Abstract
Introduction: There is little data on the incidence and consequences of hospital-acquired (HA) platelet count drops and no consensus on how to define HA-thrombocytopenia. We evaluated the incidence of relative and absolute HA-drops in platelet count among medical patients (general medicine, cardiology, and intensive care unit) to determine their association with mortality, HA-venous thromboembolism (VTE), and HA-bleeding.
Methods: Data was abstracted from the electronic medical record at the University of Vermont Medical Center, a 540-bed tertiary care hospital in Burlington, VT for admissions between 2009-12. Exclusion criteria were age <18, pregnant, admitted to a non-medical service or to the oncology service, and platelet count <150 thousand (k) at admission. HA-platelet count drops were defined as listed in the table (absolute nadir, relative drop, absolute platelet count). We used logistic regression to evaluate the association of various definitions of platelet count drops with HA-VTE, HA-bleeding (based on the International Society of Thrombosis and Haemostasis definition), and in-hospital mortality. Models were adjusted for age, sex, service, admission platelet count, and for known risk factors for HA-VTE, HA-bleeding, and mortality (Table).
Results: Of 11,863 admissions without thrombocytopenia on admission, 1,905 (16.1%) patients developed a platelet count <150k, 6,971 (58.8%) had at least a 10% drop in their platelet count, and 6,737 (56.8%) at least a 25k drop (Table). There were 939 (7.9%) deaths, 48 (0.4%) HA-VTE, and 106 (0.9%) HA-bleeding events. HA-platelet count drops were associated with increasing age, male sex, and admission to an intensive care unit (all p < 0.05). All definitions of platelet count drops were associated with mortality, HA-VTE, and HA-bleeding (Table). A 10% platelet count drop was associated with increased mortality (OR 1.52, CI: 95% 1.30-1.79), HA-VTE (OR 5.19, CI: 95% 1.83-14.74), and HA-bleeding (OR 8.83, CI: 95% 3.20-24.36) and an absolute 25k drop was associated with increased mortality (OR 1.60, CI: 95% 1.36-1.88), HA-VTE (OR 4.27, CI: 95% 1.64-11.11), and HA-bleeding (OR 5.22, CI: 95% 2.38-11.49).
Conclusion: Platelet count drops, even those considered clinically insignificant, identify a large number of hospitalized medical patients at increased risk for mortality, HA-VTE, and HA-bleeding. Our findings are not driven by severe HA-thrombocytopenia as only 2% of admissions developed platelet counts <100,000. HA-platelet count drops are likely a good marker of illness severity in this population and could identify patients at increased risk for mortality, HA-VTE and HA-bleeding allowing targeted interventions to improve patient outcomes.
Association of Hospital-Acquired Platelet Count Drops with Mortality, HA-VTE and HA-Bleeding in Medical Patients
| Platelet Drop . | Admissions = 11,863 . | Odds Ratio (95% Confidence Interval) . | ||
|---|---|---|---|---|
| . | N, % . | Mortality N = 939 . | HA-VTE N = 48 . | HA-Bleeding N = 106 . |
| Absolute Nadir | ||||
| <150k | 1,905 (16.1%) | 2.0 (1.7, 2.5) | 4.3 (2.3, 7.9) | 2.7 (1.8, 4.2) |
| <100k | 235 (2.0%) | 4.4 (3.0, 6.3) | 5.4 (2.3, 12.5) | 3.2 (1.8, 5.8) |
| Relative Drop | ||||
| 50% | 371 (3.1%) | 3.8 (2.8, 5.2) | 6.3 (3.1, 12.8) | 5.0 (3.1, 8.0) |
| 30% | 1,748 (14.7%) | 2.5 (2.1, 3.0) | 4.2 (2.2, 7.9) | 3.6 (2.3, 5.6) |
| 10% | 6,971 (58.8%) | 1.5 (1.3, 1.8) | 5.2 (1.8, 14.7) | 8.8 (3.2, 24.4) |
| Absolute Drop | ||||
| 100k | 1,186 (10.0%) | 2.7 (2.2, 3.3) | 6.4 (3.2, 12.8) | 4.3 (2.8, 6.8) |
| 75k | 2,019 (17.0%) | 2.4 (2.0, 2.3) | 5.0 (2.6, 9.9) | 4.8 (3.1, 7.7) |
| 50k | 3,594 (30.3%) | 1.9 (1.7, 2.3) | 3.3 (1.7, 6.6) | 5.5 (3.2, 9.4) |
| 25k | 6,737 (56.8%) | 1.6 (1.4, 1.9) | 4.3 (1.6, 11.1) | 5.2 (2.4, 11.5) |
| Platelet Drop . | Admissions = 11,863 . | Odds Ratio (95% Confidence Interval) . | ||
|---|---|---|---|---|
| . | N, % . | Mortality N = 939 . | HA-VTE N = 48 . | HA-Bleeding N = 106 . |
| Absolute Nadir | ||||
| <150k | 1,905 (16.1%) | 2.0 (1.7, 2.5) | 4.3 (2.3, 7.9) | 2.7 (1.8, 4.2) |
| <100k | 235 (2.0%) | 4.4 (3.0, 6.3) | 5.4 (2.3, 12.5) | 3.2 (1.8, 5.8) |
| Relative Drop | ||||
| 50% | 371 (3.1%) | 3.8 (2.8, 5.2) | 6.3 (3.1, 12.8) | 5.0 (3.1, 8.0) |
| 30% | 1,748 (14.7%) | 2.5 (2.1, 3.0) | 4.2 (2.2, 7.9) | 3.6 (2.3, 5.6) |
| 10% | 6,971 (58.8%) | 1.5 (1.3, 1.8) | 5.2 (1.8, 14.7) | 8.8 (3.2, 24.4) |
| Absolute Drop | ||||
| 100k | 1,186 (10.0%) | 2.7 (2.2, 3.3) | 6.4 (3.2, 12.8) | 4.3 (2.8, 6.8) |
| 75k | 2,019 (17.0%) | 2.4 (2.0, 2.3) | 5.0 (2.6, 9.9) | 4.8 (3.1, 7.7) |
| 50k | 3,594 (30.3%) | 1.9 (1.7, 2.3) | 3.3 (1.7, 6.6) | 5.5 (3.2, 9.4) |
| 25k | 6,737 (56.8%) | 1.6 (1.4, 1.9) | 4.3 (1.6, 11.1) | 5.2 (2.4, 11.5) |
Mortality - Adjusted additionally for: Respiratory Rate, Respiratory Dysfunction (intubated or oxygen saturation <90%), Heart Rate, Temperature, Diabetes, Cancer, and HIV (Brabrand, PLoS ONE 2015)
HA-VTE - Adjusted additionally for: Anticoagulation (prophylactic and full dose), Cancer, Heart Failure, Respiratory Dysfunction, Rheumatologic or Inflammatory Disease, and Tachycardia (Zakai, JTH 2013)
HA-Bleeding - Adjusted additionally for: Anticoagulation (prophylactic and full dose), Cancer, Renal Function, Heart Failure, Respiratory Dysfunction, Rheumatologic or Inflammatory Disease, and Tachycardia (Decousis, Chest 2011)
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal