SEARCH

SEARCH BY CITATION

Keywords:

  • haemorrhage;
  • haematological malignancies;
  • transfusion;
  • rotational thromboelastometry;
  • thrombocytopenia

Summary

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information

Previous studies have shown that total platelet count (TPC) inadequately predicts bleeding in thrombocytopenic patients with haematological malignancies. This prospective cohort study evaluated whether rotational thromboelastometry (ROTEM), coagulation or other platelet parameters were more strongly associated with bleeding than TPC. Adults treated at two UK haematology centres for haematological malignancy were enrolled if they had thrombocytopenia (TPC ≤ 50 × 109/l) at beginning of, or during treatment (International Standard Randomized Controlled Trial Number 81226121). TPC and bleeding symptoms were recorded daily for up to 30 d or until platelet count recovery, hospital discharge or death. Blood samples were tested thrice weekly using ROTEM, Platelet Function Analyser (PFA)-100®, coagulation and platelet cytometry assays. Bleeding symptoms and TPC from 49/50 enrolled participants who completed the study were recorded on 754/760 study days. Mean platelet volume and PFA-100® closure times were frequently inestimatable because of thrombocytopenia. TPC, absolute immature platelet number (AIPN) and ROTEM maximum clot firmness were significantly associated with bleeding on the day after blood sampling. Only AIPN was associated with bleeding after adjustment of test results for TPC (Odds Ratio 0·52, 95% confidence interval 0·28–0·97; P = 0·038). In a predictive model, AIPN was superior to TPC for predicting bleeding. This study indicates that AIPN may be more clinically useful than TPC at predicting bleeding.

Haematological malignancies constitute 8–9% of all new cancers diagnosed in the UK and US (http://www.cdc.gov/cancer/npcr/uscs/download_data.htm, http://www.ons.gov.uk/ons/rel/cancer-unit/cancer-incidence-and-mortality/2008-2010/index.html), and have increased in incidence by 10–14% in the last decade (http://www.cancerresearchuk.org/cancer-info/cancerstats/incidence/). The prevalence of haematological malignancies is also increasing because advances in intensive chemotherapy and haemopoietic stem cell transplantation have improved survival rates (Milligan et al, 2006; Fielding et al, 2007; Burnett et al, 2011). Despite these developments, both haematological malignancies and their treatment may lead to periods of severe thrombocytopenia (≤50 × 109/l) (Heddle et al, 2009; Wandt et al, 2012; Stanworth et al, 2013). During this thrombocytopenic period, patients are at risk of severe and life-threatening bleeding (Heddle et al, 2009; Slichter et al, 2010).

In an attempt to minimize bleeding, most haematology treatment centres across the developed world transfuse prophylactic platelet components to patients with haematological malignancy, usually when the platelet count falls below 10 x 109/l (Schiffer et al, 2001; BCSH, [BCSH] 2003; National Blood Authority, 2012). Platelets are currently the second most commonly issued allogeneic blood component (312 000 adult doses in UK last year) (Bolton-Maggs et al, 2013), of which up to 67% are issued to patients with haematological malignancies (Greeno et al, 2007; Pendry & Davies, 2011). The majority of these are given to prevent, rather than to treat, bleeding (Estcourt et al, 2012). In contrast to other blood components, demand for platelets has increased by 24% over the last decade in the UK, and by 25% in the United States from 2006 to 2011 (Taylor et al, 2010; Whitaker et al, 2011; Bolton-Maggs et al, 2013; Whitaker & Hinkins, 2013). Platelet components are a costly and limited resource (Williamson & Devine, 2013) and are associated with significant risks to recipients (Popovsky & Moore, 1985; Blumberg et al, 2009; Heddle & Webert, 2009).

In current practice, the decision to transfuse platelet components prophylactically is primarily based on the platelet count determined from daily blood tests. However, despite this strategy, bleeding remains prevalent and occurs across a wide range of platelet counts (Slichter et al, 2010; Wandt et al, 2012; Stanworth et al, 2013). In some patient groups, bleeding occurred at a similar frequency and severity irrespective of whether or not patients received prophylactic platelet transfusions according to platelet count (Stanworth et al, 2013). The most plausible explanation for these observations are that the platelet count is an inadequate marker of platelet haemostatic function in this clinical setting, and that other defects in haemostasis contribute significantly to bleeding.

The aim of our study was to evaluate whether other laboratory markers of haemostasis showed a stronger association with bleeding than the platelet count in severely thrombocytopenic patients receiving treatment for haematological malignancies. In order for laboratory markers to be clinically useful in directing treatments to prevent bleeding, each marker must be detectible using rapid, robust and inexpensive laboratory tests. Therefore, we evaluated a panel of markers that can be detected using point-of care format ROTEM thromboelastometry and Platelet Function Analyser (PFA)-100 devices or by using automated coagulation and haematology analyser tests that are in widespread use in other clinical settings. By selecting a heterogeneous cohort of patients with haematological malignancy and collecting data throughout treatment we aimed to identify predictive laboratory markers that were potentially informative with all patients in all clinical situations, irrespective of the numerous additional clinical factors which influence bleeding risk in this group.

Methods and materials

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information

The ATHENA (Risk factors for bleeding in haematology patients with low platelet counts) study (International Standard Randomized Controlled Trial Number 81226121) was an observational cohort study of 50 consecutive adults (≥16 years) with haematological malignancies admitted for intensive chemotherapy or haemopoietic stem cell transplant at the Oxford or Bristol haematology centres between September 2010 and September 2012. Patients were eligible for the study if they were thrombocytopenic (total platelet count (TPC) ≤ 50 × 109/l) at admission, or were expected to become thrombocytopenic for ≥5 d during inpatient treatment, and were able to comply with study monitoring. Patients were excluded from the study if they had any inherited disorder of haemostasis, required anti-thrombotic medication during the period of thrombocytopenia or had a prior diagnosis of immune thrombocytopenia. Potential participants were approached at admission, checked for eligibility and invited to give written informed consent prior to study enrolment in accordance with UK Regional Ethics Committee approval 10/H0505/47.

All study participants were managed using standard institutional protocols for the duration of the study. This included prophylactic transfusion of 1 adult platelet unit on each day that a morning blood sample showed TPC ≤ 10 × 109/l. Additional platelet units were administered in response to bleeding of World Health Organization (WHO) Grade ≥2 and to prevent bleeding before invasive procedures (BCSH, 2003). For each participant, the study observation period was from the first day of thrombocytopenia until platelet count recovery (unsupported TPC > 50 × 109/l for >3 d), hospital discharge, death, or for 30 d if there was prolonged thrombocytopenia.

During the observation period, participants underwent daily bleeding assessments and measurement of TPC. Further blood samples were collected thrice weekly for additional coagulation tests into Vacutainer® blood collection tubes (Becton Dickinson, Plymouth, UK) containing either 0·109 mol/l buffered sodium citrate (for ROTEM, PFA-100®, prothrombin time (PT), activated partial thromboplastin time (aPTT) and Clauss fibrinogen (CF) tests) or 7·2 mg EDTA (for platelet parameters). The whole blood laboratory tests were performed within 2 h of specimen collection. For the plasma coagulation tests, blood was centrifuged twice within 2 h of specimen collection and plasma was stored at −80°C for later batched analysis at the Oxford Haemophilia and Thrombosis Centre.

Measurement of bleeding

Bleeding symptoms were recorded using a standardized questionnaire identical to that used in a previous randomized controlled trial (Stanworth et al, 2013). A computerized algorithm was used to classify bleeding symptoms according to modified WHO criteria in which WHO grades 14 indicate bleeding with increasing severity (Stanworth et al, 2010, 2013).

Analysis of blood samples

Platelets were tested using a Sysmex XE2100 automated haematology analyser (Sysmex, Kobe, Japan) to determine the following platelet parameters- (i) Total platelet count (TPC); (ii) Mean platelet volume (MPV) (Kaito et al, 2005); (iii) Immature platelet fraction (IPF- the percentage of circulating platelets with above-threshold RNA) (Briggs et al, 2006; Barsam et al, 2011; Bat et al, 2013) and, (iv) Absolute immature platelet number (AIPN- the concentration of immature circulating platelets, calculated as IPF x TPC) (Barsam et al, 2011; Bat et al, 2013).

Thromboelastometry was performed on a ROTEM delta instrument (TEM International, Munich, Germany) with the EXTEM (thromboplastin-initiated coagulation), INTEM (contact factor-initiated coagulation) and FIBTEM (thromboplastin-initiated coagulation with the platelet inhibitor cytochalasin D) activating reagents in accordance with the manufacturer's instructions (Ganter & Hofer, 2008). For EXTEM and INTEM, the clot time (CT), clot formation time (CFT), maximum clot firmness (MCF) and maximal lysis (ML) were derived from the thromboelastometry traces. For FIBTEM only MCF and ML were derived from the thromboelastometry traces.

Platelet function was assessed by determining the PFA-100® closure times (Siemens Healthcare Diagnostics GmBH, Marburg, Germany) with the 150 μm aperture Collagen/ADP (C-ADP) cartridge, and with the INNOVANCE® PFA P2Y (aperture 100 μm, membrane coated with 20 μg Adenosine diphosphate (ADP), 5 ng prostaglandin E1 and 459 μg calcium chloride). The PT, aPTT and Clauss Fibrinogen were determined using a STA-R Evolution® Expert Series Hemostasis System (Diagnostica Stago S.A.S., Asnières sur Seine, France) using reagents and methodology according to the manufacturer's instructions.

Statistical analysis

The sample size of 50 patients was pre-specified to enable inclusion of patients with different types of haematological malignancy and treatment. Since this was a pilot study assessing previously unexplored associations between laboratory parameters and bleeding, a power calculation was not performed. Statistical analysis was performed with stata version 11·2 (StataCorp, Texas, USA). All data were non-parametric and were expressed as medians and interquartile ranges [IQRs]. Laboratory parameters were analysed as continuous variables. Associations between laboratory parameters and bleeding during the following 1 d and during the following 2 d after each blood sample were tested using a logistic regression model clustered on patient identity to account for repeated measures. This type of analysis increases the power of the study to detect a difference because the effect of a change in a laboratory parameter is measured within the same patient. Laboratory parameters other than the TPC were considered potentially predictive of bleeding if there was a statistically significant odds ratio (P < 0·05) for bleeding when adjusted for the TPC using a multivariate logistic regression model clustered on patient identity. Receiver operating characteristic (ROC) curves were derived for TPC and other potentially predictive laboratory parameters and were used to calculate the area under the curve (AUC) and 95% confidence intervals (CIs). Sensitivity, specificity, positive predictive values (PPVs) and negative predictive values (NPVs) were also calculated for various pre-specified thresholds of the TPC and any potential alternative parameter. No adjustment for confounders, such as underlying patient characteristics (e.g. type of treatment or haematological malignancy), was made because in current clinical practice, these variables are not considered when using TPC to guide administration of platelet transfusions.

Results

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information

The 50 participants enrolled into the study had mean age of 51·0 years (range 2070) and 33 were male. The underlying diagnosis was leukaemia in 16/50 participants, lymphoma in 14/50, myeloma in 9/50 and other haematological disorders in 11/50. Treatment was with intensive chemotherapy for 4/50 participants, myeloablative allograft for 10/50, reduced intensity conditioning allograft for 23/50 and autograft for 13/50. One participant, who received induction chemotherapy for acute myeloid leukaemia, was reallocated to outpatient management after study enrolment and was withdrawn from the study. The remaining 49 participants were followed throughout their study observation periods, for a total of 760 d (Fig 1). Bleeding symptom data and the TPC on the preceding day were available for a total of 757 (99·6%) and 754 (99·2%) study days respectively and are summarized in Table SI. The participants received antibiotics on 478/760 (63%) of study days (Fig 2E). The median duration of antibiotic therapy was 10 d. None of the participants received anticoagulants, anti-platelet agents or asparaginase during the study period.

image

Figure 1. ATHENA study flow diagram. TPC, total platelet count. *Patient with acute myeloid leukaemia being treated with induction chemotherapy.

Download figure to PowerPoint

image

Figure 2. The characteristics of bleeding symptoms and total platelet count in the 49 study participants. (A) The highest WHO bleeding grade for each participant during the study observation period. (B) Kaplan–Meier survival curve showing time from the start of thrombocytopenia (TPC ≤ 50 × 109/l) to first bleeding episode in days. The numbers on the curve represent censored patients because they had left the study at that time point but without bleeding. These could not be classified as ‘non-bleeders from the point of leaving the study because no bleeding score data were collected. (C) Distribution of the number of days that the participants had TPC ≤ 50 × 109/l during the study period. (D) Distribution of the number of days that the participants had TPC ≤ 20 × 109/l during the study period. (E) Distribution of the number of days the participants had TPC ≤ 10 × 109/l during the study period. (F) Distribution of the number of days the participants had therapeutic antibiotics during the study period. (G) Distribution of the number of days the participants had platelet transfusions during the study period. (H) Distribution of the number of days the participants had red cell transfusions during the study period. WHO, World Health Organization; TPC, total platelet count.

Download figure to PowerPoint

The participants had a median 3 d of bleeding of any WHO severity (IQR 0–6). A total of 65% of participants experienced at least one bleeding episode of WHO Grade I severity but 33% experienced at least one episode of bleeding of WHO Grade ≥2 (Fig 2A). The median interval between the start of thrombocytopenia and the first bleeding episode was 5 d (IQR 2–10 d; Fig 2B). The median total duration of thrombocytopenia was 11 d (IQR 8–16) for TPC ≤ 50 × 109/l and 4 d (IQR 2–8) for TPC ≤ 20 × 109/l (Figs 2C–D).

The participants received a total of 175 adult platelet units, administered as 163 separate transfusion episodes during the study observation period. These comprised 124/175 (71%) units for prophylaxis, 34/175 (19%) in response to bleeding and 17/175 (10%) to prevent bleeding during an invasive procedure. In keeping with standard care, most patients received unselected, leucocyte-reduced, apheresis platelet components that were usually ABO and RhD identical. Two patients received human leucocyte antigen (HLA)-matched platelets because of a previous history of platelet refractoriness. The 24-h platelet increments were calculated for 162 of the 163 separate transfusion episodes in which 175 adult platelet units were administered. The median 24-h platelet count increment was 11 × 109/l. In the 1 d interval after prophylactic platelet transfusions, any bleeding occurred on a total of 24 d (21% of all evaluable days after prophylactic transfusion) and WHO Grade ≥2 bleeding on 3 d (2·7%) (Table SII).

A total of 157 red cell components were given as 100 separate transfusion episodes during the study observation period. Only 1 transfusion was given in immediate response to bleeding. The median pre-transfusion haemoglobin was 81 g/l. None of the participants received plasma transfusions during the study period.

Completion of laboratory tests

A total of 278 additional blood samples were obtained from the participants during the study. The TPC, IPF and AIPN platelet parameters and the PT, aPTT, CF were determined for more than 98% of these blood samples. There was a similar very high completion rate for the parameters in the ROTEM tests, including the MCF, which may be markedly reduced in samples with a low TPC. This indicates that ROTEM is an appropriate test for evaluating haemostasis in this patient group. A summary of the laboratory results from the study cohort is shown in Table 1.

Table 1. Summary of laboratory parameters from a total of 276 blood samples obtained from 49 study participants during the study observation period.
 Rotational thromboelastography
 MedianIQRReference interval
EXTEM CT (s)5951 to 7042 to 74
EXTEM CFT (s)16686 to 28346 to 148
EXTEM MCF (mm)4537 to 5249 to 71
EXTEM ML (%)31 to 60 to 18
INTEM CT (s)156144 to 170137 to 246
INTEM CFT (s)16287 to 29540 to 100
INTEM MCF (mm)4335 to 5152 to 72
INTEM ML (%)20 to 40 to 12
FIBTEM MCF (mm)2318 to 299 to 25
FIBTEM ML (%)00 to 1
 Platelet parameters
 MedianIQRReference interval
TPC (x 109/l)2615 to 48140 to 400
IPF (%)4·92·8 to 7·30·9 to 5·2
AIPN (x 109/l)1·120·65 to 2·492·7 to 12·5
 Plasma coagulation tests
 MedianIQRReference interval
  1. EXTEM, thromboplastin-initiated coagulation; INTEM, contact factor-initiated coagulation; FIBTEM, thromboplastin-initiated coagulation with the platelet inhibitor cytochalasin D; CT, clot time; CFT, clot formation time; MCF, maximum clot firmness; ML, % maximal lysis; TPC, total platelet count; IPF, immature platelet fraction; AIPN, absolute immature platelet number; PT, prothrombin time; aPTT, activated partial thromboplastic time; IQR, interquartile range.

PT (s)13·711·3 to 15·313 to 16
aPTT (s)31·427·4 to 34·526 to 36
Fibrinogen (g/l)4·33·3 to 5·51·4 to 4·0

We selected the PFA-100 assay system as a potential approach of identifying defects in primary haemostasis in the patient cohort. PFA-100 tests using cartridges with standard aperture size (150 μm) are markedly affected by thrombocytopenia (Carcao et al, 2002). However, the recently available INNOVANCE® PFA P2Y cartridge is potentially more robust to the effects of thrombocytopenia because of a smaller membrane aperture (100 μm), but was untested in this patient group. The PFA-100® test reached a closure time endpoint in only 9·3% of samples using the standard CADP cartridge, and in 22·4% of samples using the INNOVANCE® PFA P2Y cartridge when the platelet count was ≤50 × 109/l. The median closure times, when reported, were 149·5 s (IQR 108·5–197·5 s) for the CADP cartridge and 105 s (IQR 70·5–155 s) for the INNOVANCE® PFA P2Y cartridge. The MPV was not reported by the Sysmex XE2100 analyser in 63·8% of samples with a platelet count ≤50 × 109/l. No further analyses of the PFA-100 closure times or MPV was performed because the poor completion rates of these tests prevented robust statistical analysis.

Platelet parameters

There was a significant association between bleeding (any WHO grade) within 1 d of blood sampling and the total platelet count (TPC). For every rise in the TPC of 10 × 109/l, the odds of bleeding decreased by 24% (odds ratio [OR] 0·76, 95% CI 0·62 to 0·93; P = 0·008). For every rise in the AIPN of 0·5 x 109/l, the odds of bleeding decreased by 26% (OR 0·74 95% CI 0·61 to 0·90; P = 0·002), but this effect was not seen with the IPF (OR 0·99, 95% CI 0·93 to 1·06; P = 0·824). There was also a significant association between bleeding (WHO grade ≥2) within 1 d of blood sampling and the AIPN and the TPC but not the IPF (Table 2). Data for the unadjusted TPC was calculated using all 754 study days, which reported both the TPC and bleeding assessment data, whereas data for AIPN was only available for 277 d.

Table 2. Associations between platelet parameters and WHO grade 2 or above bleeding within 1 d of blood sampling.
 WHO Grade ≥2 bleeding
 Unadjusted odds ratioOdds ratio adjusted for TPCOdds ratio adjusted for AIPN
 Odds ratioa95% CIP-valueOdds ratioa95% CIP-valueOdds ratioa95% CIP-value
  1. Unadjusted odds ratios and 95% confidence intervals (CI) for bleeding of World Health Organization (WHO) grade ≥2 in the day following blood sampling and total platelet count (TPC), immature platelet fraction (IPF) and absolute immature platelet number (AIPN), as well as OR adjusted for TPC and AIPN. All odds ratios account for repeated measures.

  2. *= P < 0·05; = **P < 0·01.

  3. a

    The odds ratio for TPC is for every 10 × 109/l rise in platelet count. The odds ratio for IPF is for every percent rise in IPF. The odds ratio for AIPN is for every 0·5 × 109/l rise in AIPN.

TPC0·690·52 to 0·920·012*0·830·57 to 1·220·340
IPF0·950·82 to 1·100·4580·920·79 to 1·080·3181·100·99 to 1·220·073
AIPN0·480·28 to 0·810·007**0·520·28 to 0·970·038*

The association between bleeding (WHO grade ≥2) within 1 d of blood sampling and AIPN remained statistically significant after adjustment of AIPN for TPC using logistic regression analysis clustered on patient identity (Table 2). However, the association between bleeding (WHO grade ≥2) within 1 d of blood sampling and TPC was no longer significant after adjustment for AIPN (Table 2). Therefore the best ‘model’ in this analysis for WHO grade 2 or above bleeding was AIPN alone. Both the AIPN and TPC were significantly associated with bleeding (WHO grade ≥2) within 2 d of blood sampling, but this effect was lost after adjustment for TPC or AIPN respectively (Table SIII).

The presence of bleeding within 1 d of blood sampling within pre-specified TPC and AIPN categories were estimated with 95% CI (Figs 3A–B). For AIPN in the range 0–0·49 × 109/l, there was bleeding of any WHO grade on 42% (95% CI 31–55%) of study days within 1 d of blood sampling. The AIPN showed a statistically significant relationship to the proportion of days with bleeding, with lower AIPNs leading to more frequent bleeding (P < 0·001; Fig 3A). For TPC in the range 1–10 × 109/l, there was bleeding of any WHO grade on 34% (95% CI 20–47%) of study days within 1 d of blood sampling. There was a statistically significant trend between the proportion of days with bleeding and TPC (P = 0·015; Fig 3B). A similar trend was seen for WHO grade 2 or above bleeding (AIPN OR 2·29, 95% CI 1·29–4·07; P = 0·005: TPC OR 1·35, 95% CI 1·01–1·79; P = 0·040).

image

Figure 3. Proportions of the study days after blood sampling in which there was bleeding of any WHO grade grouped by platelet parameter result. (A) Absolute immature platelet number (AIPN) showing a statistically significant trend between increasing proportions of days with bleeding and lower AIPN (P < 0·001). (B) Total platelet count (TPC) showing a statistically significant trend between increasing proportions of days with bleeding and lower TPC (P = 0·015) (Created using data from all 754 study days, when both TPC and bleeding assessment data known). Data were adjusted for repeated measures and are expressed as proportions with 95% confidence intervals.

Download figure to PowerPoint

Thromboelastometry and plasma coagulation test results

There were significant associations between bleeding (any and ≥2 WHO grade) within 1 d of blood sampling and the EXTEM MCF (WHO grade 2 OR 0·93 95% CI 0·88–0·97; P = 0·003) and INTEM MCF (WHO grade 2 OR 0·92 95% CI 0·87–0·98; P = 0·007), but not with the other ROTEM test results (Table 3). The EXTEM MCF and INTEM MCF were also significantly associated with bleeding (WHO grade ≥2) within 2 d of blood sampling (Table SIV).

Table 3. Association between ROTEM parameters and plasma coagulation test results and WHO grade 2 or above bleeding with 1 d of blood sampling.
 WHO Grade ≥2 bleeding
 Unadjusted odds ratioOdds ratio adjusted for TPC
 Odds ratio95% CIP-valueOdds ratio95% CIP-value
  1. The associations between ROTEM parameters and bleeding are presented as unadjusted odds ratios and as odds ratios adjusted for TPC. The plasma coagulation test (PT, aPTT and Clauss fibrinogen) results are presented as odds ratios unadjusted for TPC. All odds ratios account for repeated measures. EXTEM, thromboplastin-initiated coagulation; INTEM, contact factor-initiated coagulation; FIBTEM, thromboplastin-initiated coagulation with the platelet inhibitor cytochalasin D; CT, clot time; CFT, clot formation time; MCF, maximum clot firmness; ML, maximum lysis; PT, prothrombin time; aPTT, activated partial thromboplastin time; 95% CI, 95% confidence interval. * = P < 0·01. †All ROTEM, PT, aPTT and fibrinogen results are presented as the odds ratio for each integer rise in the specific parameter. The haematocrit odds ratio is for every 5% rise in the haematocrit. Full results are available in Table SI.

EXTEM MCF0·930·88 to 0·970·003*0·950·89 to 1·000·066
EXTEM ML0·790·62 to 1·020·0690·850·65 to 1·110·234
INTEM MCF0·930·87 to 0·980·007*0·940·87 to 1·010·090
INTEM ML0·690·47 to 1·010·0580·750·51 to 1·100·143
FIBTEM MCF0·980·91 to 1·070·7050·990·91 to 1·080·861
FIBTEM ML0·870·71 to 1·070·1870·870·69 to 1·080·209
PT0·960·77 to 1·200·706   
aPTT1·000·91 to 1·100·995   
Fibrinogen1·050·82 to 1·360·689   

As the ROTEM thromboelastometry MCF is partly dependent on TPC, odds ratios were also calculated for the association between bleeding and MCF after adjustment for TPC. The adjusted odds ratios indicated that there was no longer a statistically significant association between either the EXTEM MCF or the INTEM MCF and bleeding (any and ≥2 WHO grade) within 1 d or within 2 d of blood sampling (Table 3 and Table SIV). There were no significant associations between PT, aPTT, Clauss fibrinogen and bleeding (any and ≥2 WHO grade) within 1 d or within 2 d of blood sampling (Table 3 and Table SIV).

Prediction of bleeding using TPC and AIPN

Given that AIPN was the only platelet parameter or test result that was associated with bleeding in the study cohort independently of TPC, we plotted ROC curves to compare the sensitivity and specificity of all possible cut-off threshold values for TPC and AIPN for the prediction of bleeding within 1 d of blood sampling. The AIPN had higher values for AUC for bleeding of WHO grade ≥2 (AIPN AUC 0·75; 95% CI 0·66–0·84 versus TPC AUC 0·70; 95% CI 0·59–0·82, Fig 4) but the AUC were very similar for bleeding of any WHO grade (AIPN AUC 0·66; 95% CI 0·59–0·73 versus TPC AUC 0·66; 95% CI 0·59–0·74).

image

Figure 4. Receiver operator characteristic curves for thresholds of absolute immature platelet number and total platelet count for prediction of bleeding of World Health Organization grade 2 or above in the day following blood sampling. Receiver operator characteristic (ROC) curve for total total platelet count (TPC) created using data from all 754 study days, when both TPC and bleeding assessment data were known. AUC = area under curve.

Download figure to PowerPoint

The sensitivity, specificity, PPV and NPV of different TPC and AIPN thresholds for predicting bleeding are shown in Table SV. The TPC had poor PPV for bleeding (33·7% and 8·7% for bleeding of any and ≥2 WHO grade, respectively) with a TPC threshold of 10 × 109/l, which is the threshold usually adopted to direct prophylactic platelet transfusion. The PPV of AIPN at a threshold of 0·5 × 109/l was higher (42·3% and 17·3% for any and ≥2 WHO grade respectively) than the TPC threshold of 10 × 109/l. The NPV of the TPC threshold of 10 × 109/l (78·9% and 95·4% for bleeding of any and ≥2 WHO grade respectively) was similar to that of the AIPN threshold of 0·5 × 109/l (77·7% and 95·5% for bleeding of any and ≥2 WHO grade respectively). AIPN thresholds of ≥2 x 109/l had 100% NPV for bleeding of WHO grade ≥2.

Discussion

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information

To our knowledge this is the first study that has prospectively evaluated the associations between platelet parameters other than the TPC and other laboratory markers of haemostasis and bleeding the following day in thrombocytopenic patients undergoing treatment for haematological malignancy. It is a strength of our study that we enrolled participants with a wide range of underlying haematological diagnoses and treatments and that we collected data irrespective of whether patients experienced complications of therapy, such as sepsis, mucositis or graft-versus-host disease which may potentially alter bleeding risk. Thus, our findings are likely to be applicable to patients with a broad range of haematological malignancies.

We have focussed on platelet parameters and other markers of haemostasis that can be detected using rapid, robust and inexpensive laboratory tests. To this end, we evaluated the platelet parameters MPV, IPF and AIPN, which are determined by automated haematology analysers at the same time as the TPC in the full blood count test. We also evaluated platelet function determined using the point-of-care format PFA-100® device (Hayward et al, 2006) and coagulation factor and fibrinogen activity using the widely available PT, aPTT and Clauss fibrinogen tests. Finally, we used the point-of-care format ROTEM thromboelastometry device to evaluate coagulation factor, fibrinogen and platelet function as well as severe fibrinolysis (Luddington, 2005). ROTEM and similar technologies are used successfully in clinical settings, such as cardiac surgery and liver surgery, to detect coagulopathy, direct appropriate treatments to prevent bleeding and to improve clinical outcomes (Weber et al, 2012; Álamo et al, 2013).

Our study has several important findings relevant to the prediction of bleeding in patients with thrombocytopenia during treatment for haematological malignancy. The first is that measurement of the platelet parameter MPV and platelet function by PFA-100® closure time is not feasible in this setting because these tests were not completed in a high proportion of study blood samples. PFA-100® closure times are highly dependent on TPC and may not reach endpoint in severely thrombocytopenic samples (Carcao et al, 2002; Hayward et al, 2006). We have confimed that this effect was not substantially improved by the use of the small aperture INNOVANCE cartridge. Similarly, MPV determined by automated haematology analysers with impedance endpoints requires analysis of large numbers of platelets in each blood sample, which may not be achieved in thrombocytopenic samples using standard instrument protocols (Latger-Cannard et al, 2012). Although our data do not exclude a role for the MPV and PFA-100 closure times in predicting bleeding, the inability to reliably detect these markers in our study precludes their use in this clinical setting.

For the remaining laboratory tests, in which a very high level of completion was achieved, we confirmed a significant association between low TPC and increased bleeding within 1 d, and within 2 d of blood sampling. The study's power to detect a difference for the unadjusted TPC was strengthened by the use of data from all 754 study days, compared to only 277 d when adjusted for AIPN. One potential explanation for the poor association between bleeding and TPC in our study is that current clinical practice requires patients with TPC ≤ 10 × 109/l to receive a platelet transfusion on the day of the blood sample. Therefore, the risks of bleeding because of thrombocytopenia after a blood sample with TPC ≤ 10 × 109/l is likely to be lower than expected just by virtue of this intervention. Two recent randomized controlled trials have compared a therapeutic-only versus prophylactic platelet transfusion policy (Stanworth et al, 2013; Wandt et al, 2012). Both studies showed that prophylactic platelet transfusions reduced bleeding rates overall, but the size of the effect differed between the two studies.

We also demonstrated a significant association between low ROTEM MCF and clinically significant bleeding of WHO grade ≥2. The ROTEM MCF is a composite marker of the haemostatic function of both platelets and fibrinogen (Lang et al, 2009), and low ROTEM MCF values caused by hypofibrinogenaemia are associated with increased bleeding in settings such as cardiac surgery (Solomon et al, 2011). This is unlikely to hold true for low ROTEM MCF in haematological malignancy because there was no significant association between bleeding and the Clauss fibrinogen test result or the ROTEM FIBTEM MCF, which also reflects functional fibrinogen (Solomon et al, 2011). The significant association between ROTEM MCF and bleeding was also no longer observed after adjustment for TPC. This suggests that low MCF in this setting is a direct effect of thrombocytopenia and that this marker has no additional value to the TPC. Our finding of no association between bleeding and the ROTEM CT, ROTEM CFT, PT and aPTT test results suggests that coagulation factor defects do not contribute significantly to bleeding in haematological malignancy. Our finding of no association between the ROTEM ML test result and bleeding also suggests that severe defects in fibrinolysis are unlikely to contribute to bleeding. However, the ROTEM ML is insensitive to less severe fibrinolysis defects (Raza et al, 2013) so this coagulopathy cannot be ruled out from our data.

In contrast to the other laboratory markers of haemostasis evaluated in this study, low AIPN was associated strongly with increased bleeding both within 1 d and 2 d of blood sampling. There was also an approximately linear relationship between AIPN and the proportion of study days on which bleeding occurred in the study participants. We found that the association between AIPN and bleeding persisted after adjustment for TPC within 1 d of blood sampling. However, the association between TPC and bleeding lost statistical significance when adjusted for variation in AIPN, suggesting that the AIPN has a more direct association with bleeding.

Both the AIPN and IPF platelet parameters are determined by the Sysmex XE-2100 haematology analyser as part of the full blood count by detecting reticulated platelets, which show increased staining with a fluorescent RNA dye (Briggs et al, 2004). Reticulated platelets are an immature platelet population that are newly synthesized in the bone marrow. Increasing circulating reticulated platelets is an early marker of thrombopoietic regeneration after chemotherapy or haemopoietic stem cell transplantation that precedes increases in the TPC (Briggs et al, 2006; Have et al, 2013). The strong association between low AIPN and increased bleeding in our study participants suggests that endogenous immature platelets are an important determinant of haemostasis during treatment for haematological malignancies. This is consistent with previous reports that immature platelets have greater haemostatic function than mature or transfused platelets (Guthikonda et al, 2007; Fager et al, 2010). The IPF also provides a measure of circulating immature platelets, although expressed as a proportion of the total circulating platelet count rather than an absolute concentration. Consistent with this, transfusion of donor platelets significantly reduced the IPF in patients with haematological malignancy, but not the AIPN, at least in part because of a dilution effect from mature platelets in the donor units (Bat et al, 2013). Given that platelet transfusion was common in our thrombocytopenic study population, it is unsurprising that the circulating immature platelet concentration expressed by the IPF failed to show a consistent association with bleeding.

The potential clinical utility of measuring the AIPN in thrombocytopenic patients with haematological malignancy was highlighted by our predictive model. This showed that for prediction of bleeding, the area under the ROC curve for AIPN was greater than for TPC. Moreover, the PPV of an AIPN threshold of 0·5 × 109/l for bleeding was substantially higher than that of the TPC threshold of 10 × 109/l, which is currently used to direct prophylactic platelet transfusion, with similar NPV. If confirmed in larger scale studies, these findings suggest that decisions to administer prophylactic platelets based on AIPN and not TPC may improve the appropriateness of platelet transfusion in this group. This has the potential to reduce adverse bleeding events and to avoid unnecessary exposure to platelet components in patients with haematological malignancy.

Acknowledgements

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information

We acknowledge H. Thomas and B. Kahan for statistical advice; S. Howgate, A. Raja, J. Bailey, M. Walker and C. Reilly-Stitt for assistance with laboratory analyses; L. Fraser, K. Harding, C. O'Donovan and J. Proctor for recruitment and assessment of the study participants; C. Dyer for study administration, and the haematology units at Oxford University Hospitals NHS Trust and University Hospitals Bristol NHS Foundation Trust led by Dr Tim Littlewood and Prof D. Marks respectively.

Author Contributions

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information

LE designed and executed the study, performed data analysis and wrote the manuscript. PH, AM and MM supervized laboratory analysis and data interpretation. GP performed bleeding assessments and coordinated data collection. SS and MM supervized study design and data interpretation. All authors contributed to the final manuscript.

Funding

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information

The study was funded through NHS Blood and Transplant Trust Fund (TR037) and British Society of Haematology Start-up Grant. This research was also supported by the National Institute for Health Research (NIHR) Bristol Biomedical Research Unit in Cardiovascular Disease at the University Hospitals Bristol NHS Foundation Trust and the University of Bristol. This article presents independent research from the NIHR. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.

References

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information
  • Álamo, J.M., León, A., Mellado, P., Bernal, C., Marín, L.M., Cepeda, C., Suárez, G., Serrano, J., Padillo, J. & Gómez, M.Á. (2013) Is “intra–operating room” thromboelastometry useful in liver transplantation? a case-control study in 303 patients. Transplantation Proceedings, 45, 36373639.
  • Barsam, S.J., Psaila, B., Forestier, M., Page, L.K., Sloane, P.A., Geyer, J.T., Villarica, G.O., Ruisi, M.M., Gernsheimer, T.B., Beer, J.H. & Bussel, J.B. (2011) Platelet production and platelet destruction: assessing mechanisms of treatment effect in immune thrombocytopenia. Blood, 117, 57235732.
  • Bat, T., Leitman, S.F., Calvo, K.R., Chauvet, D. & Dunbar, C.E. (2013) Measurement of the absolute immature platelet number reflects marrow production and is not impacted by platelet transfusion. Transfusion, 53, 12011204.
  • BCSH (2003) British Committee for Standards in Haematology: guidelines for the use of platelet transfusions. British Journal of Haematology, 122, 1023.
  • Blumberg, N., Spinelli, S.L., Francis, C.W., Taubman, M.B. & Phipps, R.P. (2009) The platelet as an immune cell-CD40 ligand and transfusion immunomodulation. Immunology Research, 45, 251260.
  • Bolton-Maggs, P.H.B. (Ed), Poles, D., Watt, A., Thomas, D. & Cohen, H.. on behalf of the Serious Hazards of Transfusion (SHOT) Steering Group (2013) The 2012 Annual SHOT Report, p. 200. © Serious Hazards of Transfusion (SHOT) 2012. ISBN 978-0-9668648-5-8.
  • Briggs, C., Kunka, S., Hart, D., Oguni, S. & Machin, S.J. (2004) Assessment of an immature platelet fraction (IPF) in peripheral thrombocytopenia. British Journal of Haematology, 126, 9399.
  • Briggs, C., Hart, D., Kunka, S., Oguni, S. & Machin, S.J. (2006) Immature platelet fraction measurement: a future guide to platelet transfusion requirement after haematopoietic stem cell transplantation. Transfusion Medicine, 16, 101109.
  • Burnett, A.K., Hills, R.K., Milligan, D., Kjeldsen, L., Kell, J., Russell, N.H., Yin, J.A.L., Hunter, A., Goldstone, A.H. & Wheatley, K. (2011) Identification of patients with acute myeloblastic leukemia who benefit from the addition of Gemtuzumab Ozogamicin: results of the MRC AML15 trial. Journal of Clinical Oncology, 29, 369377.
  • Carcao, M.D., Blanchette, V.S., Stephens, D., He, L., Wakefield, C.D., Butchart, S., Christie, D.J. & Rand, M.L. (2002) Assessment of thrombocytopenic disorders using the Platelet Function Analyzer (PFA-100®). British Journal of Haematology, 117, 961964.
  • Estcourt, L.J., Birchall, J., Lowe, D., Grant-Casey, J., Rowley, M. & Murphy, M.F. (2012) Platelet transfusions in haematology patients: are we using them appropriately? Vox Sanguinis, 103, 284293.
  • Fager, A.M., Wood, J.P., Bouchard, B.A., Feng, P. & Tracy, P.B. (2010) Properties of procoagulant platelets: defining and characterizing the subpopulation binding a functional prothrombinase. Arteriosclerosis, Thrombosis, and Vascular Biology, 30, 24002407.
  • Fielding, A.K., Richards, S.M., Chopra, R., Lazarus, H.M., Litzow, M.R., Buck, G., Durrant, I.J., Luger, S.M., Marks, D.I., Franklin, I.M., McMillan, A.K., Tallman, M.S., Rowe, J.M. & Goldstone, A.H. (2007) Outcome of 609 adults after relapse of acute lymphoblastic leukemia (ALL); an MRC UKALL12/ECOG 2993 study. Blood, 109, 944950.
  • Ganter, M.T. & Hofer, C.K. (2008) Coagulation monitoring: current techniques and clinical use of viscoelastic point-of-care coagulation devices. Anesthesia and Analgesia, 106, 13661375.
  • Greeno, E., McCullough, J. & Weisdorf, D. (2007) Platelet utilization and the transfusion trigger: a prospective analysis. Transfusion, 47, 201205.
  • Guthikonda, S., Lev, E.I., Patel, R., Delao, T., Bergeron, A.L., Dong, J.F. & Kleiman, N.S. (2007) Reticulated platelets and uninhibited COX-1 and COX-2 decrease the antiplatelet effects of aspirin. Journal of Thrombosis and Haemostasis, 5, 490496.
  • Have, L.W., Hasle, H., Vestergaard, E.M. & Kjaersgaard, M. (2013) Absolute immature platelet count may predict imminent platelet recovery in thrombocytopenic children following chemotherapy. Pediatric Blood & Cancer, 60, 11981203.
  • Hayward, C.P.M., Harrison, P., Cattaneo, M., Ortel, T.L. & Rao, A.K. & The Platelet Physiology Subcommittee Of The Scientific Standardization Committee Of The International Society On Thrombosis Haemostasis (2006) Platelet function analyzer (PFA)-100® closure time in the evaluation of platelet disorders and platelet function. Journal of Thrombosis and Haemostasis, 4, 312319.
  • Heddle, N. & Webert, K.. (2009) Investigation of acute transfusion reactions. In: Practical Transfusion Medicine (ed. by M. Murphy & D. Pamphilion), pp. 6389. Blackwell, Oxford.
  • Heddle, N., Cook, R., Tinmouth, A., Kouroukis, C., Hervig, T., Klapper, E., JM, B., Szczepiorkowski, Z., AuBuchon, J., Barty, R. & Lee, K. (2009) A randomized controlled trial comparing standard and low dose strategies for transfusion of platelets (SToP) to patients with thrombocytopenia. Blood, 113, 15641573.
  • Kaito, K., Otsubo, H., Usui, N., Yoshida, M., Tanno, J., Kurihara, E., Matsumoto, K., Hirata, R., Domitsu, K. & Kobayashi, M. (2005) Platelet size deviation width, platelet large cell ratio, and mean platelet volume have sufficient sensitivity and specificity in the diagnosis of immune thrombocytopenia. British Journal of Haematology, 128, 698702.
  • Lang, T., Johanning, K., Metzler, H., Piepenbrock, S., Solomon, C., Rahe-Meyer, N. & Tanaka, K. (2009) The effects of fibrinogen levels on thromboelastometric variables in the presence of thrombocytopenia. Anesthesia and Analgesia, 108, 751758.
  • Latger-Cannard, V., Hoarau, M., Salignac, S., Baumgart, D., Nurden, P. & Lecompte, T. (2012) Mean platelet volume: comparison of three analysers towards standardization of platelet morphological phenotype. International Journal of Laboratory Hematology, 34, 300310.
  • Luddington, R.J. (2005) Thrombelastography/thromboelastometry. Clinical & Laboratory Haematology, 27, 8190.
  • Milligan, D.W., Wheatley, K., Littlewood, T., Craig, J.I.O. & Burnett, A.K. & for the NCRI Haematological Oncology Clinical Studies Group (2006) Fludarabine and cytosine are less effective than standard ADE chemotherapy in high-risk acute myeloid leukemia, and addition of G-CSF and ATRA are not beneficial: results of the MRC AML-HR randomized trial. Blood, 107, 46144622.
  • National Blood Authority (2012) The national blood authority's patient blood management guideline: module 3 - medical. In: The National Blood Authority's Patient Blood Management Guideline (ed. by National Blood Authority), National Blood Authority, Canberra, Australia.
  • Pendry, K. & Davies, T. (2011) An audit of the use and wastage in the North West of England and North Wales - where have all the platelets gone? Blood and Transfusion Matters, 34, 1719.
  • Popovsky, M. & Moore, S. (1985) Diagnostic and pathogenetic considerations in transfusion-related acute lung injury. Transfusion, 25, 573577.
  • Raza, I., Davenport, R., Rourke, C., Platton, S., Manson, J., Spoors, C., Khan, S., De'Ath, H., Allard, S., Hart, D., John Pasi, K., Hunt, B.J., Stanworth, S., MacCallum, P. & Brohi, K. (2013) The incidence and magnitude of fibrinolytic activation in trauma patients. Journal of Thrombosis and Haemostasis, 11, 307314.
  • Schiffer, C.A., Anderson, K.C., Bennett, C.L., Bernstein, S., Elting, L.S., Goldsmith, M., Goldstein, M., Hume, H., McCullough, J.J., McIntyre, R.E., Powell, B.L., Rainey, J.M., Rowley, S.D., Rebulla, P., Troner, M.B. & Wagnon, A.H. (2001) Platelet transfusion for patients with cancer: clinical practice guidelines of the American Society of Clinical Oncology. Journal of Clinical Oncology, 19, 15191538.
  • Slichter, S.J., Kaufman, R.M., Assmann, S.F., McCullough, J., Triulzi, D.J., Strauss, R.G., Gernsheimer, T.B., Ness, P.M., Brecher, M.E., Josephson, C.D., Konkle, B.A., Woodson, R.D., Ortel, T.L., Hillyer, C.D., Skerrett, D.L., McCrae, K.R., Sloan, S.R., Uhl, L., George, J.N., Aquino, V.M., Manno, C.S., McFarland, J.G., Hess, J.R., Leissinger, C. & Granger, S. (2010) Dose of prophylactic platelet transfusions and prevention of hemorrhage. New England Journal of Medicine, 362, 600613.
  • Solomon, C., Cadamuro, J., Ziegler, B., Schöchl, H., Varvenne, M., Sørensen, B., Hochleitner, G. & Rahe-Meyer, N. (2011) A comparison of fibrinogen measurement methods with fibrin clot elasticity assessed by thromboelastometry, before and after administration of fibrinogen concentrate in cardiac surgery patients. Transfusion, 51, 16951706.
  • Stanworth, S.J., Dyer, C., Choo, L., Bakrania, L., Copplestone, A., Llewelyn, C., Norfolk, D., Powter, G., Littlewood, T., Wood, E.M. & Murphy, M.F. (2010) Do all patients with hematologic malignancies and severe thrombocytopenia need prophylactic platelet transfusions? Background, rationale, and design of a clinical trial (trial of platelet prophylaxis) to assess the effectiveness of prophylactic platelet transfusions. Transfusion Medicine Reviews, 24, 163171.
  • Stanworth, S.J., Estcourt, L.J., Powter, G., Kahan, B., Dyer, C., Choo, L., Bakrania, L., Llewelyn, C., Littlewood, T., Soutar, R., Norfolk, D., Copplestone, A., Smith, N., Kerr, P., Jones, G., Raj, K., Westerman, D., Szer, J., Jackson, N., Bardy, P., Plews, D., Lyons, S., Bielby, L., Wood, E.M. & Murphy, M. (2013) A no-prophylaxis platelet transfusion strategy for hematologic cancers. New England Journal of Medicine, 368, 17711780.
  • Taylor, C.E., Cohen, H., Mold, D., Jones, H., Davies, T., Mistry, H., Ball, J., Asher, D., Cawley, C., Chaffe, B., Chapman, C., Gray, A., Jones, J., Milkins, C., New, H., Norfolk, D., Regan, F., Still, E. & Tinegate, H. on behalf of the Serious Hazards of Transfusion (SHOT) Steering Group (2010) The 2009 annual SHOT report. p. 153. © Serious Hazards of Transfusion (SHOT) 2009. ISBN 978-0-9558648-2-7.
  • Wandt, H., Schaefer-Eckart, K., Wendelin, K., Pilz, B., Wilhelm, M., Thalheimer, M., Mahlknecht, U., Ho, A., Schaich, M., Kramer, M., Kaufmann, M., Leimer, L., Schwerdtfeger, R., Conradi, R., Dolken, G., Klenner, A., Hanel, M., Herbst, R., Junghanss, C. & Ehninger, G. (2012) Therapeutic platelet transfusion versus routine prophylactic transfusion in patients with haematological malignancies: an open-label, multicentre, randomised study. Lancet, 380, 13091316.
  • Weber, C.F., Gorlinger, K., Meininger, D., Herrmann, E., Bingold, T., Moritz, A., Cohn, L.H. & Zacharowski, K. (2012) Point-of-care testing: a prospective, randomized clinical trial of efficacy in Coagulopathic cardiac surgery patients. Anesthesiology, 117, 531547.
  • Whitaker, B.I. & Hinkins, S. (2013) The 2011 National Blood Collection and Utilization Survey Report. (ed. by the US Department of Health and Human Services). US Department of Health and Human Services, Rockville, Maryland.
  • Whitaker, B.I., Schlumpf, K., Schulman, J. & Green, J.. (2011) Report of the US Department of Health and Human Services. The 2009 national blood collection and utilization survey report. US Department of Health and Human Services, Office of the Assistant Secretary for Health, Washington, DC.
  • Williamson, L.M. & Devine, D.V. (2013) Challenges in the management of the blood supply. The Lancet, 381, 18661875.

Supporting Information

  1. Top of page
  2. Summary
  3. Methods and materials
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Author Contributions
  8. Conflicts of interest
  9. Funding
  10. References
  11. Supporting Information
FilenameFormatSizeDescription
bjh12928-sup-0001-TableS1-S5.docxWord document32K

Table SI. Sites of bleeding. Patients may have had more than one site of bleeding on any 1 d.

Table SII. Presence of bleeding after administration of platelet transfusions, categorised in two reasons for administration of platelet transfusion.

Table SIII. Associations between platelet parameters and bleeding within 2 d of blood sampling.

Table SIV. Association between ROTEM parameters and plasma coagulation test results and bleeding with 2 d of blood sampling.

Table SV. Predictive value of total platelet count (TPC) and absolute immature platelet number (AIPN) for any bleeding within 1 d of blood sampling.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.