Characteristics of patients
Between 2011 and 2016, 259 patients required platelet transfusion on ≥ 2 occasions and were included in the study, collectively receiving 1357 platelet transfusions. The characteristics of the study patients are summarized in Table 1. The median age was 58 years (Q1, Q3: 43.5, 70), most (57.5%) patients were males and 84.9% were admitted for a medical reason. About a third (35.5%) had cancer, almost two thirds (63.3%) were in shock on ICU admission, 59.8% were septic and 20.1% had a form of bleeding.
Table 1: General characteristics of patients by platelet transfusion refractoriness status
|
Total (N=259)
|
Yes (N=142)
|
No (N=117)
|
P value
|
Age (years)
|
Median (Q1, Q3)
|
58.0
(43.5, 70.0)
|
57.0
(42.2, 68.0)
|
60.0
(46.0, 72.0)
|
0.11
|
Male Sex
|
N (%)
|
149 (57.5%)
|
81 (57.0%)
|
68 (58.1%)
|
0.86
|
Body Mass Index (kg/m2)
|
Median (Q1, Q3)
|
27.3
(22.4, 32.3)
|
27.9
(22.3, 32.9)
|
26.6
(22.6, 31.4)
|
0.40
|
Obese > 30 kg/m2
|
N (%)
|
94 (37.5%)
|
54 (39.4%)
|
40 (35.1%)
|
0.60
|
Body Surface Area (m2)
|
Median (Q1, Q3)
|
1.8
(1.6, 1.9)
|
1.8
(1.7, 1.9)
|
1.8
(1.6, 1.9)
|
0.42
|
Admission Category
|
N (%)
|
|
|
|
0.03
|
Medical
|
|
220 (84.9%)
|
125 (88.0%)
|
95 (81.2%)
|
|
Surgical
|
|
30 (11.6%)
|
16 (11.3%)
|
14 (12.0%)
|
|
Trauma
|
|
9 (3.5%)
|
1 (0.7%)
|
8 (6.8%)
|
|
Chronic Cardiac Disease /238 patients⁑
|
N (%)
|
35 (15.2%)
|
18 (13.6%)
|
17 (17.3%)
|
0.44
|
Chronic Immune Disease /238 patients⁑
|
N (%)
|
89 (38.7%)
|
47 (35.6%)
|
42 (42.9%)
|
0.26
|
Chronic Liver Disease /238 patients⁑
|
N (%)
|
48 (20.9%)
|
35 (26.5%)
|
13 (13.3%)
|
0.01
|
Chronic Respiratory Disease /238 patients⁑
|
N (%)
|
17 (7.4%)
|
12 (9.1%)
|
5 (5.1%)
|
0.25
|
Chronic Renal Disease /238 patients*
|
N (%)
|
29 (12.6%)
|
14 (10.6%)
|
15 (15.3%)
|
0.29
|
APACHE II Score
|
Median (Q1, Q3)
|
25.0
(20.0, 30.0)
|
25.0
(21.0, 30.0)
|
24.0
(20.0, 28.5)
|
0.40
|
Glasgow Coma Scale on ICU admission
|
Median (Q1, Q3)
|
14.0
(9.5, 15.0)
|
14.0
(9.0, 15.0)
|
14.0
(10.0, 15.0)
|
0.97
|
Mechanical Ventilation /238 patients*
|
N (%)
|
191 (83.0%)
|
110 (83.3%)
|
81 (82.7%)
|
0.89
|
PaO2/FiO2 Ratio
|
Median (Q1, Q3)
|
178.5
(113.2, 270.5)
|
176.0
(112.2, 258.0)
|
182.0
(114.0, 282.5)
|
0.38
|
Shock
|
N (%)
|
164 (63.3%)
|
91 (64.1%)
|
73 (62.4%)
|
0.78
|
Sepsis
Septic shock
|
N (%)
N (%)
|
155 (59.8%)
149 (57.5%)
|
88 (62.0%)
84 (59.2%)
|
67 (57.3%)
65 (55.6%)
|
0.44
0.56
|
|
|
|
|
|
|
DIC likelihood score
|
Median (Q1, Q3)
|
5.0
(4.0, 6.0)
|
5.0
(4.0, 6.0)
|
5.0
(4.0, 5.0)
|
0.04
|
Active bleeding
|
N (%)
|
52 (20.1%)
|
27 (19.0%)
|
25 (21.4%)
|
0.64
|
Active cancer
|
N (%)
|
92 (35.5%)
|
47 (33.1%)
|
45 (38.5%)
|
0.37
|
Hematological Cancer
|
N (%)
|
75 (29.0%)
|
38 (26.8%)
|
37 (31.6%)
|
0.39
|
Organ transplant
|
N (%)
|
13 (5.0%)
|
9 (6.3%)
|
4 (3.4%)
|
0.28
|
Splenomegaly /182 patients*
|
N (%)
|
65 (36.5%)
|
40 (43.5%)
|
25 (29.1%)
|
0.046
|
Platelet count on ICU admission (x109/L)
|
Median (Q1, Q3)
|
53.5
(28.0, 98.8)
|
49.0
(26.0, 93.0)
|
59.0
(35.0, 108.0)
|
0.11
|
Thrombocytopenia on admission ( < 150 x109/L)
|
N (%)
|
226 (87.6%)
|
123 (87.2%)
|
103 (88.0%)
|
0.85
|
Severe thrombocytopenia on admission (< 50 x109/L)
|
N (%)
|
118 (45.7%)
|
72 (51.1%)
|
46 (39.3%)
|
0.06
|
Lowest platelet count during ICU Stay (x109/L)
|
Median (Q1, Q3)
|
18.0
(10.0, 29.0)
|
14.0
(7.0, 28.0)
|
19.0
(11.0, 34.0)
|
0.001
|
INR
|
Median (Q1, Q3)
|
1.6
(1.3, 2.0)
|
1.6
(1.3, 2.2)
|
1.6
(1.2, 2.0)
|
0.38
|
Lactic acid (mmol/L)
|
Median (Q1, Q3)
|
2.9
(1.7, 6.0)
|
2.9
(1.5, 5.9)
|
3.0
(1.8, 6.1)
|
0.47
|
Creatinine (μmol/L)
|
Median (Q1, Q3)
|
142.5
(73.8, 256.2)
|
145.0
(82.2, 260.2)
|
134.5
(65.8, 232.0)
|
0.33
|
Bilirubin (μmol/L)
|
Median (Q1, Q3)
|
40.0
(19.0, 107.2)
|
40.0
(19.0, 106.0)
|
40.0
(19.5, 107.5)
|
0.99
|
Transfusion count
|
Median (Q1, Q3)
|
4.0
(2.0, 6.0)
|
5.0
(3.0, 8.0)
|
3.0
(2.0, 4.0)
|
< 0.001
|
Platelet transfusion refractory (CCI definition)
|
N (%)
|
146 (57.0%)
|
139 (98.6%)
|
7 (6.1%)
|
< 0.001
|
- Transfusion Characteristics
|
Total (N=1357)
|
No (N=431)
|
Yes (N=926)
|
P value
|
Platelet Count Prior to Transfusion (x109/L)
|
Median (Q1, Q3)
|
31.0
(17.0, 52.0)
|
33.0
(18.0, 54.0)
|
31.0
(16.0, 52.0)
|
0.17
|
Transfusion Dose (Units)
|
Median (Q1, Q3)
|
6.0
(6.0, 12.0)
|
6.0
(6.0, 6.0)
|
6.0
(6.0, 12.0)
|
< 0.001
|
Platelet Increment (x109/L)
|
Median (Q1, Q3)
|
6.0
(-5.0, 24.0)
|
21.0
(9.0, 42.0)
|
2.0
(-8.0, 14.0)
|
< 0.001
|
Corrected Count Increment (x109/L)
|
Median (Q1, Q3)
|
2.8
(-2.0, 10.8)
|
10.3
(3.8, 18.8)
|
0.6
(-3.7, 5.8)
|
< 0.001
|
Platelet product
Apheresis
Apheresis-Irradiated
Pooled
Pooled-Irradiated
Mixed**
|
N (%)
|
138 (10.2%)
196 (14.4%)
750 (55.3%)
182 (13.4%)
91 (6.7%)
|
40 (9.3%)
72 (16.7%)
242 (56.1%)
59 (13.7%)
18 (4.2%)
|
98 (10.6%)
124 (13.4%)
508 (54.9%)
123 (13.3%)
73 (7.9%)
|
0.06
|
* For variables with missing data, the number of valid observations is reported.
** Mixed” indicates aggregate transfusions that were derived from transfusions with 2 or more different platelet products.
APACHE: Acute Physiologic Assessment and Chronic Health Evaluation; DIC: disseminated intravascular coagulation; ICU: intensive care unit; INR: international normalized ratio, Q1: first quartile, Q3: third quartile
The majority (87.6%) of the patients had thrombocytopenia (platelet count < 150⨉109/L) on admission with count < 50⨉109/L present in 45.7%. The rest (12.4%) developed thrombocytopenia during their ICU stay. The median platelet count on admission was 53.5 ⨉109/L (Q1, Q3: 28, 98.8). The median platelet nadir was 18 ⨉109/L (Q1, Q3: 10, 29). Each patient received platelet transfusion on a median of 4 (Q1, Q3: 2, 6) occasions. The median dose of the platelet transfusions was 6 pooled units (Q1, Q3: 6, 12) or their equivalent. The majority of the platelets transfused were pooled un-irradiated (55.3%), while pooled-irradiated platelets accounted for 13.4%, apheresis platelets accounted for 10.2%, and apheresis-irradiated platelets accounted for 14.4%.
The median time interval to next transfusion was 1 day (1, 2). The post-transfusion platelet count was taken a median of 5.0 hours (Q1, Q3: 2.94, 7.58; range: 0, 40 hours) after transfusion. There was no significant difference in PI across the different intervals to next transfusion as described in Figure 1.
Prevalence of platelet transfusion refractoriness
More than half of the patients (54.8%) had platelet transfusion refractoriness by PI and 57.0% by CCI increment. The median PI was 6⨉109/L (Q1, Q3: -5, 24), and the median CCI was 2.8 (Q1, Q3: -2.0, 10.8). There was strong agreement between refractoriness diagnosis by PI and CCI (kappa coefficient: 0.929, 95% CI: 0.884-0.968; p<0.001). Assuming CCI increment is more accurate for diagnosing platelet transfusion refractoriness, diagnosis by PI had 95.2% sensitivity (95% CI: 90.4-98.1%) and 98.2% specificity (95% CI: 93.6-99.8%). Figure 2A describes the relationship between PI and CCI according to the platelet type.
Predictors of platelet transfusion refractoriness
As shown in Table 1, the admission category, splenomegaly, chronic liver disease, and higher number of transfusions were associated with platelet transfusion refractoriness on univariate analysis. Trauma patients were the least likely to develop refractoriness (11.1%) compared with medical (56.8%) and non-trauma surgical patients (53.3%). Particularly of note, the admitting services of Hepatology, Liver Transplant, and Hematology had the highest rates of refractoriness (69.7%, 69.2%, and 55.6%, respectively) (Figure 2B). Of patients with splenomegaly, 61.5% developed refractoriness compared to 46.0% of those with non-enlarged spleens (p=0.046). There was a modest, but statistically significant increase in DIC likelihood among refractory (median: 5, Q1, Q3: 4, 6) compared to non-refractory patients (median: 5, Q1, Q3: 4, 5) (p=0.04). On the other hand, no association of statistical significance could be discerned between refractoriness and presence of sepsis on admission, bleeding, hematological cancer, or shock.
The multivariable logistic regression model showed younger age to be the only statistically significant predictor of refractory status (OR: 0.975, 95% CI: 0.951-0.999). APACHE II score tended to be significant (OR: 1.064, 95% CI: 0.996-1.136; p=0.065).
Factors associate with the time to next platelet transfusion
Several factors predisposed to an earlier next transfusion (Table 2). These include the pre-transfusion platelet count (p<0.001), platelet dose (p=0.04), PI following transfusion (p<0.001), and the particular platelet product used (p< 0.001). Re-transfusion within 2 days was less frequent with non-irradiated pooled platelets (50.7%, compared to 63.6% for apheresis non-irradiated, 62.0% for apheresis irradiated, and 69.4% for pooled irradiated).
Table 2: Characteristics of platelet transfusions by the time to next transfusion
|
|
Total (N=1118)
|
Less than 2 days (N=643)
|
2 days or more (N=475)
|
P value
|
Transfusion Dose (Units)
|
Median (Q1, Q3)
|
6.0
(6.0, 12.0)
|
6.0
(6.0, 12.0)
|
6.0
(6.0, 6.0)
|
0.04
|
Platelet Increment (x109/L)
|
Median (Q1, Q3)
|
6.0
(-5.0, 24.0)
|
5.0
(-6.0, 19.0)
|
9.0
(-4.0, 29.0)
|
< 0.001
|
Platelet Count Prior to Transfusion (x109/L)
|
Median (Q1, Q3)
|
31.0
(16.0, 50.0)
|
28.0
(14.0, 45.0)
|
37.0
(19.0, 58.0)
|
< 0.001
|
Platelet product
|
N (%)
|
111 (9.9%)
163 (14.6%)
608 (54.4%)
157 (14.0%)
79 (7.1%)
|
71 (11.0%)
101 (15.7%)
310 (48.2%)
109 (17.0%)
52 (8.1%)
|
40 (8.4%)
62 (13.1%)
298 (62.7%)
48 (10.1%)
27 (5.7%)
|
< 0.001
|
Apheresis
Apheresis-Irradiated
Pooled
Pooled-Irradiated
Mixed⁑
|
* Mixed indicates aggregate transfusions that were derived from transfusions with 2 or more different platelet products.
Q1: first quartile, Q3: third quartile
Table 3 describes the outcomes of patients. The median ICU length of stay for all patients was 13 days (Q1, Q3: 7, 23), and the median hospital length of stay was 28 days (Q1, Q3: 16, 58). ICU and hospital mortality were both high, at 59.1% and 73%, respectively.
Table 3: Outcomes of patient by platelet refractory tatus
|
|
Total (N=259)
|
Yes (N=142)
|
No (N=117)
|
P value
|
ICU mortality
|
N (%)
|
153 (59.1%)
|
86 (60.6%)
|
67 (57.3%)
|
0.55
|
Hospital mortality
|
N (%)
|
189 (73.0%)
|
105 (73.9%)
|
84 (71.8%)
|
0.70
|
New tracheostomy /238 patients*
|
N (%)
|
25 (10.9%)
|
19 (14.4%)
|
6 (6.1%)
|
0.046
|
Duration of mechanical ventilation (days) /238 patients*
|
Median (Q1, Q3)
|
10.0
(6.0, 17.0)
|
10.0
(5.0, 15.8)
|
12.0
(6.0, 19.5)
|
0.15
|
ICU length of stay (days)
|
Median (Q1, Q3)
|
13.0
(7.0, 23.0)
|
16.0
(8.0, 26.0)
|
12.0
(6.0, 21.0)
|
0.02
|
Hospital length of stay (days)
|
Median (Q1, Q3)
|
28.0
(16.0, 58.0)
|
33.0
(16.0, 62.5)
|
27.0
(14.0, 47.0)
|
0.11
|
* For variables with missing data, the number of valid observations is reported.
ICU: intensive care unit, Q1: first quartile, Q3: third quartile