SEARCH

SEARCH BY CITATION

Keywords:

  • critically ill;
  • deep vein thrombosis;
  • pulmonary embolism;
  • risk factors;
  • thrombocytosis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

Summary.  Background:  It is uncertain whether reactive thrombocytosis is associated with an increased risk of venous thromboembolism. This study assessed the incidence of reactive thrombocytosis, defined as platelet count ≥ 500 × 109 L−1, at intensive care unit discharge and its association with subsequent venous thromboembolism.

Methods and Results:  This cohort study involved linkage of routinely collected intensive care unit, laboratory, radiology and death registry data of critically ill patients admitted to the intensive care unit between January 2009 and March 2010. The census date for survival and radiologically confirmed venous thromboembolism was 31 October 2011. Of the 1446 patients who survived to intensive care unit discharge, 139 patients had reactive thrombocytosis (9.6%, 95% confidence interval [CI] 8.2–11.2%). Twenty-nine patients developed venous thromboembolism after discharge (2%, 95% CI 1.4–2.9%; 67 per 100 person-years, 95% CI 45–97) and the median time to develop venous thromboembolism was 25 days (interquartile range 8–148). Reactive thrombocytosis was associated with an increased risk of subsequent venous thromboembolism (hazard ratio 5.3, 95% CI 1.7–16.4), after adjusting for other covariates. Platelet counts explained about 34% of the variability in the risk of venous thromboembolism and had a relatively linear relationship with the risk of venous thromboembolism when the platelet counts were > 400 × 109 L−1. Venous thromboembolism after intensive care unit discharge was associated with an increased risk of mortality (hazard ratio 2.0, 95% CI 1.1–3.9), after adjusting for reactive thrombocytosis.

Conclusions:  Reactive thrombocytosis during the recovery phase of critical illness was associated with an increased risk of subsequent venous thromboembolism.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

Reactive thrombocytosis, defined by an abnormally high platelet count in the absence of a chronic myeloproliferative disease, is associated with many medical illnesses, including infection, inflammation and hemorrhage [1]. An elevated platelet count in these situations is primarily driven by an increase in production of thrombopoietin, catecholamines, interleukin-6, interleukin-11 and other cytokines [2]. It is well established that essential thrombocytosis in myeloproliferative diseases is associated with an increased risk of thrombosis due to the presence of giant platelets [1]. In reactive thrombocytosis, the size and function of platelets are normal in nature and, as such, their interactions with the blood vessel wall should remain qualitatively normal, making thrombosis or bleeding no more likely than in patients with normal platelet counts [1,3]. Recent observational studies have, however, reported that reactive thrombocytosis is associated with an increased risk of venous thromboembolism (VTE) in patients after major trauma [4,5]. Whether the association between VTE and reactive thrombocytosis in these studies was mainly due to confounding from the association between severity of injury and reactive thrombocytosis remains uncertain [6].

The incidence of clinically silent VTE, including pulmonary embolism (PE), in critically ill patients is very high despite pharmacological prophylaxis [7]. In one small cohort study, up to 10% of the patients already had unsuspected deep vein thrombosis (DVT) at the time of intensive care unit (ICU) admission [8]. Furthermore, VTE after ICU discharge also appears to be common and may contribute substantially to the morbidity and mortality of critically ill patients [9,10].

We hypothesized that reactive thrombocytosis is common in critical illness and is associated with an increased risk of subsequent VTE during the recovery phase of critical illness. In this linked data cohort study, we assessed the incidence of reactive thrombocytosis at ICU discharge and whether reactive thrombocytosis was associated with an increased risk of subsequent VTE in a heterogeneous group of critically ill patients.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

This was a cohort study. After obtaining approvals from the Clinical Safety and Quality Unit (120314-1) and Ethics Committee (EC 2012/107) of Royal Perth Hospital, the data of all patients who were admitted to the Intensive Care Unit of Royal Perth Hospital, between 1 January 2009 and 23 March 2010, were linked to the laboratory and radiology databases and death registry. Royal Perth hospital is an 800-bed university teaching hospital and the 23-bed multidisciplinary ICU admits about 1500 critically ill adult patients of all specialties per annum. The standard thromboprophylaxis for all critically ill patients included elastic stockings, sequential compression devices to the lower limbs, and unfractionated heparin 5000 units three times per day if there were no contraindications. An anti-platelet agent was not used for reactive thrombocytosis in this study center.

All data were routinely collected for clinical reasons and were subsequently retrieved from the ICU, laboratory and radiology databases. The clinical predictors analyzed included age, sex, Acute Physiology and Chronic Health Evaluation (APACHE) II score and predicted mortality [11], Sequential Organ Failure Assessment (SOFA) scores [12–14], platelet count on admission and discharge, inflammatory markers including white blood cell counts, C-reactive protein and fibrinogen concentrations, chronic medical conditions, and use of inferior vena cava (IVC) filter. Serum CRP concentrations were measured by an immunoenzyme analyzer (Hitachi 917, Tokyo, Japan; normal range, 0–5 mg L−1) and fibrinogen concentrations were measured by the Clauss technique (Diagnostica Stago, Paris, France).

For patients who were readmitted to the ICU during the same hospitalization, only data from the first ICU admission were used. All VTE events included in this study were symptomatic deep vein thrombosis or pulmonary embolism confirmed by color Doppler compression ultrasound and CT pulmonary angiography, respectively. However, only symptomatic VTE events were investigated in this study. The radiologists who reported VTE events in this study were unaware of the study and also the platelet count of the patients at ICU discharge. The date of census for mortality and VTE outcome was 31 October 2011. No patients were lost to follow-up or had missing mortality data.

Statistical analysis

In this study, reactive thrombocytosis was defined as platelet count ≥ 500 × 109 L−1 [15]. The sample size was estimated to have 90% power if the incidence of VTE of patients without reactive thrombocytosis was 3%, incidence of reactive thrombocytosis was 20% and its associated relative risk of VTE was three. The associations between VTE and reactive thrombocytosis or other risk factors were tested with univariable analyses followed by multivariable analysis. Categorical variables and continuous variables with skewed distributions were analyzed using chi-square and Mann–Whitney tests, respectively. After confirming the proportionality assumption of the risk factors for VTE (see online data supplement), Cox proportional hazards regression was used to assess whether reactive thrombocytosis at ICU discharge was associated with an increased risk of subsequent VTE, after adjusting for risk factors for thrombocytosis. Because VTE was infrequent, a parsimony model with predictors associated with a P-value < 0.25 was presented to improve precision and avoid over-fitting.

In a separate Cox proportional hazards regression analysis, we modelled the relationship between platelet counts, as a continuous variable, and hazard of VTE by allowing a non-linear relationship using a 3-knot restricted cubic spline function [16]. This was performed to assess whether a certain threshold of platelet count was needed to increase the risk of VTE and whether their association was ‘dose-related’, which would suggest a cause and effect relationship. The relative contribution of each risk factor to explaining the variability in VTE was assessed using the chi-square statistic minus the degrees of freedom [16]. In this study a P-value < 0.05 was taken as significant. All tests were two-tailed and performed by SPSS for Windows (version 19.0; IBM, Armonk, NY, USA, 2011) and S-PLUS (version 8.0, 2007; Insightful Corp., Seattle, Washington, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

Of the 1599 patients admitted to the ICU during the study period, 1446 (90.4%) and 1268 patients (79.3%) survived to ICU discharge and the date of census, respectively (Fig. 1).

image

Figure 1.  Flow chart showing inclusion and exclusion of patients in the study.

Download figure to PowerPoint

Incidence of VTE in the ICU and after ICU discharge

Of the 1599 patients admitted to the ICU, 14 patients (0.9%, 95% confidence interval [CI] 0.5–1.5%; 58 per 100 person-years, 95% CI 32–98) developed symptomatic VTE (eight DVT, six PE) during their ICU stay. Two of these 14 patients had an inferior vena cava (IVC) filter inserted because of contraindications to systemic anticoagulation for acute pulmonary embolism and one patient developed lower limb DVT 19 day after an IVC filter was inserted for primary VTE prophylaxis. Including these three IVC filters, a total of 56 patients (3.5%) received an IVC filter during their hospital stay; 47 patients (2.9%) needed the filter because of contraindications to pharmacological thromboprophylaxis and another nine patients (0.6%) needed an IVC filter because of contraindications to full systematic anticoagulation after the diagnosis of lower limb DVT. Of the 47 patients who received an IVC filter as primary thromboprophylaxis, four subsequently developed new lower limb or pelvic vein DVT (8.5%, 95% CI 3.4–19.9%), occurring between 9 and 37 days after the placement of the IVC filter. None of 56 patients who had an IVC filter developed PE after the placement of the IVC filter.

Of the 1446 patients who survived to ICU discharge, a total of 29 patients (2%, 95% CI 1.4–2.9%; 68 per 100 person-years, 95% CI 45–97) developed VTE after ICU discharge (17 patients while in hospital and 12 patients after hospital discharge; 19 DVT, 8 PE and 2 with both). The median time to develop VTE after ICU discharge was 25 days (interquartile range 8–148 days).

Incidence of reactive thrombocytosis and its relationship to VTE after ICU discharge

Reactive thrombocytosis occurred in 29 patients (1.9%, 95% CI 1.3–2.7%) on admission to the ICU, but reactive thrombocytosis was substantially more common by the time of ICU discharge (= 139, 9.6%, 95% CI 8.2–11.2%; 579 per 100 person-years, 95% CI 487–684) (Fig. 2). Although the absolute platelet counts on admission to the ICU were not significantly different between those who developed subsequent VTE and those who did not, reactive thrombocytosis at ICU discharge was significantly more common among those who developed VTE after ICU discharge than those who did not (21 vs. 11%, = 0.048) (Table 1). Reactive thrombocytosis remained associated with an increased risk of subsequent VTE (4.3% vs. 1.8%, hazard ratio [HR] 5.3, 95% CI 1.7–16.4; = 0.004), after adjusting for age, gender, APACHE II predicted mortality, SOFA score at ICU discharge, high C-reactive protein concentrations, diabetes mellitus requiring insulin, and leukemia or myeloma (Table 2)(Fig. 3).

image

Figure 2.  Distribution of platelet counts at intensive care unit (ICU) discharge (= 1411). Platelet counts were measured in 109 L−1 and reactive thrombocytosis is defined as platelet counts ≥ 500 × 109 L−1.

Download figure to PowerPoint

Table 1.   Characteristics of patients with and without venous thromboembolism (VTE) after intensive care unit (ICU) discharge (= 1446)
VariablesVTE (Yes, = 29)VTE (No, = 1417) P value*
  1. Values are median unless stated otherwise. APACHE, Acute Physiology and Chronic Health Evaluation; CRP, C-reactive protein; IQR, interquartile range; IVC, inferior vena cava; SOFA, Sequential Organ Failure Assessment. *P values were generated by chi-square or Mann–Whitney tests and are in italics if statistically significant. Data from 755 patients (52%) who had CRP concentrations measured on the day of ICU discharge. Chronic respiratory condition is defined as requiring home oxygen or presence of pulmonary hypertension; cardiovascular disease is defined by the New York Heart Association classification IV status.

Age, years (IQR)54 (33–73)50 (32–65)0.268
Male, no. (%)15 (52)501 (35)0.079
APACHE II score (IQR)21 (18–27)19 (14–24) 0.012
APACHE II predicted mortality (IQR)27.5 (14–52)17.5 (8–38) 0.004
Admission diagnosis:
 Cardiac or respiratory arrest1 (3)43 (3)0.137
 Sepsis1 (3)81 (6)
 Trauma7 (24)238 (17)
 Pneumonia1 (3)86 (6)
 Obstructive airway disease1 (3)37 (3)
 Drug overdose0 (0)138 (10)
 Diabetic ketoacidosis1 (3)4 (0.3)
 Congestive heart failure0 (0)12 (0.8)
 Major vascular or cardiac surgery6 (21)348 (25)
 Others11 (38)348 (25)
Admission SOFA score (IQR)8 (5–10)5 (3–8) 0.007
Maximum SOFA score (IQR)9 (7–10)6 (4–9) 0.001
SOFA on day of ICU discharge (IQR)3 (1–6)2 (1–4)0.288
Platelet count on ICU admission, × 109 L−1 (IQR)232 (133–283)212 (156–286)0.646
No. of patients with platelet count ≥ 500 × 109 L−1on ICU admission (%)1 (3.5)28 (2.0)0.448
Platelet count at ICU discharge,× 109 L−1 (IQR)220 (129–269)223 (155–319)0.938
No. of patients with platelet count ≥ 500 × 109 L−1at ICU discharge (%)6 (21)133 (11) 0.048
White blood cell count at ICU discharge (IQR)11 (9–14)11 (8–14)0.493
Fibrinogen concentrations at ICU discharge (IQR)4.8 (3.3–6.1)4.4 (3.0–6.0)0.651
CRP at ICU discharge, mg per liter (IQR)55 (29–160)85 (37–150)0.553
No. of patients with CRP ≥ 100 mg L−1 (%)6 (35)306 (42)0.804
Co-morbid conditions, No. (%)
 Diabetes mellitus requiring insulin2 (7)46 (3)0.250
 Metastatic cancer0 (0)16 (1)0.999
 Leukemia or myeloma2 (7)14 (1)0.039
 Lymphoma1 (3)8 (0.6)0.167
 History of hepatic failure0 (0)4 (0.3)0.999
 Cirrhosis1 (3)30 (2)0.470
 Chronic respiratory disease0 (0)67 (5)0.642
 Chronic cardiovascular disease5 (17)169 (12)0.383
 End-stage renal failure3 (10)57 (4)0.116
 Immunosuppressive disease0 (0)8 (0.6)0.999
 Immunosuppressive treatment1 (3)35 (2.5)0.522
 HIV infection0 (0)3 (0.2)0.999
IVC filter placement in ICU, No. (%)10 (34)46 (3) 0.001
Length of ICU stay, days (IQR)3.8 (1.1–12.6)1.9 (1.0–5.2)0.063
Length of hospital stay, days (IQR)32 (14–60)13 (6–25) 0.001
Mortality in the ward after ICU discharge, No. (%)4 (14)99 (7)0.146
Mortality by 31 October 2011, No. (%)9 (31)169 (12) 0.006
Survival time after ICU discharge until censored on 31 October 2011 in months (IQR)23 (14–30)27 (22–31) 0.036
Table 2.   Cox proportional hazards regression analysis showing the relationship between reactive thrombocytosis, defined as platelet counts ≥ 500 × 109 L−1, at ICU discharge and subsequent risk of venous thromboembolism (= 1446)
VariablesHazard ratio (95% confidence interval) P value
  1. CRP, C-reactive protein; ICU, intensive care unit; SOFA, Sequential Organ Failure Assessment. Chronic respiratory disease, chronic cardiovascular disease, end-stage renal failure, immunosuppressive treatment, immunosuppressive disease, lymphoma, metastatic cancer, cirrhosis, history of hepatic failure and Acute Physiology and Chronic Health Evaluation (APACHE) II predicted mortality were removed during the modelling process because their associated P values were > 0.25.

Reactive thrombocytosis at ICU discharge5.3 (1.71–16.4)0.004
SOFA score at ICU discharge1.3 (1.18–1.5) (per score increment)0.004
Leukemia or myeloma12.0 (2.5–57)0.002
Diabetes mellitus requiring insulin2.9 (0.6–13.1)0.169
High CRP at ICU discharge (> 100 mg L−1)0.5 (0.2–1.6)0.247
Male2.1 (0.8–5.6)0.128
Age1.0 (0.9–1.1) (per year increment)0.227
image

Figure 3.  Hazard of developing venous thromboembolism over time in patients with and without reactive thrombocytosis (≥ 500 × 109 L−1) at intensive care unit discharge, after adjusting for other covariates.

Download figure to PowerPoint

When platelet counts at ICU discharge were analyzed as a continuous variable in the multivariable analysis, an asymmetrical bimodal relationship between platelet counts and risk of VTE after ICU discharge was observed (Fig. 4). The risk of subsequent VTE appeared to increase linearly with the platelet counts when the platelet counts at ICU discharge were > 400 × 109 L−1. Furthermore, platelet count at ICU discharge explained about 34% of the variability and was more important than other risk factors in explaining the risk of subsequent VTE in the final multivariable model.

image

Figure 4.  Relationship between platelet count at intensive care unit (ICU) discharge and risk of developing thromboembolism, after adjusting for age, gender, myeloma/leukemia, diabetes mellitus requiring insulin therapy, Sequential Organ Failure Assessment score, and a high C-reactive protein concentration (> 100 mg L−1) at ICU discharge. Dotted lines delineate the confidence interval. Platelet counts were measured in 109 L−1.

Download figure to PowerPoint

Other risk factors for VTE after ICU discharge

VTE after ICU discharge was also associated with severity of acute illness leading to admission to the ICU, as reflected by higher Acute Physiology and Chronic Health Evaluation (APACHE) II predicted mortality on admission and maximum Sequential Organ Failure Assessment (SOFA) scores in patients who developed subsequent VTE than those who did not (Table 1). However, in the multivariable analysis only SOFA score at ICU discharge and leukemia or myeloma were other significant risk factors for subsequent VTE (Table 2).

Associations between reactive thrombocytosis or subsequent VTE and long-term survival

Survival after ICU discharge was shorter among those who developed VTE after ICU discharge than those who did not (mortality 31 vs. 12%, = 0.006; median survival time 23 vs. 27 months, = 0.036) (Table 1). In the multivariable analysis, VTE after ICU discharge remained significantly associated with an increased risk of mortality (HR 2.0, 95% CI 1.1–3.9; = 0.045), after adjusting for other prognostic factors. Reactive thrombocytosis was, however, not significantly associated with mortality (HR 1.6, 95% CI 0.9–2.6; = 0.092), after adjusting for occurrence of VTE and other covariates in the final model (Table 3).

Table 3.   Cox proportional hazards regression analysis showing the factors associated with an increased risk of mortality after ICU discharge (= 1446)
VariablesHazard ratio (95% confidence interval) P value
  1. ICU, intensive care unit; SOFA, Sequential Organ Failure Assessment. Chronic cardiovascular disease, end-stage renal failure, immunosuppressive treatment, history of hepatic failure, a high C-reactive protein at ICU discharge and gender were removed during the modelling process because their associated P values were > 0.25.

Reactive thrombocytosis at ICU discharge1.55 (0.93–2.60)0.092
VTE after ICU discharge2.00 (1.10–3.95)0.045
Age1.03 (1.02–1.04) (per year increment)0.001
APACHE II predicted mortality1.19 (1.12–1.26) (per 10% increment)0.001
Metastatic cancer4.53 (2.09–9.80)0.001
Leukemia or myeloma4.25 (2.06–8.77)0.001
Lymphoma4.67 (1.85–11.8)0.001
Diabetes mellitus requiring insulin3.45 (2.15–5.54)0.001
Cirrhosis2.52 (1.30–4.90)0.007
Chronic respiratory disease2.13 (1.30–3.48)0.003
Immunosuppressive disease2.52 (0.77–8.26)0.127
SOFA score at ICU discharge1.04 (0.99–1.10) (per score increment)0.134

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

This cohort study showed that reactive thrombocytosis was common during the recovery phase of critical illness and was associated with an increased risk of subsequent VTE. Symptomatic VTE during the recovery phase of critical illness was not rare (2%) and was associated with an increased risk of mortality after ICU and hospital discharge. These results are significant and require careful consideration.

First, our results are consistent with observational studies on trauma, general medical, cardiology, cardiothoracic and vascular surgical patients, suggesting that reactive thrombocytosis may not be completely innocuous [4,5,17–21]. It is possible that reactive thrombocytosis may, in part, represent a ‘surrogate’ marker of a prothrombotic state, because of its associations with other prothrombotic factors or mediators such as fibrinogen [15,22], resulting in a higher risk of VTE. It is, however, well recognized that reactive thrombocytosis after splenectomy is associated with an increased risk of VTE. In a prospective study of 147 splenectomy patients over a median period of 23 months, 14.3% of the patients with severe reactive thrombocytosis (> 650 × 109 L−1) without myeloproliferative diseases developed portal, mesenteric or splenic vein thrombosis compared with no VTE events among those without severe reactive thrombocytosis [23]. Our recent study also showed that reactive thrombocytosis in critically ill patients was associated with an increased in vitro thrombotic tendency and platelet counts had a relatively linear ‘dose-related’ relationship with the strength of thrombotic tendency when measured by a thromboelastograph (Pearson’s correlation coefficient = 0.53, < 0.001) [15]. This ‘dose-related’in vitro thrombotic tendency of reactive thrombosis is indeed consistent with the findings of this study – the risk of VTE is linearly related to the platelet counts when platelet counts are > 400 × 109 L−1. A recent study also showed that by inducing in vitro thrombocytosis in whole blood donated from healthy volunteers there was an increase in platelet adhesion activity [24] and an observational study has suggested that platelet transfusion was associated with an increased risk of VTE [25]. Furthermore, randomized controlled trials have suggested that aspirin is, in fact, effective in preventing primary and recurrent VTE in patients without reactive thrombocytosis [26,27]. Collectively, these results suggest that platelets may play a more substantial role in the pathogenesis of VTE in the critically ill than previously recognized.

Second, unexpected mortality in critically ill patients after ICU and hospital discharge is common [16,28]. Our results showed that reactive thrombocytosis was common during the recovery phase of critical illness (9.6%) and was associated with an increased risk of VTE that was independently associated with mortality after ICU discharge. As such, VTE may explain, at least in part, why mortality may continue to occur among patients who have an apparent recovery from their critical illness after their ICU discharge. We also noted that only VTE, but not reactive thrombocytosis, was independently associated with mortality, suggesting that reactive thrombocytosis itself may not increase mortality if VTE can be prevented.

Although critically ill patients are known to be at high risk for VTE [8], the epidemiology of VTE after critical illness has not been thoroughly studied [9,10]. Results from a recent small observational study showed that PE was common (6%) in patients with traumatic brain injury after ICU discharge, despite negative twice-weekly lower limb compression ultrasound before their ICU discharge [10]. The incidence of clinically confirmed VTE in our patients after ICU discharge was, in fact, higher than the incidence of VTE during ICU stay (2% vs. 0.9%, respectively). Although IVC filters appeared to be effective in preventing subsequent PE in our patients, lower limb or pelvic vein DVT could still occur (8.5%). As such, clinicians should not be complacent about the risk of VTE during the recovery phase of critical illness and pharmacological thromboprophylaxis is likely to be as important in the recovery phase as in the initial phase of critical illness [29].

The last consideration is the limitations of this study. First, although the sample size of this study was not small, the incidence of VTE was relatively low (2.8% in total), limiting the precision of the results and the power of the study to confirm multiple VTE risk factors. This is a single-center study and, hence, further confirmation of our results by other studies is required. Second, most of our patients had reactive thrombocytosis over a period of time during the recovery phase of their critical illness. Whether the duration of reactive thrombocytosis is as important as the maximum platelet counts in the pathogenesis of VTE during the recovery phase of critical illness remains uncertain. Finally, antiplatelet agents were not used routinely in our patients with reactive thrombocytosis. Whether antiplatelet agents can reduce VTE and improve patient-centered outcomes in patients with reactive thrombocytosis during the recovery phase of critical illness remains uncertain, but this merits further investigation by an adequately powered randomized controlled study.

In summary, reactive thrombocytosis was common during the recovery phase of critical illness and was associated with an increased risk of subsequent VTE. Symptomatic VTE during the recovery phase of critical illness was not rare (2%) and was associated with an increased risk of mortality after ICU and hospital discharge. More research is needed to assess whether using antiplatelet agents for reactive thrombocytosis during the recovery phase of critical illness can prevent VTE and improve other patient-centered outcomes.

Addendum

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

KMH initiated the idea of the study, and was involved in the collection, analysis and interpretation of the data. CBY and OD were involved in the collection and interpretation of data. All authors were involved in the drafting of the manuscript and agreed on the content of the final manuscript. KMH is the guarantor of the study.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

We would like to thank all the intensive care specialists and researcher coordinators for collecting the Acute Physiology and Chronic Health Evaluation II and Sequential Organ Failure Assessment scores that were used in this study. This work was solely funded by Department of Intensive Care Medicine, Royal Perth Hospital, Perth, Australia. No funding was received from the National Institute of Health (NIH), Wellcome Trust or National Health and Medical Research Council (NHMRC).

Disclosure of Conflict of Interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References

The authors state that they have no conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  • 1
    Schafer AI. Thrombocytosis. N Engl J Med 2004; 350: 12119.
  • 2
    Schafer AI. Thrombocytosis and thrombocythemia. Blood Rev 2001; 15: 15966.
  • 3
    Buss DH, Cashell AW, O’Connor ML, Richards F, Douglas Case LD. Occurrence, etiology and clinical significance of extreme thrombocytosis: a study of 280 cases. Am J Med 1994; 96: 24753.
  • 4
    Salim A, Hadjizacharia P, DuBose J, Kobayashi L, Inaba K, Chan LS, Margulies DR. What is the significance of thrombocytosis in patients with trauma? J Trauma 2009; 66: 134954.
  • 5
    Kashuk JL, Moore EE, Johnson JL, Biffl WL, Burlew CC, Barnett C, Sausia A. Progressive postinjury thrombocytosis is associated with thromboembolic complications. Surgery 2010; 148: 66774; discussion 674-675.
  • 6
    Valade N, Decailliot F, Rebufat Y, Heurtematte Y, Duvaldestin P, Stephan F. Thrombocytosis after trauma: incidence, aetiology and clinical significance. Br J Anaesth 2005; 94: 1823.
  • 7
    Schultz DJ, Brasel KJ, Washington L, Goodman LR, Quickel RR, Lipchik RJ, Clever T, Weigelt J. Incidence of asymptomatic pulmonary embolism in moderately to severely injured trauma patients. J Trauma 2004; 56: 72731; discussion 731-733.
  • 8
    Cook DJ, Crowther MA. Thromboprophylaxis in the intensive care unit: focus on medical-surgical patients. Crit Care Med 2010; 38: S7682.
  • 9
    Ho KM, Burrell M, Rao S, Baker R. Incidence and risk factors for fatal pulmonary embolism after major trauma: a nested cohort study. Br J Anaesth 2010; 105: 596602.
  • 10
    Praeger AJ, Westbrook AJ, Nichol AD, Wijemunige R, Davies AR, Lyon SM, Wills JL, Bailey M, Rosenfeld JV, Cooper DJ. Deep vein thrombosis and pulmonary embolus in patients with traumatic brain injury: a prospective observational study. Crit Care Resusc 2012; 14: 103.
  • 11
    Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13: 81829.
  • 12
    Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, Reinhart CK, Suter PM, Thijs LG. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996; 22: 70710.
  • 13
    Kajdacsy-Balla Amaral AC, Andrade FM, Moreno R, Artigas A, Cantraine F, Vincent JL. Use of the sequential organ failure assessment score as a severity score. Intensive Care Med 2005; 31: 2439.
  • 14
    Ho KM. Combining sequential organ failure assessment (SOFA) score with acute physiology and chronic health evaluation (APACHE) II score to predict hospital mortality of critically ill patients. Anaesth Intensive Care 2007; 35: 51521.
  • 15
    Duff O, Ho KM, Maybury SM. In-vitro thrombotic tendency of reactive thrombocytosis in critically ill patients: a prospective case-control study. Anaesth Intensive Care 2012; 40: 4728.
  • 16
    Ho KM, Knuiman M, Finn J, Webb SA. Estimating long-term survival of critically ill patients: the PREDICT model. PLoS ONE 2008; 3: e3226.
  • 17
    Schmuziger M, Christenson JT, Maurice J, Simonet F, Velebit V. Reactive thrombocytosis after coronary bypass surgery. An important risk factor. Eur J Cardiothorac Surg 1995;9:3937; discussion 397-398.
  • 18
    Lancaster RT, Conrad MF, Patel VI, Cambria RP, Lamuraglia GM. Predictors of early graft failure after infrainguinal bypass surgery: a risk-adjusted analysis from the NSQIP. Eur J Vasc Endovasc Surg 2012; 43: 549555.
  • 19
    Tchebiner JZ, Nutman A, Boursi B, Shlomai A, Sella T, Wasserman A, Guzner-Gur H. Diagnostic and prognostic value of thrombocytosis in admitted medical patients. Am J Med Sci 2011; 342: 395401.
  • 20
    Vidwan P, Lee S, Rossi JS, Stouffer GA. Relation of platelet count to bleeding and vascular complications in patients undergoing coronary angiography. Am J Cardiol 2010; 105: 121922.
  • 21
    Iijima R, Ndrepepa G, Mehilli J, Bruskina O, Schulz S, Schömig A, Kastrati A. Relationship between platelet count and 30-day clinical outcomes after percutaneous coronary interventions. Pooled analysis of four ISAR trials. Thromb Haemost 2007; 98: 8527.
  • 22
    Schafer AI. Thrombocytosis: when is an incidental finding serious? Cleve Clin J Med 2006; 73: 76774.
  • 23
    Stamous KM, Toutouzas KG, Kekis PB, Nakos S, Gafou A, Manouras A, Krepis E, Katsaragakis S, Bramis J. Prospective study of the incidence and risk factors of postsplenectomy thrombosis of the portal, mesenteric, and splenic veins. Arch Surg 2006; 141: 6639.
  • 24
    Peerschke EI, Silver RT, Weksler BB, Yin W, Bernhardt B, Varon D. Examination of platelet function in whole blood under dynamic flow conditions with the cone and plate(let) analyzer: effect of erythrocytosis and thrombocytosis. Am J Clin Pathol 2007; 127: 4228.
  • 25
    Cook D, Crowther M, Meade M, Rabbat C, Griffith L, Schiff D, Geerts W, Guyatt G. Deep venous thrombosis in medical-surgical critically ill patients: prevalence, incidence, and risk factors. Crit Care Med 2005; 33: 156571.
  • 26
    Becattini C, Agnelli G, Schenone A, Eichinger S, Bucherini E, Silingardi M, Bianchi M, Moia M, Ageno W, Vandelli MR, Grandone E, Prandoni P, WARFASA Investigators. Aspirin for preventing the recurrence of venous thromboembolism. N Engl J Med 2012; 366: 195967.
  • 27
    Becker RC. Aspirin and the prevention of venous thromboembolism. N Engl J Med 2012; 366: 202830.
  • 28
    Ho KM, Lee KY, Dobb GJ, Webb SA. C-reactive protein concentration as a predictor of in-hospital mortality after ICU discharge: a prospective cohort study. Intensive Care Med 2008; 34: 4817.
  • 29
    Ho KM, Chavan S, Pilcher D. Omission of early thromboprophylaxis and mortality in critically ill patients: a multicenter registry study. Chest 2011; 140: 143646.