Thrombocytosis in intensive care

A. M. Gurung, B. Carr and I. Smith*

Directorate of Anaesthesia and Intensive Care, North Staffordshire Hospital, Stoke-on-Trent ST4 6QG, UK*Corresponding author

Accepted for publication: August 13, 2001


    Abstract
 Top
 Abstract
 Introduction
 Methods and results
 Comment
 References
 
We conducted a retrospective study of platelet count in 226 patients admitted for critical care over a 5-month period, to explore the incidence of thrombocytosis and its relation to admission category, duration of ICU stay and outcome. Our findings indicate that thrombocytosis is not rare in ICU patients. At least one platelet count greater than 450x109 litre–1 was found in 21.7% of patients and was associated with lower ICU mortality (P=0.003), lower hospital mortality (P=0.006), but longer duration of ICU stay (P<0.0001). Thrombocytosis may serve as an independent predictor of favourable outcome in ICU patients.

Br J Anaesth 2001; 87: 926–8

Keywords: intensive care, mortality; blood, platelets, thrombocytosis


    Introduction
 Top
 Abstract
 Introduction
 Methods and results
 Comment
 References
 
Marked variations of platelet count between patients and in the same patient over time are common in intensive care unit (ICU) patients. The causes of thrombocytopenia have been extensively reviewed1 and thrombocytopenia has been shown to be predictive of an adverse outcome.2 We are unaware of any previous literature specific to the prognostic value of thrombocytosis in intensive care medicine.

The purpose of this study was to explore the incidence of thrombocytosis and its relation to duration of ICU stay and outcome. Thrombocytosis may serve as an additional variable in prediction of outcome in ICU patients.


    Methods and results
 Top
 Abstract
 Introduction
 Methods and results
 Comment
 References
 
All patients admitted to the general ICU (GICU) and trauma ICU (TICU) of the North Staffordshire Hospital, Stoke-on-Trent, UK from January 1 to May 31, 2000 were included in this retrospective study. Data related to age, sex, duration of ICU stay, outcome, admission category, APACHE II score, and predicted mortality were obtained from HISS (Hospital Information Support System) and Acubase – ICNARC (Intensive Care National Audit Research Centre) databases. Duration of stay was rounded to the nearest whole day, with any duration of stay less than 24 h counted as 1 whole day. If the patient was admitted more than once, or transferred from one unit to another, then it was considered as a single admission. The initial platelet count (for the first admission) and the highest and the lowest platelet counts during the whole duration of ICU stay were recorded. On the basis of these results, patients were retrospectively divided into three groups: those with a platelet count less than 150x109 litre–1 on at least one occasion; those whose platelet count remained between 150 and 450x109 litre–1 throughout; and those whose platelet count exceeded 450x109 litre–1 on any occasion. The day of ICU stay on which the latter occurred was also recorded. Unless contraindicated, thromboprophylaxis with graduated compression stockings and subcutaneous heparin was used in all patients.

Continuous variables are expressed as number and percentage of the group from which they were obtained, or median and interquartile range. Continuous variables were analysed using the Kruskal–Wallis test. The Mann–Whitney test was used to compare duration of ICU stay between patients with early and late onset of thrombocytosis. Discrete variables were analysed using chi-squared tests. Statistical analyses were performed using StatView® version 4.02 for Macintosh and a P value of <0.05 was considered significant.

A total of 231 patients were admitted during the study period: to the GICU, 104 and to the TICU, 127. Four patients with no documented platelet count and one patient transferred to an ICU in another hospital were excluded. The platelet count of 226 patients and admission category, predicted mortality, ICU outcome and hospital outcome were analysed (Table 1). Thrombocytosis, defined as a platelet count of greater than 450x109 litre–1 on at least one occasion, was observed in 49 patients (21.7%). The median onset of thrombocytosis was 7 (1–11) days after ICU admission, although 39% of patients either had thrombocytosis at ICU admission or developed this within 4 days. Patients with at least one platelet count higher than 450x109 litre–1 had lower ICU and hospital mortality but a longer duration of ICU stay, despite similar predicted mortality (Table 1). The median (range) duration of ICU stay (18.5, 13–33 days) was significantly longer in patients who developed thrombocytosis after 4 or more days in the ICU compared with those with an earlier thrombocytosis (3, 1–6 days; P<0.0001), although the mortality of these two subgroups did not differ significantly.


View this table:
[in this window]
[in a new window]
 
Table 1 Patient characteristics in the three groups. Values are number (percentage) of occurrences or median (interquartile range). Between group comparisons were analysed using the Kruskal–Wallis test{dagger} for continuous variables, and the chi-squared test§ for discrete variables. A three by three chi-squared test was performed to study the distribution of medical surgical and trauma patients between the three groups (P=0.061)
 

    Comment
 Top
 Abstract
 Introduction
 Methods and results
 Comment
 References
 
Thrombocytosis is common in ICU patients. Among our patients, 21.7% had at least one platelet count greater than 450x109 litre–1. Thrombocytosis can be primary or reactive to a variety of causes. Extreme thrombocytosis should not be considered a rare event in a general, acute care hospital population, and usually represents a reactive phenomenon.3 Reactive thrombocytosis has been reported to occur as a response to various conditions, such as post-splenectomy and in haematopoietic disorders, major trauma and surgery, neoplasms, and inflammation.4 Increases in platelet count have previously been shown to be associated with improved survival in ICU,5 although none of those patients had platelet counts in excess of 450x109 litre–1. Indeed, most of the deaths in that series5 occurred in patients with thrombocytopenia, which is well recognized to be associated with poor survival.2

We are unaware of any previous literature describing the association between a high platelet count and outcome in critically ill patients. We found that a platelet count higher than 450x109 litre–1 on at least one occasion during the ICU stay was associated with lower ICU and hospital mortality. This was equally true for patients presenting to ICU with an established thrombocytosis and for those developing it later. Indeed, a small cohort of patients (n=14) developed a thrombocytosis after presenting with an initial platelet count of less than 150x109 litre–1. These patients were included with the thrombocytosis group, but even when analysed separately, they were still associated with a more favourable outcome compared with patients without a thrombocytosis. As thrombocytopenia is associated with poor outcome,2 we compared our patients with thrombocytosis against patients with thrombocytopenia as well as those with a normal platelet count throughout; outcome was more favourable compared with either group. Overall, thrombocytosis was associated with a prolonged ICU stay, although this was far less marked in patients presenting to ICU with thrombocytosis or developing it at an early stage.

Despite these findings, this study has some limitations. It is a retrospective study. Some patients were not on the ICU for sufficient time for all data required to calculate predicted mortality to be collected. However, the distribution of these patients did not differ between the three groups defined by platelet counts. Being the major trauma receiving hospital in the region, the proportion of patients admitted with multiple trauma including head injury requiring neurosurgery is high, which is the reason for grouping these categories of patient together in our study. Thrombocytosis was more common in this patient category.

In conclusion, thrombocytosis is a common finding in ICU. It appears to be associated with lower ICU and hospital mortality but longer duration of ICU stay. It may serve as an additional variable in predicting outcome in ICU patients. A large prospective study is required to confirm these findings.


    References
 Top
 Abstract
 Introduction
 Methods and results
 Comment
 References
 
1 Drews RE, Weinberger SE. Thrombocytopenic disorders in critically ill patients. Am J Resp Crit Care Med 2000; 162: 347–51[Free Full Text]

2 Vanderschueren S, De Weerdt A, Malbrain M, et al. Thrombocytopenia and prognosis in intensive care. Crit Care Med 2000; 28: 1871–6[ISI][Medline]

3 Buss DH, Cashell AW, O’Connor ML, Richard F, II, Case LD. Occurrence, etiology, and clinical significance of extreme thrombocytosis: A study of 280 cases. Am J Med 1994; 96: 247–53[ISI][Medline]

4 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: 393–7[Abstract]

5 Nijsten MWN, ten Duis H-J, Zijlstra JG, et al. Blunted rise in platelet count in critically ill patients is associated with worse outcome. Crit Care Med 2000; 28: 3843–6[ISI][Medline]