of interest to disclose.)

We note that you received funding from a commercial source: (Sanofi Genzyme, Biomarin, Pfizer, Bayer, Takeda, Sobi, Roche, Bayer, Rovi, Novo Nordisk, Gilead, Janssen, MSD, Pfizer, Angelini, Menarini, ViiV, and Grifols)

Please provide an amended Competing Interests Statement that explicitly states this commercial funder, along with any other relevant declarations relating to employment, consultancy, patents, products in development, marketed products, etc.

Within this Competing Interests Statement, please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your amended Competing Interests Statement within your cover letter. We will change the online submission form on your behalf.

4. One of the noted authors is a group or consortium (the COVID-19 Network working group). In addition to naming the author group, please list the individual authors and affiliations within this group in the acknowledgments section of your manuscript. Please also indicate clearly a lead author for this group along with a contact email address."

the COVID-19 Network working group

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The manuscript entitled “The usefulness of D-dimer as a predictive marker for mortality in patients with COVID-19 hospitalized during the first wave in Italy” by Shermarke Hassan et al., aims to demonstrate the value of D-dimer as a predictive marker for in hospital mortality in COVID-19 patients.

The manuscript addresses a simple and important question using a multicenter study involving a large sample size. The manuscript is well written.

There are some points that need to be addressed:

1. The statistical analyses need to be clearer.

Three prediction models were tested to show the causal relationship between D-dimer levels and in hospital mortality: a model containing only D-dimer as a predictor, a model containing

age, sex and comorbidities, and a model containing D-dimer, age, sex and comorbidities.

How was the calibration performed for each of the models using each of the quartiles?

The authors show the calibration plots in figure 2, but is not clear for these calibration plots which quartile of D-dimer are referring to?

Would be worth showing the calibration plot for each of their models developed using each of the quartiles.

Reviewer #2: The topic of the paper is timely and of clinical interest. Yet, I found the article to contain some incorrect statements and linguistic and grammatical errors. Thus, the article requires a complete revision. Specific comments:

-Page 2, section “Results”: add percentage of number of patients with evaluable data

-Page 3, section “Introduction”: numbers on people affected by COVID-19 and who died from the infection should be updated to April 2022

-Page 3, section “Introduction”, sentence “it was important to understand the role of early predictive markers”: please specify. Predictive of mortality? Disease worsening? Disability? Long-term consequences?

-Page 3, section “Introduction”: please rephrase the sentence “thereby decreasing the burden on the healthcare system” and avoid brackets

-Page 3, section “Introduction”, sentence “The secondary aim of this paper”: please remove the words “of this paper”, as they are a repetition

-Page 3, section “Methods-Study design and population”: please specify the starting time point of follow-up period: follow-up was of 30 days starting from hospital admission? Viral infection confirmed by PCR result?

-Page 3, section “Methods-Study design and population”: please rephrase the sentence “In cases where it was not possible to obtain informed consent […] assuming the patient’s consent.” Grammar is not correct; please avoid the full stop between the words “death” and “Data collection”

-Page 4, section “Methods-Data collection and definition of variables”: please remove the sentence “The primary endpoint […] mortality”, as endpoints have already been previously described

-Page 4, section “Methods-Data collection and definition of variables”: please define the Charlson comorbidity index, add reference

-Page 4, section “Methods-Data collection and definition of variables”, sentence “The total list of comorbidities was as follows;”: please change semicolons with colons

-Page 4, section “Methods-Data collection and definition of variables”, words “HIV aids”: please specify HIV infection and AIDS

-Page 4, section “Statistical analysis, general approach”: please define abbreviations before using them (e.g. SD, IQR…)

- Page 4, section “Statistical analysis, general approach”, sentence “in that it (practically) precludes”: please rephrase avoiding the use of brackets

- Page 4, section “Statistical analysis, general approach”, sentence “patients with missing values for the exposure”: please remove “the”

-Page 5, section “Statistical analysis, the predictive value of D-dimer for in-hospital mortality”: please replace semicolons with colons

-Page 5, section “Statistical analysis, the predictive value of D-dimer for in-hospital mortality”, paragraph “Model calibration was measured […] line to the data”: the whole paragraph is not clear and it is hard to follow. Could you please explain it better?

-Page 5, section “Results – Baseline characteristics”, sentence “either anticoagulant […] and hospital admission”: please rephrase this sentence, as English grammar is not correct

- Page 5, section “Results – Baseline characteristics”, sentence “between March 6th 2020 to September 6th 2020”: please replace TO with AND

-Page 5, section “Results – Baseline characteristics”, sentence “Almost all […] COVID-19 wave”: please rephrase avoiding the use of brackets

- Page 6, section “Results – Baseline characteristics”, sentence “Who died doing so”: please rephrase

- Page 6, section “Results – Baseline characteristics”: could you please specify how quartiles are defined and which are D-Dimer levels?

-Page 7, section “Discussion”, sentence “The mechanisms underlying this COVID-19 induced coagulopathy”: please remove “the” and add a “-“ between COVID-19 and induced (COVID-19-induced coagulopathy”;

- Page 7, section “Discussion”, sentence “The mechanisms underlying this COVID-19 induced coagulopathy may […] by monocytes”: add a reference explaining mechanisms of other bacteria-induced septic coagulopathies

-Page 7, section “Discussion”, sentence “measured within 3 days of admission”: please substitute “of” with “from”

-Page 8, section “Discussion”, sentence “9.1% of patients were on anticoagulant use […] during the study”: please rephrase avoiding the repetition of “anticoagulant use”

-Page 8, section “Discussion”: please define abbreviations: VTE, DIC, TMA…

-Page 8, section “Discussion”, sentence “For example in Lombardy […] were already being prescribed anticoagulant treatment”: please rephrase

-As suggested also by Authors, an important limitation of this study is the absence of D-Dimer levels in more than 500 patients included in the study, which represent approximately half of the analyzed cohort. Moreover, D-dimer is a non-specific marker of inflammation, therefore it is likely to be high in COVID-19 infected patients, regardless of the presence of coagulopathy and/or infection-related thrombotic event.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2022 Jul 22;17(7):e0264106. doi: 10.1371/journal.pone.0264106.r002

Author response to Decision Letter 0

7 Jun 2022

Editor

Request: 1. When submitting your revision, we need you to address these additional requirements. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Reply: I have applied the PLOS ONE style requirements

Request: 2. Thank you for including your consent statement: "Written informed consent was obtained from patients before data collection. In cases where it was not possible to obtain informed consent, due to severe illness or death. Data collection was still performed assuming the patient’s consent." Please state in the ethics statement in the Methods and online submission form whether the ethics committee that approved your study waived the requirement for written informed consent from all patients.

Reply: I have amended the consent statement as follows:

“This observational study was approved by the Medical Ethics Committee of the Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico. The need to obtain informed consent was waived by the Medical Ethics Committee in cases where it was not possible to obtain informed consent, due to severe illness or death. In all other cases, written informed consent was obtained.”

I have also changed this in the online submission form.

Request: 3. Thank you for stating the following in the Competing Interests section: (B. Ferrari has received consulting fees and travel support from Sanofi Genzyme. R. Gualtierotti reports participation in advisory boards for Biomarin, Pfizer, Bayer and Takeda as well as participation at educational seminars sponsored by Pfizer, Sobi and Roche. I. Martinelli reports personal and non-financial support from Bayer, Roche, Rovi and Novo Nordisk outside of the submitted work. A. Gori has received grants for research support, honoraria, consultation fees, and travel support from Gilead, Janssen, MSD, Pfizer, Angelini, Menarini, ViiV. F. Peyvandi has received honoraria for participating as a speaker at educational meetings, symposia and advisory boards of Roche, Sobi, Sanofi, Grifols and Takeda. All other authors have no conflicts of interest to disclose.)

Reply: I am not sure if I understood this request correctly. What is meant by “Please provide an amended Competing Interests Statement that explicitly states this commercial funder”? Isn’t this exactly what I have already done? (i.e. mentioning these commercial parties in the competing interest statement) Also, what is meant by “We note that you received funding”? Do you mean funding for this project? None of the commercial parties mentioned in the competing interest statement have funded the current work. Could you please explain to me what further action needs to be taken? For now, I have added "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” to the Competing Interest statement:

“B. Ferrari has received consulting fees and travel support from Sanofi Genzyme. R. Gualtierotti reports participation in advisory boards for Biomarin, Pfizer, Bayer and Takeda as well as participation at educational seminars sponsored by Pfizer, Sobi and Roche. I. Martinelli reports personal and non-financial support from Bayer, Roche, Rovi and Novo Nordisk outside of the submitted work. A. Gori has received grants for research support, honoraria, consultation fees, and travel support from Gilead, Janssen, MSD, Pfizer, Angelini, Menarini, ViiV. F. Peyvandi has received honoraria for participating as a speaker at educational meetings, symposia and advisory boards of Roche, Sobi, Sanofi, Grifols and Takeda. All other authors have no conflicts of interest to disclose. This does not alter our adherence to PLOS ONE policies on sharing data and materials.”

Request: 4. One of the noted authors is a group or consortium (the COVID-19 Network working group). In addition to naming the author group, please list the individual authors and affiliations within this group in the acknowledgments section of your manuscript. Please also indicate clearly a lead author for this group along with a contact email address."

Reply: I have added the information of the lead author of this group to the acknowledgements section. I have also added the full list of authors and affiliations, but because it is very long, I have added it as a supplement instead of putting it in the acknowledgements.

Reviewer 1

The manuscript entitled “The usefulness of D-dimer as a predictive marker for mortality in patients with COVID-19 hospitalized during the first wave in Italy” by Shermarke Hassan et al., aims to demonstrate the value of D-dimer as a predictive marker for in hospital mortality in COVID-19 patients. The manuscript addresses a simple and important question using a multicenter study involving a large sample size. The manuscript is well written.

There are some points that need to be addressed:

1. The statistical analyses need to be clearer.

Three prediction models were tested to show the causal relationship between D-dimer levels and in hospital mortality: a model containing only D-dimer as a predictor, a model containing

age, sex and comorbidities, and a model containing D-dimer, age, sex and comorbidities.

Question: How was the calibration performed for each of the models using each of the quartiles? The authors show the calibration plots in figure 2, but is not clear for these calibration plots which quartile of D-dimer are referring to? Would be worth showing the calibration plot for each of their models developed using each of the quartiles.

Reply: Thank you for the insightful questions, we did not clearly mention how D-dimer was modeled in these analyses, which may have caused some confusion in understanding the method of analysis. When assessing the predictive value of D-dimer for in-hospital mortality (by looking at model discrimination and model calibration), we modeled D-dimer as a continuous variable, because this yields the best predictive value. (as categorizing a variable always leads to some reduction in predictive power) So figure 2 (the calibration plot) consists of prediction models that use D-dimer as a continuous variable. When assessing the association between D-dimer and in-hospital mortality we wanted to model the variable in a way that was more understandable from a clinicians’ point of view. A statement like “the relative risk of in-hospital mortality increases by X% for every 1 ng/mL increase in D-dimer levels” is a little hard to understand and doesn’t give you any sense of what this means in clinical practice. Therefore, we decided to model D-dimer using quartiles of D-dimer levels for this analysis. We have modified the manuscript so that this is clearer:

“All information was obtained from electronic patient records, using a standardized case report form. The exposure of interest, D-dimer levels (expressed as ng/mL) was used as either a continuous variable or a categorical variable depending on the analysis of interest. When showing descriptive statistics and estimating the association between D-dimer levels and in-hospital mortality, D-dimer was converted to a categorical variable with 4 levels that correspond to the 1st, 2nd, 3rd and 4th quartile of D-dimer levels. This was done to make the results easier to interpret for the reader. D-dimer was analyzed as a continuous variable when assessing the predictive value of D-dimer levels at admission for in-hospital mortality. This is because categorizing a variable always leads to some loss in predictive power.” (page 4)

We also rewrote the part of the manuscript where we explain model calibration:

“Model calibration refers to the degree to which the risk of mortality predicted by a model and the actual observed mortality rate in a group of patients are similar. For example, if the predicted mortality risk for a group of 100 patients is equal to 12% than the observed proportion of patients that died should also be around 12%. If the predicted mortality risk is much higher or lower than the observed mortality rate than the model is miscalibrated. To assess model calibration, we first fitted a model to the data, and then calculated the individual predicted mortality risk for each patient using this model. We then divided the patient population into ten groups (deciles) based on their predicted mortality risk. For each group, the predicted mortality risk for the whole group was compared with the observed mortality rate in that same group. This was done visually in a scatterplot that showed the predicted mortality risk on the X-axis and the observed mortality on the Y-axis for each decile. Furthermore, to examine calibration across the whole range, we also fitted a LOWESS (Locally Weighted Scatterplot Smoothing) line to the data.” (page 5/6)

Reviewer 2

The topic of the paper is timely and of clinical interest. Yet, I found the article to contain some incorrect statements and linguistic and grammatical errors. Thus, the article requires a complete revision. Specific comments:

Reply: Thank you for the important feedback, I have addressed your comments point-by-point below:

Question: Page 2, section “Results”: add percentage of number of patients with evaluable data

Reply: Done

Question: Page 3, section “Introduction”: numbers on people affected by COVID-19 and who died from the infection should be updated to April 2022

Reply: Done

Question: Page 3, section “Introduction”, sentence “it was important to understand the role of early predictive markers”: please specify. Predictive of mortality? Disease worsening? Disability? Long-term consequences?

Reply: Changed to “it was important to look for biomarkers measured at admission that could predict mortality and other in-hospital adverse outcomes, in order to better triage patients.”

Question: Page 3, section “Introduction”: please rephrase the sentence “thereby decreasing the burden on the healthcare system” and avoid brackets

Reply: Changed to “Patients predicted to have a low enough mortality risk wouldn’t need to be hospitalized, thereby decreasing the burden on the healthcare system.”

Question: Page 3, section “Introduction”, sentence “The secondary aim of this paper”: please remove the words “of this paper”, as they are a repetition

Reply: Done

Question: Page 3, section “Methods-Study design and population”: please specify the starting time point of follow-up period: follow-up was of 30 days starting from hospital admission? Viral infection confirmed by PCR result?

Reply: This was already mentioned in page 3 in the following way “The study population consisted of patients aged > 18 years who were hospitalized and who were positive for COVID-19 based on real-time PCR at 9 Italian hospitals, during the first COVID-19 wave in Lombardy, Italy (Feb-May 2020).

Question: Page 3, section “Methods-Study design and population”: please rephrase the sentence “In cases where it was not possible to obtain informed consent […] assuming the patient’s consent.” Grammar is not correct; please avoid the full stop between the words “death” and “Data collection”

Reply: Done

Question: Page 4, section “Methods-Data collection and definition of variables”: please remove the sentence “The primary endpoint […] mortality”, as endpoints have already been previously described

Reply: I have left this sentence in and modified it to “The outcome used for all analyses was in-hospital mortality” as it is the first time the outcome is mentioned in the methods section.

Question: Page 4, section “Methods-Data collection and definition of variables”: please define the Charlson comorbidity index, add reference

Reply: Added reference and changed text to “Lastly, the number of comorbidities (continuous variable) that each patient had at admission was calculated. Comorbidities to be assessed were selected based on their usage in the Charlson comorbidity index [8], a well-known risk score used to predict 10-year survival in patients with several comorbidities. To this list of comorbidities, we also added obesity (as defined by the clinician). The final list of comorbidities used was as follows: cardiovascular disease, chronic obstructive pulmonary disease, chronic kidney disease, diabetes mellitus, cancer, liver disease, dementia, connective tissue disease, acquired immunodeficiency syndrome (AIDS) and clinician-defined obesity.”

Question: Page 4, section “Methods-Data collection and definition of variables”, sentence “The total list of comorbidities was as follows;”: please change semicolons with colons

Reply: Done

Question: Page 4, section “Methods-Data collection and definition of variables”, words “HIV aids”: please specify HIV infection and AIDS

Reply: Changed to “acquired immunodeficiency syndrome (AIDS)”

Question: Page 4, section “Statistical analysis, general approach”: please define abbreviations before using them (e.g. SD, IQR…)

Reply: Done

Question: Page 4, section “Statistical analysis, general approach”, sentence “in that it (practically) precludes”: please rephrase avoiding the use of brackets

Reply: Removed brackets, changed “practically” to “almost completely”

Question: Page 4, section “Statistical analysis, general approach”, sentence “patients with missing values for the exposure”: please remove “the”

Reply: Done

Question: Page 5, section “Statistical analysis, the predictive value of D-dimer for in-hospital mortality”: please replace semicolons with colons

Reply: Done

Question: Page 5, section “Statistical analysis, the predictive value of D-dimer for in-hospital mortality”, paragraph “Model calibration was measured […] line to the data”: the whole paragraph is not clear and it is hard to follow. Could you please explain it better?

Reply: I rewrote the paragraph as follows: “Model calibration refers to the degree to which the risk of mortality predicted by a model and the actual observed mortality rate in a group of patients are similar. For example, if the predicted mortality risk for a group of 100 patients is equal to 12% than the observed proportion of patients that died should also be around 12%. If the predicted mortality risk is much higher or lower than the observed mortality rate than the model is miscalibrated. To assess model calibration, we first fitted a model to the data, and then calculated the individual predicted mortality risk for each patient using this model. We then divided the patient population into ten groups (deciles) based on their predicted mortality risk. For each group, the predicted mortality risk for the whole group was compared with the observed mortality rate in that same group. This was done visually in a scatterplot that showed the predicted mortality risk on the X-axis and the observed mortality on the Y-axis for each decile. Furthermore, to examine calibration across the whole range, we also fitted a LOWESS (Locally Weighted Scatterplot Smoothing) line to the data.”

Question: Page 5, section “Results – Baseline characteristics”, sentence “either anticoagulant […] and hospital admission”: please rephrase this sentence, as English grammar is not correct

Reply: I changed the sentence to “47 patients were excluded due to missing data for one or more confounding factors”

Question: Page 5, section “Results – Baseline characteristics”, sentence “between March 6th 2020 to September 6th 2020”: please replace TO with AND

Reply: Done

Question: Page 5, section “Results – Baseline characteristics”, sentence “Almost all […] COVID-19 wave”: please rephrase avoiding the use of brackets

Reply: Changed the sentence to “Almost all (98%) patients were enrolled before May 31st 2020, which roughly corresponds to the first three months of the initial COVID-19 epidemic in Italy”

Question: Page 6, section “Results – Baseline characteristics”, sentence “Who died doing so”: please rephrase

Reply: Changed to “75% of deaths occurred in the first 10 days”

Question: Page 6, section “Results – Baseline characteristics”: could you please specify how quartiles are defined and which are D-Dimer levels?

Reply: See the following text on page 4: “All information was obtained from electronic patient records, using a standardized case report form. The exposure of interest, D-dimer levels (expressed as ng/mL) was used as either a continuous variable or a categorical variable depending on the analysis of interest. When showing descriptive statistics and estimating the association between D-dimer levels and in-hospital mortality, D-dimer was converted to a categorical variable with 4 levels that correspond to the 1st, 2nd, 3rd and 4th quartile of D-dimer levels. This was done to make the results easier to interpret for the reader. D-dimer was analyzed as a continuous variable when assessing the predictive value of D-dimer levels at admission for in-hospital mortality. This is because categorizing a variable always leads to some loss in predictive power.”

Question: Page 7, section “Discussion”, sentence “The mechanisms underlying this COVID-19 induced coagulopathy”: please remove “the” and add a “-“ between COVID-19 and induced (COVID-19-induced coagulopathy”;

Reply: Done

Question: Page 7, section “Discussion”, sentence “The mechanisms underlying this COVID-19 induced coagulopathy may […] by monocytes”: add a reference explaining mechanisms of other bacteria-induced septic coagulopathies

Reply: Added reference

Question: Page 7, section “Discussion”, sentence “measured within 3 days of admission”: please substitute “of” with “from”

Reply: Done

Question: Page 8, section “Discussion”, sentence “9.1% of patients were on anticoagulant use […] during the study”: please rephrase avoiding the repetition of “anticoagulant use”

Reply: Changed to “). At baseline, 9.1% of patients were on anticoagulants and no information was given about thromboprophylaxis use during the study.”

Question: Page 8, section “Discussion”: please define abbreviations: VTE, DIC, TMA…

Reply: I have written them all out and removed the abbreviations as they are only mentioned 1-2 times.

Question: Page 8, section “Discussion”, sentence “For example in Lombardy […] were already being prescribed anticoagulant treatment”: please rephrase

Reply: Changed sentence to “For example, prophylactic anticoagulation for hospitalized COVID-19 patients became much more common over time.”

Question: As suggested also by Authors, an important limitation of this study is the absence of D-Dimer levels in more than 500 patients included in the study, which represent approximately half of the analyzed cohort. Moreover, D-dimer is a non-specific marker of inflammation, therefore it is likely to be high in COVID-19 infected patients, regardless of the presence of coagulopathy and/or infection-related thrombotic event.

Reply: You are correct, D-dimer is indeed also a marker of inflammation and therefore not specific to coagulopathies. For our purposes this was actually a good thing as we were interested in using D-dimer to predict overall mortality, not just thrombotic events. So the predictive value of D-dimer in this paper was based on its ability to act as a proxy measurement for 1) problems in coagulation as well as 2) general inflammation, which are both factors that lower the survival rate.

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file.^{(23.5KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0264106.r003

Decision Letter 1

Massimo Filippi

29 Jun 2022

The usefulness of D-dimer as a predictive marker for mortality in patients with COVID-19 hospitalized during the first wave in Italy.

PONE-D-22-03309R1

Dear Dr. Hassan,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Massimo Filippi

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

PLoS One. doi: 10.1371/journal.pone.0264106.r004

Acceptance letter

Massimo Filippi

13 Jul 2022

PONE-D-22-03309R1

The usefulness of D-dimer as a predictive marker for mortality in patients with COVID-19 hospitalized during the first wave in Italy.

Dear Dr. Hassan:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Massimo Filippi

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Cumulative incidence of death, per time-period.

(DOCX)

Click here for additional data file.^{(12.1KB, docx)}

S1 List. List of participants in COVID-19 Network working group.

(DOCX)

Click here for additional data file.^{(14.8KB, docx)}

S1 File. File containing the data and code used to generate the results.

(ZIP)

Click here for additional data file.^{(20.8KB, zip)}

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file.^{(23.5KB, docx)}

Data Availability Statement

All relevant data are within the paper and its Supporting Information files.

[pone.0264106.ref001] 1.Arabi YM, Murthy S, Webb S. COVID-19: a novel coronavirus and a novel challenge for critical care. Intensive Care Med. 2020/03/04. 2020;46: 833–836. doi: 10.1007/s00134-020-05955-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref002] 2.Grasselli G, Pesenti A, Cecconi M. Critical Care Utilization for the COVID-19 Outbreak in Lombardy, Italy: Early Experience and Forecast During an Emergency Response. Jama. 2020/03/14. 2020. doi: 10.1001/jama.2020.4031 [DOI] [PubMed] [Google Scholar]

[pone.0264106.ref003] 3.Xie J, Tong Z, Guan X, Du B, Qiu H, Slutsky AS. Critical care crisis and some recommendations during the COVID-19 epidemic in China. Intensive Care Med. 2020/03/04. 2020;46: 837–840. doi: 10.1007/s00134-020-05979-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref004] 4.Dong E, Du H, Gardner L. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect Dis. 2020/02/23. 2020;20: 533–534. doi: 10.1016/S1473-3099(20)30120-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref005] 5.Wu Z, McGoogan JM. Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for Disease Control and Prevention. Jama. 2020/02/25. 2020. doi: 10.1001/jama.2020.2648 [DOI] [PubMed] [Google Scholar]

[pone.0264106.ref006] 6.Tang N, Li D, Wang X, Sun Z. Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia. J Thromb Haemost. 2020/02/20. 2020;18: 844–847. doi: 10.1111/jth.14768 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref007] 7.Al-Ani F, Chehade S, Lazo-Langner A. Thrombosis risk associated with COVID-19 infection. A scoping review. Thromb Res. 2020;192: 152–160. doi: 10.1016/j.thromres.2020.05.039 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref008] 8.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987;40: 373–383. doi: 10.1016/0021-9681(87)90171-8 [DOI] [PubMed] [Google Scholar]

[pone.0264106.ref009] 9.Harrell FE, Lee KL, Califf RM, Pryor DB, Rosati RA. Regression modelling strategies for improved prognostic prediction. Stat Med. 1984;3: 143–152. doi: 10.1002/sim.4780030207 [DOI] [PubMed] [Google Scholar]

[pone.0264106.ref010] 10.Wolbers M, Koller MT, Witteman JCM, Steyerberg EW. Prognostic models with competing risks methods and application to coronary risk prediction. Epidemiology. 2009;20: 555–561. doi: 10.1097/EDE.0b013e3181a39056 [DOI] [PubMed] [Google Scholar]

[pone.0264106.ref011] 11.Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996/12/01. 1996;49: 1373–1379. doi: 10.1016/s0895-4356(96)00236-3 [DOI] [PubMed] [Google Scholar]

[pone.0264106.ref012] 12.Alicandro G, Remuzzi G, La Vecchia C. Italy’s first wave of the COVID-19 pandemic has ended: no excess mortality in May, 2020. The Lancet. Lancet Publishing Group; 2020. pp. e27–e28. doi: 10.1016/S0140-6736(20)31865-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref013] 13.Weitz JI, Fredenburgh JC, Eikelboom JW. A Test in Context: D-Dimer. Journal of the American College of Cardiology. Elsevier USA; 2017. pp. 2411–2420. doi: 10.1016/j.jacc.2017.09.024 [DOI] [PubMed] [Google Scholar]

[pone.0264106.ref014] 14.Panigada M, Bottino N, Tagliabue P, Grasselli G, Novembrino C, Chantarangkul V, et al. Hypercoagulability of COVID-19 patients in intensive care unit: A report of thromboelastography findings and other parameters of hemostasis. J Thromb Haemost. 2020;18: 1738–1742. doi: 10.1111/jth.14850 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref015] 15.Peyvandi F, Artoni A, Novembrino C, Aliberti S, Panigada M, Boscarino M, et al. Hemostatic alterations in COVID-19. Haematologica. 2020;105: haematol.2020.262634. doi: 10.3324/haematol.2020.262634 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref016] 16.Iba T, Levy JH, Levi M, Thachil J. Coagulopathy in COVID-19. Journal of Thrombosis and Haemostasis. Blackwell Publishing Ltd; 2020. pp. 2103–2109. doi: 10.1111/jth.14975 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref017] 17.Simmons J, Pittet JF. The coagulopathy of acute sepsis. Curr Opin Anaesthesiol. 2015;28: 227–236. doi: 10.1097/ACO.0000000000000163 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref018] 18.Cugno M, Meroni PL, Gualtierotti R, Griffini S, Grovetti E, Torri A, et al. Complement activation and endothelial perturbation parallel COVID-19 severity and activity. J Autoimmun. 2021;116: 102560. doi: 10.1016/j.jaut.2020.102560 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref019] 19.Cugno M, Meroni PL, Gualtierotti R, Griffini S, Grovetti E, Torri A, et al. Complement activation in patients with COVID-19: A novel therapeutic target. J Allergy Clin Immunol. 2020;146: 215–217. doi: 10.1016/j.jaci.2020.05.006 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref020] 20.Sakka M, Connors JM, Hékimian G, Martin-Toutain I, Crichi B, Colmegna I, et al. Association between D-Dimer levels and mortality in patients with coronavirus disease 2019 (COVID-19): a systematic review and pooled analysis. JMV-Journal de Medecine Vasculaire. Elsevier Masson SAS; 2020. pp. 268–274. doi: 10.1016/j.jdmv.2020.05.003 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref021] 21.Zhan H, Chen H, Liu C, Cheng L, Yan S, Li H, et al. Diagnostic Value of D-Dimer in COVID-19: A Meta-Analysis and Meta-Regression. Clin Appl Thromb. 2021;27. doi: 10.1177/10760296211010976 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref022] 22.Zhang L, Yan X, Fan Q, Liu H, Liu X, Liu Z, et al. D-dimer levels on admission to predict in-hospital mortality in patients with Covid-19. J Thromb Haemost. 2020;18: 1324–1329. doi: 10.1111/jth.14859 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref023] 23.Naymagon L, Zubizarreta N, Feld J, van Gerwen M, Alsen M, Thibaud S, et al. Admission D-dimer levels, D-dimer trends, and outcomes in COVID-19. Thromb Res. 2020;196: 99–105. doi: 10.1016/j.thromres.2020.08.032 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref024] 24.Oulhaj A, Ahmed LA, Prattes J, Suliman A, Alsuwaidi AR, Al-Rifai RH, et al. The competing risk between in-hospital mortality and recovery: A pitfall in COVID-19 survival analysis research. medRxiv. 2020; 2020.07.11.20151472. doi: 10.1101/2020.07.11.20151472 [DOI] [Google Scholar]

[pone.0264106.ref025] 25.Van Calster B, McLernon DJ, Van Smeden M, Wynants L, Steyerberg EW, Bossuyt P, et al. Calibration: The Achilles heel of predictive analytics. BMC Med. 2019;17. doi: 10.1186/s12916-019-1466-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref026] 26.Thachil J, Longstaff C, Favaloro EJ, Lippi G, Urano T, Kim PY. The need for accurate D‐dimer reporting in COVID‐19: Communication from the ISTH SSC on fibrinolysis. J Thromb Haemost. 2020;18: 2408–2411. doi: 10.1111/jth.14956 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264106.ref027] 27.Linkins LA, Takach Lapner S. Review of D-dimer testing: Good, Bad, and Ugly. International Journal of Laboratory Hematology. Blackwell Publishing Ltd; 2017. pp. 98–103. doi: 10.1111/ijlh.12665 [DOI] [PubMed] [Google Scholar]

PERMALINK

The usefulness of D-dimer as a predictive marker for mortality in patients with COVID-19 hospitalized during the first wave in Italy

Shermarke Hassan

Barbara Ferrari

Raffaella Rossio

Vincenzo la Mura

Andrea Artoni

Roberta Gualtierotti

Ida Martinelli

Alessandro Nobili

Alessandra Bandera

Andrea Gori

Francesco Blasi

Valter Monzani

Giorgio Costantino

Sergio Harari

Frits Richard Rosendaal

Flora Peyvandi

Roles

Abstract

Background

Aims

Methods

Results

Conclusion

Introduction

Methods

Study design and population

Data collection and definition of variables

Statistical analysis: General approach

Statistical analysis: Causal relationship between D-dimer levels and in-hospital mortality

Statistical analysis: The predictive value of D-dimer for in-hospital mortality

Sample size calculation

Results

Baseline characteristics

Table 1. Baseline characteristics of COVID-19 patients hospitalized in the region of Lombardy, Italy, during the first COVID-19 wave (Feb-May 2020).

Fig 1. Cumulative incidence function of 501 COVID-19 patients hospitalized in the region of Lombardy, Italy, during the first COVID-19 wave (Feb-May 2020).

Table 2. Association between D-dimer values and in-hospital mortality.

Causal relationship between D-dimer levels and in-hospital mortality

The predictive value of D-dimer for in-hospital mortality

Fig 2. Calibration plot of prediction models.

Discussion

Conclusions

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Massimo Filippi

Roles

Author response to Decision Letter 0

Decision Letter 1

Massimo Filippi

Roles

Acceptance letter

Massimo Filippi

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases