A multinational study of thromboprophylaxis practice in critically ill children

Abstract

Objectives: Although critically ill children are at increased risk for developing deep venous thrombosis, there are few pediatric studies establishing the prevalence of thrombosis or the efficacy of thromboprophylaxis. We tested the hypothesis that thromboprophylaxis is infrequently used in critically ill children even for those in whom it is indicated.

Design: Prospective multinational cross-sectional study over four study dates in 2012.

Setting: Fifty-nine PICUs in Australia, Canada, New Zealand, Portugal, Singapore, Spain, and the United States.

Patients: All patients less than 18 years old in the PICU during the study dates and times were included in the study, unless the patients were 1) boarding in the unit waiting for a bed outside the PICU or 2) receiving therapeutic anticoagulation.

Interventions: None.

Measurements and main results: Of 2,484 children in the study, 2,159 (86.9%) had greater than or equal to 1 risk factor for thrombosis. Only 308 children (12.4%) were receiving pharmacologic thromboprophylaxis (e.g., aspirin, low-molecular-weight heparin, or unfractionated heparin). Of 430 children indicated to receive pharmacologic thromboprophylaxis based on consensus recommendations, only 149 (34.7%) were receiving it. Mechanical thromboprophylaxis was used in 156 of 655 children (23.8%) 8 years old or older, the youngest age for that device. Using nonlinear mixed effects model, presence of cyanotic congenital heart disease (odds ratio, 7.35; p < 0.001) and spinal cord injury (odds ratio, 8.85; p = 0.008) strongly predicted the use of pharmacologic and mechanical thromboprophylaxis, respectively.

Conclusions: Thromboprophylaxis is infrequently used in critically ill children. This is true even for children at high risk of thrombosis where consensus guidelines recommend pharmacologic thromboprophylaxis.

Figure 1

The frequency of pharmacologic (A) and mechanical (B) thromboprophylaxis was highly variable depending on the presence of risk factor. Each bar represents the proportion of patients with the risk factor. The denominator is the total number of patients in the study. The circles represent the proportion of patients at risk receiving thromboprophylaxis. The denominator is the total number of patient with the risk factor. Because mechanical thromboprophylaxis can only be used in children 8 years old or older, the proportion of patients at risk in (B) represents only the subset of patients in this age group. CHD = congenital heart disease.

Figure 2

The frequency of pharmacologic thromboprophylaxis was highest at the extremes of age while the frequency of mechanical thromboprophylaxis initially increased but seemed to plateau at 13 years old or older. Filled circles represent pharmacologic thromboprophylaxis; filled squares represent mechanical thromboprophylaxis; gray bars represent proportion of patients per year; solid line represents fitted values (pharmacologic); and dashed line represents fitted values (mechanical).

Figure 3

The frequency of pharmacologic (A) and mechanical (B) thromboprophylaxis was highly variable across PICUs. Each bubble represents a PICU. The size of the bubble represents the relative number of patients in the PICU during the study dates. The x-axis represents the identification number assigned to each PICU. The solid line represents the mean frequency of thromboprophylaxis for the entire patient cohort while the dashed lines represent the range.

Figure 4

Different patient and PICU characteristics were independently associated with pharmacologic (A) and mechanical (B) thromboprophylaxis. The x-axes are in the logarithmic scale. Only the risk factors with p values less than 0.05 in the multivariable model are presented. CHD = congenital heart disease.