Association of platelet count with 28-day mortality in medical-surgical ICU patients with sepsis: a multicenter retrospective cohort study

doi:10.21203/rs.3.rs-4689196/v1

Download PDF

Research Article

Association of platelet count with 28-day mortality in medical-surgical ICU patients with sepsis: a multicenter retrospective cohort study

https://doi.org/10.21203/rs.3.rs-4689196/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

The association between platelet count and 28-day mortality in medical-surgical intensive care unit (ICU) patients with sepsis remains inconclusive. The aim of this study was to investigate whether platelet count is associated with 28-day mortality in these patients.

Methods

This retrospective cohort study extracted 6,122 adult patients with sepsis in medical-surgical ICU from the eICU Collaborative Research Database (eICU-CRD). The logistic regression models were used to estimate the covariates and investigate the relatioshiop between platelet count and 28-day mortality rate. Then, a generalized additive model (GAM) was used to investigate the dose-response relationship between the platelet count (every 10-unit change in platelet count) and 28-day mortality rate. Moreover, a two-piece-wise linear regression model was applied to assess the threshold effect of the platelet count and 28-day mortality rate.

Results

After adjustment for the covariates, the platelet count had a nonlinear relationship with 28-day mortality (P < 0.001). On the left side of the inflection point (platelet count < 127 x10ˆ9/L), an increase of 10 in the platelet count was associated with a 10% decreased risk 0f 28-day mortality rate (OR = 0.90, 95% CI = 0.87–0.93, P < 0.001). Nevertheless, when the platelet count ≥ 127 x10ˆ9/L, every 10-unit increase in platelet count was not significantly associated with 28-day mortality rate.

Conclusion

The relationship between platelet count and 28-day mortality rate in medical-surgical ICU patients with sepsis was nonlinear. This indicates that low platelet count may receive attention in medical-surgical ICU patients with sepsis.

platelet count

mortality

eICU

Since intensive care unit (ICU) patients are exposed to sepsis, pneumonia, and other inflammation^{[1, 2]}, their high mortality rates range from 11–18%^{[3, 4]}. Additionally, the patients admitted to the ICU after surgery may be associated with a progressive increase in adverse outcomes^[5]. Some survivors may suffer from psychosocial, physical, and cognitive sequelae^[6]. Therefore, these conditions may cause a considerable burden on public health^{[7, 8]}.

Platelets, which are cytoplasmic fragments produced by megakaryocytes, participate in the process of hemostasis and maintaining the integrity of the vascular endothelium^[9]. Platelet dysfunction or quantitative abnormalities may trigger immune dysregulation, prolonged ICU stay, and mortality^[10]. It is worth noting that thrombocytopenia is frequent in ICU patients^[11]. Consequently, platelets have always been a focus of ICU physicians. In an international prospective cohort study, patients with thrombocytopenia had a higher risk of 90-day mortality. A retrospective cohort study utilizing the Electronic ICU database suggested that in critically ill patients with tumors, the association between platelet count and in-hospital mortality was curvilinear^[12]. Another retrospective cohort study using the eICU database showed that patients who develop thrombocytopenia during their stay in the neurological ICU have a higher odds ratio of in-hospital mortality^[13]. Another study using the eICU database also suggested that the mortality risk of patients with infectious diseases in the intensive care unit increased as the nadir platelet count decreased below 130 × 10⁹/L^[13]. However, in patients with medical-surgical ICU, whether platelet count is related to a 28-day mortality rate is limited.

So, we used the eICU-CRD to conduct a retrospective multicentre cohort study, which explored the relationship between platelet count and 28-day mortality rate in medical-surgical ICU patients with sepsis and further investigate whether the threshold of platelet count where the mortality risk significantly decreases, which is a high priority in patients with sepsis.

Data source

This was a retrospective observational study, with data extracted from the eICU-CRD, an international online critical care database^[15]. The eICU-CRD is a multicenter ICU database containing high-granularity data on over 200,000 admissions to the ICU monitored by the eICU program across the United States. From 2014 to 2015, all data were automatically stored through the Philips Healthcare eICU program and retrieved electronically^[15]. The eICU-CRD has been utilized for observational studies^{[16, 17]}. Access to data was approved after finishing the CITI program for "Data or Specimen Only Research". Due to the retrospective design without direct patient intervention and the security schema for which the reidentification risk was certified as meeting Safe Harbor standards by Privacert (Cambridge, MA), this study was exempted from approval from the Institutional Review Board of the Massachusetts Institute of Technology (our record ID: 13249328).

For the same reason, informed consent was waived. The study was conducted according to the Declaration of Helsinki. All methods were performed in line with the relevant guidelines and regulations.

Study population.

All patients admitted to the ICU and diagnosed with sepsis were included. Sepsis was defined as suspected or documented infection plus an acute increase of greater than 2 points^[18] in the Sequential Organ Failure Assessment (SOFA) score recorded in the Acute Physiology and Chronic Health Evaluation (APACHE) IV dataset^[19]. Infection was identified from the ICD-9 code in the eICU Collaborative Research Database. The following exclusion criteria were used: (1) not first ICU admission, (2) ICU stay < 48 hours, (3) age < 18 years old, (4) missing platelet count, (5) non-medical-surgical patients, and (5) system error of platelet count. The study flowchart is presented in Fig. 1.

Variables

The eICU database includes demographic records, physiological indicators of bedside monitors, diagnosis via ICD-9 codes, and other laboratory data obtained during routine medical care.

Baseline characteristics such as age, gender, ethnicity, and body mass index (BMI, kg/m²) were collected from the patient and ApachePatientResult tables. The laboratory indices of platelet count ('x10ˆ9/L) were collected from the laboratory tables. The physiological variables, including temperature (°C), respiratory rate, heart rate (HR), and mean arterial pressure (MAP), were obtained from the apacheApsVar table. Comorbidities, including acute immunodeficiency syndrome (AIDS), chronic obstructive pulmonary disease (COPD), diabetes, chronic heart failure, hepatic failure, Cirrhosis, leukemia, metastatic cancer, lymphoma, and immunosuppression were extracted from the APACHE IV score. Severity at admission was measured by the Glasgow Coma Scale (GCS), Apache IV score, and Acute Physiology Score III. In addition, life support interventions (e.g., the use of mechanical ventilation) were incorporated in this study.

Outcomes

The outcome of this study was that all-cause ICU mortality occurred within 28 days after admission to the ICU.

Statistical analysis

Continuous variables are presented as the means ± standard deviation (SD) or median and interquartile ranges (IQR). Categorical data are shown as numbers and percentages. The difference according to the tertiles of the platelet count was compared using one-way analysis of variance (ANOVA) for continuous data and chi-squared tests for categorical variables. To improve the statistical strength of the results, we transformed the platelet count (per change in the platelet count of 10) as the targeted independent variable in the regression, the smooth and threshold effect analyses.

The statistical analysis included the following main steps. First, platelet count was classified into three groups (tertiles) according to distribution. Second, logistic regression models were applied to assess the relationship between the platelet count and the 28-day mortality rate. The results are presented as odds ratios (ORs) with 95% confidence intervals (95% CIs). The regression model included a non-adjusted model, Model I, and Model II. These confounders were selected based on their association with the outcomes of interest or changes in effect estimates of more than 10%^[20]. After considering the clinical significance, we adjusted for the following covariates: gender, age, ethnicity, BMI, site of infection, Apache IV score, AIDS, COPD, diabetes, chronic heart failure, mechanical ventilation use, hepatic failure, Cirrhosis, leukemia, metastatic cancer, lymphoma, and immunosuppression. Third, a GAM was used to investigate the dose-response relationship between the platelet count (per change in the platelet count of 10) and 28-day mortality rate (Fig. 2). Finally, we utilized a two-piece-wise linear regression model to examine the threshold effect of the platelet count (per change in the platelet count of 10) and 28-day mortality rate(Table 3). The turning point for the platelet count (per change in the platelet count of 10) was determined using "exploratory" analyses to move the trial turning point along the pre-defined interval and pick up the one that gave maximum model likelihood. In addition, we also performed a log-likelihood ratio test and compared the one-line linear regression model with the two-piece-wise linear model. To calculate the 95% CI for the turning point^[21], we used the bootstrap resampling method described in the previous analysis^{[22, 23]}. Dummy variables were used to indicate missing covariate values, which was performed when continuous variables were missing more than 1% of the value.

Table 1

Baseline characteristics of participants according to the tertiles of the platelet count (N = 6122)
Characteristic	Platelet Count (×10⁹/L) T1 (3-144) T2 (145–226) T3(227–594) N = 2036 N = 2028 N = 2058			P-value
Demographics Age (years, mean ± SD)	65.2 ± 15.8	66.9 ± 16.1	65.5 ± 15.6	0.002
Gender, n (%) Male Female Missing Ethnicity, n (%) Caucasian African American Hispanic Asian Native American Other/Unknown Missing BMI (kg/m2, mean ± SD)	911 (44.7%) 1125 (55.3%) 0(0.0%) 1545 (75.9%) 184 (9.0%) 127 (6.2%) 101 (5.0%) 27 (1.3%) 27 (1.3%) 25 (1.2%) 28.4 ± 8.5	954 (47.0%) 1074 (53.0%) 0(0.0%) 1572 (77.5%) 169 (8.3%) 122 (6.0%) 77 (3.8%) 23 (1.1%) 43 (2.1%) 22 (1.1%) 29.8 ± 9.8	1118 (54.4%) 939 (45.6%) 1(0.0%) 1596 (77.6%) 197 (9.6%) 98 (4.8%) 100 (4.9%) 16 (0.8%) 31 (1.5%) 20 (1.0%) 29.3 ± 9.6	< 0.001 0.105 < 0.001
Vital signs Temperature(°C)	36.6 ± 1.5	36.6 ± 1.3	36.6 ± 1.2	0.221
Respiratory rate (bpm)	30.3 ± 14.1	30.0 ± 14.7	30.8 ± 14.5	0.239
Heart rate (/min)	115.3 ± 29.7	113.3 ± 28.5	115.2 ± 26.7	0.043
MAP (mmHg, median, [IQR])	53(40–200)	56(40–200)	56(40–200)	< 0.001
Severity of illness (mean ± SD) GCS score	12.1 ± 3.7	12.0 ± 3.7	12.2 ± 3.5	0.184
Apache IV score	77.9 ± 27.0	72.5 ± 24.7	71.5 ± 24.8	< 0.001
Acute Physiology Score III	63.9 ± 26.3	58.9 ± 23.5	58.5 ± 23.8	< 0.001
Site of infection, n (%)				< 0.001
Sepsis, pulmonary	742 (36.4%)	928 (45.8%)	904 (43.9%)
Sepsis, renal/UTI (including bladder)	457 (22.4%)	456 (22.5%)	434 (21.1%)
Sepsis, GI	260 (12.8%)	210 (10.4%)	249 (12.1%)
Sepsis, unknown	278 (13.7%)	183 (9.0%)	190 (9.2%)
cutaneous/soft tissue	146 (7.2%)	147 (7.2%)	170 (8.3%)
other	150 (7.4%)	101 (5.0%)	104 (5.1%)
gynecologic	3 (0.1%)	3 (0.1%)	7 (0.3%)
Comorbidities AIDS, n (%) No Yes Missing COPD, n (%)	2010 (99.6%) 9 (0.4%) 17 (0.8%)	1991 (99.5%) 10 (0.5%) 27 (1.3%)	2029 (99.9%) 2 (0.1%) 27 (1.3%)	0.062 < 0.001
No	1882 (92.4%)	1805 (89.0%)	1833 (89.1%)
Yes	154 (7.6%)	223 (11.0%)	225 (10.9%)
Diabetes status, n (%)				0.002
No	1586 (77.9%)	1503 (74.1%)	1497 (72.7%)
Yes	433 (21.3%)	498 (24.6%)	534 (25.9%)
Missing	17 (0.8%)	27 (1.3%)	27 (1.3%)
Chronic heart failure, n (%)				0.679
No	1838 (90.3%)	1827 (90.1%)	1870 (90.9%)
Yes	198 (9.7%)	201 (9.9%)	188 (9.1%)
Hepatic failure, n (%)				< 0.001
No	1936 (95.1%)	1981 (97.7%)	2012 (97.8%)
Yes	83 (4.1%)	20 (1.0%)	19 (0.9%)
Missing	17 (0.8%)	27 (1.3%)	27 (1.3%)
Cirrhosis, n (%)				< 0.001
No	1892 (92.9%)	1977 (97.5%)	2014 (97.9%)
Yes	127 (6.2%)	24 (1.2%)	17 (0.8%)
Missing	17 (0.8%)	27 (1.3%)	27 (1.3%)
Leukaemia, n (%)				< 0.001
No	1967 (96.6%)	1983 (97.8%)	2018 (98.1%)
Yes	52 (2.6%)	18 (0.9%)	13 (0.6%)
Missing	17 (0.8%)	27 (1.3%)	27 (1.3%)
Metastatic cancer, n (%)				0.003
No	1939 (95.2%)	1957 (96.5%)	1982 (96.3%)
Yes	80 (3.9%)	44 (2.2%)	49 (2.4%)
Missing	17 (0.8%)	27 (1.3%)	27 (1.3%)
Mechanical ventilation use, n (%) No Yes Missing 28-day mortality, n (%) No Yes	1403 (68.9%) 616 (30.3%) 17 (0.8%) 1792 (88.0%) 244 (12.0%)	1269 (62.6%) 732 (36.1%) 27 (1.3%) 1878 (92.6%) 150 (7.4%)	1294 (62.9%) 737 (35.8%) 27 (1.3%) 1891 (91.9%) 167 (8.1%)	< 0.001 < 0.001
Abbreviations: BMI, body mass index. COPD, chronic obstructive pulmonary disease.
AIDS, Acquired Immune Deficiency Syndrome. MAP, mean arterial pressure,
Missing data were grouped into a set of dummy variables and incorporated into the analysis.
Data are expressed as the mean ± SD, median (interquartile range), or percentage. Among the 6122 patients, the amount of missing values for the covariates were 1 (0.02%) for gender, 67 (1.09%) for ethnicity, 111(1.81%) for BMI, 378 (6.17%) for temperature, 77 (1.25%) for heart rate, 91 (1.58%) for respiratory rate, 83 (1.35%) for MAP, 162 (2.64%) for GCS score, 699 (11.41%) for apache IV score, 699 (11.41%) for acute Physiology Score III, 71 (1.15%) for AIDS, 71 (1.15%) for diabetes, 71 (1.15%) for hepatic failure, 71 (1.15%) for Cirrhosis, 71 (1.15%) for leukaemia, 71 (1.15%) for metastatic cancer, and 71 (1.15%) for mechanical ventilation use.

Table 2

Association of platelet count (per change in the platelet count of 10) with 28‑day mortality rate
Variable	Non-adjusted model¹	Model Ⅰ²	Model Ⅱ³
Variable	OR (95% CI)	OR (95% CI)	OR (95% CI)
platelet count	0.98 (0.97,0.99)	0.98 (0.97, 0.99)	0.99 (0.98, 1.00)
platelet count T1 T2 T3 P for trend	Reference 0.59(0.47,0.73) 0.65(0.53,0.80) <0.001	Reference 0.57 (0.46, 0.71) 0.65 (0.53, 0.80) < 0.001	Reference 0.63 (0.50, 0.79) 0.73 (0.58, 0.91) 0.004
Abbreviations: CI, confidence interval; OR, odds ratio.
The independent variable was an increase in platelet count of 10 units.
¹ This model was not adjusted for any covariate.
² Model I was adjusted for gender, age, and ethnicity.
³ Model II was further adjusted for BMI, site of infection, apache IV score, AIDS, COPD, diabetes, chronic heart failure, mechanical ventilation use, hepatic failure, Cirrhosis, leukemia, metastatic cancer, lymphoma, and immunosuppression.

Table 3

Threshold effect analysis of the platelet count and 28-day mortality
Models	Per 10-unit increase
Models	OR (95%CI)	P value
One line effect	0.99 (0.98, 1.00)	0.008
Turning point (K)	12.7
platelet count < K	0.90 (0.87, 0.93)	< 0.0001
platelet count ≥ K	1.01 (1.00, 1.02)	0.219
P value for LRT test*		< 0.001
Data were presented as OR (95% CI) P value; Model I, linear analysis; Model II, nonlinear analysis. Adjusted for gender, age, ethnicity, BMI, site of infection, apache IV score, AIDS, COPD, diabetes, chronic heart failure, mechanical ventilation use, hepatic failure, Cirrhosis, leukemia, metastatic cancer, lymphoma, and immunosuppression.
CI, confidence interval; OR, odds ratio; LRT, logarithm likelihood ratio test.
* P < 0.05 indicates that Model II significantly differs from Model I.
The independent variable was an increase in per 10-unit platelet count.

To examine the robustness of the results, we conducted sensitivity analyses. The platelet count was divided into a categorical variable by tertiles and a trend test to observe whether the P value of the trend test was consistent with the P value when the platelet count was a continuous variable.

All the statistical analyses were performed using R software version 4.2.0 (http://www.r-proje ct. org).

1. Baseline characteristics of the study population

Table 1 presents the baseline characteristics of the study population by tertiles of platelet count. Age, gender, BMI, MAP, apache IV score, acute Physiology Score III, site of infection, COPD, diabetes, hepatic failure, Cirrhosis, leukemia, metastatic cancer, and mechanical ventilation use were significantly associated with platelet count (P < 0.05). In contrast, ethnicity, temperature, respiratory, heart rate, GCS sore, AIDS, and chronic heart failure were not.

2. Association of platelet counts with the 28‑day mortality rate in medical-surgical ICU patients with sepsis

Table 2 shows that an increase in platelet count (per change in the platelet count of 10) was associated with the 28‑day mortality rate (OR = 0.98, 95%CI = 0.97–0.99) according to the non-adjusted model. After we adjusted for several covariates (Table 2), an increase of 10 units in platelet count was also associated with the 28-day mortality rate (OR = 0.98, 95% CI = 0.97–0.99). To eliminate the influence of other covariates, we further adjusted for additional covariates, and platelet count was also significantly related to the 28-day mortality rate (OR = 0.99, 95% CI = 0.98–1.00 for every 10-units increase in platelet count) (Table 2). The variable of platelet count (per change in the platelet count of 10) was divided into categorical variables (tertiles) for a sensitivity analysis. We observed the same trend in conformity as that for the continuous variables (P for trend < 0.05) (Table 2).

3. Identification of nonlinear relationship

We observed a nonlinear dose-response relationship between platelet count (per change in the platelet count of 10) and 28‑day mortality rate (Fig. 2 and Table 3). The inflection point was 12.7 (per change in the platelet count of 10). On the left side of the inflection point (platelet count < 127 x10ˆ9/L), an increase of 10 units in the platelet count was significantly associated with a 10% decreased risk of mortality rate (OR = 0.90, 95% CI = 0.87–0.93, P < 0.0001). When the platelet count was ≥ 127 x10ˆ9/L, the mortality rate increased with an adjusted OR of 1.01 (95% CI = 1.00–1.02, P = 0.219) for every 10-unit increase in platelet count.

The GAM detected an L-shaped association between the platelet count and the 28-day mortality rate (Table 3). The linear regression model and a two-piece-wise linear regression model were compared, and the P value of the log-likelihood ratio test was < 0.001. The 95% CI for the turning point of the platelet count (per change in the platelet count of 10) was 12.7 (Table 3). This result indicates that the two-piece-wise linear regression model should be used to fit the model.

This retrospective cohort study found a curvilinear relationship between the platelet count and the 28-day mortality rate in medical-surgical ICU patients with sepsis. The nonlinear analysis showed that when the platelet count was less than 127 x10ˆ9/L, every 10-unit increase in platelet count was associated with a significantly decreased risk of the 28-day mortality rate. When the platelet count was more significant than 127 x10ˆ9/L, an increase in the platelet count per 10 change was not associated with the mortality rate. To our knowledge, this is the first study to report the association between platelet count and 28-day mortality in medical-surgical ICU septic patients.

Most studies investigating the relationship found that platelet count was associated with mortality risk. A prospective cohort study of adult ICU patients in 52 ICUs across ten countries found that patients with thrombocytopenia had worse outcomes, including 90-day mortality and bleeding^[24]. Furthermore, some studies have found that the platelet count is associated with 28-day mortality in different ICU patients. A retrospective study utilizing the eICU database suggested that patients in the neurological ICU who developed thrombocytopenia had a higher risk for in-hospital mortality^[13]. Using the eICU database, another study found that the association between platelet count and in-hospital mortality was curvilinear in critically ill patients with tumors ^[12]. Another study using the eICU database revealed a curvilinear relationship between platelet counts and mortality risk for patients with infectious diseases in the intensive care unit, with an increased risk of mortality when the platelet count fell below 130 × 109/L^[14]. The above studies have investigated the relationship between platelet count and mortality rate, representing a limited perspective on the complex physiological processes in the human body. This relationship cannot be generalized to all patient populations.

Our study revealed that the platelet count was associated with the 28-day mortality rate in medical-surgical ICU patients. And this association was L-shaped. When the platelet count was < 127× 10⁹/L, an increase of 10 units in the platelet count was associated with a 10% decreased risk of a 28-day mortality rate. Nevertheless, this relationship was insignificant when the platelet count was ≥ 127 x10ˆ9/L. This shows that the safe range for the platelet count in medical-surgical ICU patients with sepsis is 127 x10ˆ9/L.

Medical-surgical ICU patients can experience various platelet-related issues, with multiple factors often at play concurrently. The main mechanisms involve (1) hemodilution^[25], (2)immune thrombocytopaenia^{[26, 27]}, (3)blood coagulation mechanism abnormalities^[28], (4)the use of heparin and low-molecular-weight heparin (LMWH)^[29], (5)pseudothrombocytopenia^[30]. This study analyzed data from medical-surgical patients with sepsis in the ICU and identified the relationship between platelet count and the 28-day mortality rate was nonlinear. This information may help guide early clinical interventions in patients who are at risk, thereby prolonging their survival. When the platelet count of medical-surgical ICU patients falls below 127 x 10^9/L, clinicians may pay close attention to their condition and consider whether intervention measures are necessary.

This study has several strengths. First, the large sample size from eICU-CRD strengthened the statistical power and credibility. We utilized multiple logistic regression models to assess the correlation between platelet count and 28-day mortality. Second, we used strict statistical adjustment to reduce residual confounding to avoid the influence of unavoidable potential confounding in this observational study. Third, we used a GAM to investigate a curvilinear relationship between the platelet count and 28-day hospital mortality in medical-surgical ICU patients with sepsis. Fourth, we conducted a two-piece-wise linear regression model to examine the threshold effect of the platelet count and 28-day mortality. Finally, to minimize the bias introduced by missing data, we grouped the missing data into a set of dummy variables and included them in all analyses.

This study has several limitations. First, this was a retrospective study subject to the inherent limitations of a retrospective design. Second, despite employing multivariate logistic regression to adjust for potential confounding variables, some potential confounding factors were not included in the analysis, leading to biased results. Third, due to differences in population characteristics, blood testing methods, ethnic blood traits, and even the timing of data collection, these values vary across different laboratories. Since the data were collected from 2014 to 2015, observation of changes over a longer timeframe is impossible. Fourth, the eICU database only evaluates and treats critical illnesses and does not include information on treating platelets.

The relationship between platelet count and 28-day mortality in medical-surgical ICU patients with sepsis was nonlinear. When patients' platelet count was < 127 'x10ˆ9/L, an increased 10-unit platelet count was associated with a decreased risk of 28-day mortality. This indicates that low platelet count may receive attention in medical-surgical ICU patients with sepsis.

intensive care unit, ICU; Collaborative Research Database, CRD; generalized additive model, GAM; body mass index, BMI; mean arterial pressure, MAP; acute immunodeficiency syndrome, AIDS; chronic obstructive pulmonary disease, COPD; Glasgow Coma Scale, GCS; standard deviation, SD; interquartile ranges, IQR; odds ratios, ORs; confidence intervals, CI; Health Insurance Portability and Accountability Act, HIPAA; low-molecular-weight heparin, LMWH.

Data availability statement

The data are available on the official eICU-CRD website (https://eicu-crd.mit.edu/).

Consent for publication

Not applicable.

Consent to participate

Not applicable.

Ethical approval

After finishing the web-based training courses and the Protecting Human Research Participants examination (No. 13249328), we obtained permission to extract data from the eICU Collaborative Research Database (eICU-CRD). The database is publicly and freely accessible to researchers, according to data usage. As this study is based on a secondary analysis of eICU-CRD data, ethical approval is not required according to Chinese ethical requirements.

Author Contributions

Hai-Yang He designed the research. Yue-Lian Ma analyzed the data, performed the statistical analyses, and wrote the manuscript. Xiong Chen created the tables and figures and revised the manuscript. All the authors reviewed the data, reviewed the manuscript, and approved the ﬁnal manuscript.

Acknowledgements

Thanks to Dr. Yi Li for his suggestions on the language and statistical analysis of the manuscript.

Conflict of interest statement

The authors declare that they have no conflicts of interest.

Funding

There was no funding to support the study

van Vught LA, Klein KP, Spitoni C et al. Incidence, Risk Factors, and Attributable Mortality of Secondary Infections in the Intensive Care Unit After Admission for Sepsis[J]. JAMA 2016,315(14):1469–79.
Cillóniz C, Torres A, Niederman MS. Management of pneumonia in critically ill patients[J]. BMJ,2021,375:e065871.
Zimmerman JE, Kramer AA, Knaus WA. Changes in hospital mortality for United States intensive care unit admissions from 1988 to 2012[J]. Crit Care,2013,17(2):R81.
Pirracchio R, Petersen ML, Carone M, et al. Mortality prediction in intensive care units with the Super ICU Learner Algorithm (SICULA): a population-based study[J]. Lancet Respir Med. 2015;3(1):42–52.
Zarbock A, Weiss R, Albert F et al. Epidemiology of surgery associated acute kidney injury (EPIS-AKI): a prospective international observational multicenter clinical study[J]. Intensive Care Med 2023,49(12):1441–55.
Geense WW, Zegers M, Peters M et al. New Physical, Mental, and Cognitive Problems 1 Year after ICU Admission: A Prospective Multicenter Study[J]. Am J Respir Crit Care Med 2021,203(12):1512–21.
Schell CO, Gerdin WM, Hvarfner A, et al. The global need for essential emergency and critical care[J]. Crit Care. 2018;22(1):284.
Adhikari NK, Fowler RA, Bhagwanjee S et al. Critical care and the global burden of critical illness in adults[J]. Lancet 2010,376(9749):1339–46.
van der Meijden P, Heemskerk J. Platelet biology and functions: new concepts and clinical perspectives[J]. Nat Rev Cardiol 2019,16(3):166–79.
Blumberg N, Cholette JM, Schmidt AE et al. Management of Platelet Disorders and Platelet Transfusions in ICU Patients[J]. Transfus Med Rev 2017,31(4):252–7.
Williamson DR, Lesur O, Tétrault JP et al. Thrombocytopenia in the critically ill: prevalence, incidence, risk factors, and clinical outcomes[J]. Can J Anaesth 2013,60(7):641–51.
Zhao C, Qin Z, Tang Y, et al. Association between platelets and in-hospital mortality in critically ill patients with tumours: a retrospective cohort study[J]. BMJ Open. 2022;12(4):e053691.
Zhou D, Li Z, Wu L, et al. Thrombocytopenia and platelet course on hospital mortality in neurological intensive care unit: a retrospective observational study from large database[J]. BMC Neurol. 2020;20(1):220.
Li J, Li R, Jin X, et al. Association of platelet count with mortality in patients with infectious diseases in intensive care unit: a multicenter retrospective cohort study[J]. Volume 33. Platelets; 2022. pp. 1168–74. 8.
Pollard TJ, Johnson A, Raffa JD, et al. The eICU Collaborative Research Database, a freely available multicenter database for critical care research[J]. Sci Data. 2018;5:180178.
van den Boom W, Hoy M, Sankaran J et al. The Search for Optimal Oxygen Saturation Targets in Critically Ill Patients: Observational Data From Large ICU Databases[J]. Chest,2020,157(3):566–73.
Serpa NA, Deliberato RO, Johnson A et al. Mechanical power of ventilation is associated with mortality in critically ill patients: an analysis of patients in two observational cohorts[J]. Intensive Care Med 2018,44(11):1914–22.
Singer M, Deutschman CS, Seymour CW et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3)[J]. JAMA,2016,315(8):801–10.
Zimmerman JE, Kramer AA, Mcnair DS et al. Acute Physiology and Chronic Health Evaluation (APACHE) IV: hospital mortality assessment for today's critically ill patients[J]. Crit Care Med 2006,34(5):1297–310.
Jaddoe VW, de Jonge LL, Hofman A et al. First trimester fetal growth restriction and cardiovascular risk factors in school age children: population based cohort study[J]. BMJ,2014,348:g14.
Lin L, Chen CZ, Yu XD. [The analysis of threshold effect using Empower Stats software][J]. Zhonghua Liu Xing Bing Xue Za Zhi 2013,34(11):1139–41.
Yu X, Cao L, Yu X. Elevated cord serum manganese level is associated with a neonatal high ponderal index[J]. Environ Res. 2013;121:79–83.
Yu X, Chen J, Li Y, et al. Threshold effects of moderately excessive fluoride exposure on children's health: A potential association between dental fluorosis and loss of excellent intelligence[J]. Environ Int. 2018;118:116–24.
Anthon CT, Pène F, Perner A et al. Thrombocytopenia and platelet transfusions in ICU patients: an international inception cohort study (PLOT-ICU)[J]. Intensive Care Med 2023,49(11):1327–38.
Thiele T, Selleng K, Selleng S, et al. Thrombocytopenia in the intensive care unit-diagnostic approach and management[J]. Semin Hematol. 2013;50(3):239–50.
Cooper N, Ghanima W. Immune Thrombocytopenia[J]. N Engl J Med. 2019;381(10):945–55.
Hunt BJ. Bleeding and coagulopathies in critical care[J]. N Engl J Med 2014,370(9):847–59.
Levi M, Opal SM. Coagulation abnormalities in critically ill patients[J]. Crit Care. 2006;10(4):222.
Zachariah G, Walczyszyn B, Smith K, et al. Characteristics of the post-surgical decrease in platelet counts in orthopedic surgery patients, observations and insights[J]. Hematol Transfus Cell Ther; 2023.
Greinacher A, Selleng S. How I evaluate and treat thrombocytopenia in the intensive care unit patient[J]. Blood 2016,128(26):3032–42.

No competing interests reported.

Download PDF

Editor assigned by journal
08 Jul, 2024
Submission checks completed at journal
08 Jul, 2024
First submitted to journal
04 Jul, 2024

You are reading this latest preprint version

Association of platelet count with 28-day mortality in medical-surgical ICU patients with sepsis: a multicenter retrospective cohort study

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Methods

Data source

Variables

Outcomes

Statistical analysis

Results

Discussion

Conclusion

List Of Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1