Abstract
Objective
To identify patient, hospital and central venous catheter (CVC) factors that may influence the use of low dose heparin infusion (LDHI) for CVC patency in critically ill-children.
Design
Secondary analysis of an international multicenter observational study.
Setting
59 Pediatric Intensive Care Units (PICUs) over four study dates in 2012, involving 7 countries.
Patients
Children less than 18 years of age with a CVC, admitted to a participating unit and enrolled in the completed PROTRACT study were included. All overflow patients were excluded.
Interventions
None.
Measurements and Main Results
Of the 2,484 patients in the PROTRACT study, 1,312 patients had a CVC. 507 of those patients used LDHI. The frequency of LDHI was compared across various patient, hospital and CVC factors using chi-squared, Mann-Whitney and Fisher's exact tests. In the multivariate analysis, age was not a significant factor for LDHI use. Patients with pulmonary hypertension had decreased LDHI use while those with active surgical or trauma diagnoses had increased LDHI use. All central CVC insertion sites were more likely to use LDHI when compared to peripherally inserted CVCs. The Asia-Pacific region showed increased LDHI use, along with community hospitals and smaller ICUs (<10 beds).
Conclusion
Patient, CVC, and hospital factors are associated with the use of LDHI in critically ill children. Further study is needed to evaluate the efficacy and persistence of LDHI use.
Keywords: Pediatrics, Critical Care, Central Venous Access Device, Anticoagulation, Venous Thromboembolism, Intensive Care Units, Pediatric
Introduction
Central venous catheters (CVCs) are widely used in the care and management of critically ill children. However, CVCs have the potential to lose functionality from occlusion if they are not kept patent (1). Current studies in neonates find that the use of low dose heparin infusion (LDHI), to maintain CVC patency is of some benefit, in that it reduces catheter occlusion and prolongs CVC usability without increasing adverse effects (2) (3) (4) (5). However, there is minimal evidence supporting the efficacy of LDHI use for CVC patency in critically-ill children outside the neonatal period. Multiple studies in this population were inconclusive (6) (7) (8) (9) (10) (11) (12) (13) or found no significant effect on the patency of CVCs with LDHI (14). Additionally, no studies were found to detail the frequency of LDHI use to maintain CVC patency in pediatric medicine. In this study we aimed to identify the global prevalence of LDHI (defined as <10U/hr of unfractionated heparin) and the factors associated with LDHI, as a secondary analysis of the Prophylaxis against Thrombosis Practice (PROTRACT) study, a multinational, point-prevalence study of thromboprophylaxis practice in critically-ill children (15). Documenting the use of LDHI using the large and diverse population represented in PROTRACT, presented the opportunity to improve the global knowledge of care and maintenance of CVCs in children. We hypothesized that patient, hospital, and CVC factors would influence the use of LDHI as a technique for maintaining CVC patency.
Materials and Methods
We performed a secondary analysis of the PROTRACT study, a prospective, observational study involving seven countries, over four study dates (February 1, May 1, August 1, and November 1, 2012) to account for the seasonality of admissions to the Pediatric Intensive Care Unit (PICU) (15). The study enrolled children less than 18 years of age that were hospitalized in one of the 59 participating PICUs.
Overflow patients and those who received extracorporeal membrane oxygen, continuous renal replacement therapy, vascular assist devices or therapeutic anticoagulation at the time of study were excluded. At the local time of 9 am on each study date the following data were collected for each PROTRACT study patient: demographics, preadmission history, and ongoing factors presumptively associated with venous thromboembolism (VTE) risk, including the number, type and location of CVCs. Pharmacologic and mechanical VTE prophylaxis were recorded. Data were gathered from the medical records into standardized case report forms by trained abstractors, who were typically physicians or nurses, complying with a manual containing detailed descriptions of each data point.
Design
After local IRB approval was obtained, we conducted a retrospective secondary analysis of the PROTRACT Study. Children less than 18 years of age with a CVC, who were admitted to a participating PICU and enrolled in the PROTRACT study were included. Data from the four study dates were combined for analysis. The primary outcome measure was the prevalence of LDHI. Secondary outcomes were patient, CVC and hospital factors associated with LDHI use.
Data Collection
The following data were collected on each study date:
Patient factors: gender, race, ethnicity, age, weight, diagnoses, bleeding tendency (platelet count, international normalized ratio [INR], partial thromboplastin time [PTT]), family history, patient history and therapeutic interventions.
CVC factors: CVC type and anatomic site of insertion.
Hospital factors: PICU size and type, geographic location, hospital/institution type.
Definitions
LDHI was defined as administration of <10U/hr of unfractionated heparin through the CVC presumably for line patency. Higher doses were considered prophylactic anticoagulation.
Practitioners have identified the presence of cardiac disease, surgery or bleeding tendency of patients as factors that influence the decision to prescribe pharmacologic prophylaxis (16). As these factors may also influence the decision to prescribe LDHI, combined categories were created. Bleeding tendency included diagnoses of cerebrovascular accident, non-intracranial hemorrhage, and intra-cranial hemorrhage. Heart disease included patients with cyanotic congenital heart disease, acyanotic congenital heart disease, dilated cardiomyopathy and those listed for heart transplant. Surgical/Trauma included patients with burns, spinal cord Injury, lower extremity fracture, craniotomy, traumatic brain injury, lower extremity surgery, scoliosis repair/spine surgery, and solid organ transplant.
Analysis
Using chi-squared, Mann-Whitney and Fisher's exact tests, the frequency of LDHI use was compared across each patient, hospital and CVC factor. For patient factors, frequencies were generated in terms of LDHI use, as well as the percentages of LDHI use with specific patient, CVC and hospital factors present and absent. Chi-square/Fisher's Exact test was used to compare the proportion of LDHI usage across the predictor (patient related, central venous catheter related and hospital related) categories. A multivariable logistic regression was used and a stepwise selection method (with α = 0.10) was applied to find the best set of predictors for the LDHI usage. P value <0.05 was considered significant. SAS v9.4 (SAS Institute Inc., Cary, NC) was used for the analyses.
Results
Prevalence of Low Dose Heparin Infusion
One thousand three hundred twelve critically-ill children with CVCs out of the 2,484 patients that met the inclusion criteria for PROTRACT study enrollment, were included in this secondary analysis. Of these, 507 used LDHI (38%; 95% CI: 36.0%-41.3%). The median LDHI concentration was 1 U/ml (IQR 1-2). This did not change with age.
Factors Associated with Low Dose Heparin Infusion
Patient factors associated with LDHI use from univariate analysis are presented in Table 1. LDHI was associated with younger patients, with ages >13 years having the lowest percentage of LDHI use (30%). Patients of Asian race were more likely to use LDHI (54% with odds ratio [OR] = 2.03) compared to other races, and Hispanic/Latino ethnicity showed greater likelihood of LDHI use (45.2% with OR=1.36) than non-Hispanic/Latinos.
Table 1.
Patient-related Predictors of Low Dose Heparin Infusion (Univariate)
| Total N | LDHIa (yes)% | OR | 95/ CI (OR) | p | |
|---|---|---|---|---|---|
| Race | |||||
| White = Ref | 923 | 339 (36.7%) | |||
| Black | 193 | 78 (40.4%) | 1.17 | 0.85-1.60 | 0.34 |
| Asian | 61 | 33 (54.1%) | 2.03 | 1.21-3.42 | 0.008 |
| Other | 135 | 57 (42.2%) | 1.26 | 0.87-1.82 | 0.22 |
| Ethnicity | |||||
| Hispanic/Latino | 221 | 100 (45.2%) | 1.36 | 1.01-1.82 | 0.04 |
| Non-Hispanic/Latino=Ref | 982 | 372 (37.9%) | |||
| Unknown | 109 | 35 (32.1%) | 0.78 | 0.51-1.18 | 0.24 |
| Age | |||||
| 0-1 =Ref | 740 | 313 (42.3%) | |||
| 1 - 12 | 417 | 148 (35.5%) | 0.75 | 0.59-0.96 | 0.03 |
| ≥13 | 155 | 46 (29.7%) | 0.58 | 0.40-0.84 | 0.004 |
| Active Diagnosis | |||||
| Acyanotic CHDb | 175 | 94 (53.7%) | 2.03 | 1.47-2.80 | <0.0001 |
| Cyanotic CHDb | 234 | 135 (57.7%) | 2.59 | 1.94-3.45 | <0.0001 |
| Dilated Cardiomyopathy | 28 | 11 (39.3%) | 1.03 | 0.48-2.21 | 0.94 |
| Listed for Heart Transplant | 7 | 2 (28.6%) | 0.53 | 0.08-3.40 | 0.51 |
| Cerebrovascular accident | 28 | 14 (50%) | 1.60 | 0.76-3.39 | 0.22 |
| Pulmonary Hypertension | 117 | 33 (28.2%) | 0.60 | 0.39-0.91 | 0.02 |
| Cancer | 104 | 28 (26.9%) | 0.56 | 0.36-0.88 | 0.01 |
| IBD | 4 | 2 (50%) | 1.59 | 0.22-11.32 | 0.64 |
| Sepsis/Bacteremia | 189 | 72 (38.1%) | 0.97 | 0.71-1.34 | 0.87 |
| Nephrotic Syndrome | 6 | 3 (50%) | 1.59 | 0.32-7.92 | 0.57 |
| End Stage Renal Disease | 20 | 9 (45%) | 1.30 | 0.54-3.17 | 0.56 |
| Burns | 8 | 2 (25%) | 0.53 | 0.11-2.62 | 0.43 |
| Spinal Cord Injury | 11 | 5 (45.4%) | 1.33 | 0.40-4.37 | 0.64 |
| Lower Extremity fracture | 15 | 2 (13.3%) | 0.24 | 0.05-1.07 | 0.06 |
| Craniotomy | 54 | 28 (51.8%) | 1.75 | 1.02-3.02 | 0.04 |
| Traumatic brain injury | 43 | 14 (32.6%) | 0.76 | 0.40-1.45 | 0.41 |
| Lower Extremity Surgery | 9 | 4 (44.4%) | 1.27 | 0.34-4.76 | 0.72 |
| Scoliosis Repair/ Other Spine Surgery | 42 | 11 (26.2%) | 0.55 | 0.28-1.11 | 0.10 |
| Intra-cranial hemorrhage | 49 | 19 (38.8%) | 1.01 | 0.56-1.81 | 0.98 |
| Non-intracranial hemorrhage | 34 | 15 (44.1%) | 1.26 | 0.64-2.50 | 0.51 |
| Solid Organ Transplant | 32 | 13 (40.6%) | 1.09 | 0.53-2.23 | 0.81 |
| Grouped Diagnoses | |||||
| Bleedingc | 92 | 40 (43.5%) | 1.24 | 0.81-1.90 | 0.32 |
| Heart Diseased | 419 | 228 (54.4%) | 2.63 | 2.07-3.34 | <0.0001 |
| Surgical/Traumae | 179 | 69 (38.6%) | 0.995 | 0.72-1.38 | 0.98 |
| Interventions | |||||
| Parenteral Nutrition | 368 | 154 (41.8%) | 1.20 | 0.94-1.54 | 0.14 |
| Packed RBCsf | 232 | 103 (44.4%) | 1.34 | 1.00-1.78 | 0.05 |
| Fresh Frozen Plasma Transfusion | 100 | 48 (48%) | 1.51 | 1.01-2.28 | 0.05 |
| Platelet Transfusion | 107 | 48 (44.9%) | 1.32 | 0.89-1.97 | 0.17 |
| Cryoprecipitate transfusion | 20 | 16 (80%) | 6.52 | 2.17-19.63 | 0.0008 |
| Invasive mechanical ventilation | 783 | 317 (40.5%) | 1.21 | 0.97-1.52 | 0.10 |
| Non-invasive mechanical ventilation | 187 | 62 (33.2%) | 0.76 | 0.55-1.05 | 0.10 |
| L-asparaginase therapy | 8 | 3 (37.5%) | 0.95 | 0.23-4.00 | 0.95 |
| Complete bed rest | 633 | 246 (38.9%) | 1.02 | 0.82-1.27 | 0.87 |
| Vasoactive agent | 411 | 233 (56.7%) | 3.00 | 2.35-3.81 | <0.0001 |
| Anticipated surgery | 97 | 44 (45.4%) | 1.35 | 0.89-2.04 | 0.16 |
| Prior Surgery | 478 | 223 (46.6%) | 1.69 | 1.35-2.13 | <0.0001 |
| Anticoagulant Use | |||||
| Aspirin | 93 | 43 (46.2%) | 1.40 | 0.92-2.14 | 0.12 |
| Low Molecular Weight Heparin | 83 | 28 (33.7%) | 0.80 | 0.50-1.27 | 0.34 |
| Inravenous Heparin | 61 | 43 (70.5%) | 4.05 | 2.31-7.11 | <0.0001 |
| Subcutaneous heparin | 7 | 3 (42.9%) | 1.19 | 0.27-5.35 | 0.82 |
| Vitamin K antagnostic | 9 | 3 (33.3%) | 0.79 | 0.20-3.18 | 0.74 |
| Grouped Anticoagulantg | 223 | 110 (49.3%) | 1.70 | 1.27-2.27 | 0.0004 |
| Patient History/Pre-Study Info | |||||
| Personal History Thrombosis | 83 | 35 (42.2%) | 1.01 | 0.64-1.59 | 0.95 |
| Congenital thrombophilia | 5 | 2 (40%) | 0.97 | 0.16-5.85 | 0.98 |
| Acquired Hypercoagulability | 6 | 2 (33.3%) | 0.70 | 0.13-3.85 | 0.68 |
| Sickle Cell Disease | 6 | 5 (83.3%) | 7.46 | 0.87-64.02 | 0.04 |
| Family Hx of Thrombosis | 3 | 2 (66.7%) | 2.98 | 0.27-33.05 | 0.37 |
| Obesity | 23 | 8 (34.8%) | 0.84 | 0.36-2.01 | 0.70 |
| Home Parenteral Nutrition | 29 | 3 (10.3%) | 0.18 | 0.05-0.58 | 0.004 |
| Multiple Cardiac Catheterizations | 113 | 58 (51.3%) | 1.70 | 1.15-2.50 | 0.008 |
| Complex Surgical Cardiac Repair | 345 | 200 (58%) | 2.96 | 2.30-3.82 | <0.0001 |
| Cavopulmonary Anastomosis | 69 | 45 (65.2%) | 3.17 | 1.91-5.27 | <0.0001 |
LDHI = Low Dose Heparin Infusion
CHD = Congenital Heart Disease
Bleeding = Cerebrovascular accident, Non-intracranial hemorrhage, Intra-cranial hemorrhage
Heart Disease = Cyanotic CHD, Acyanotic CHD, Dilated Cardiomyopathy, Listed for Heart Transplant
Surgical/Trauma = Burns, Spinal Cord Injury, Lower Extremity Fracture, Craniotomy, Traumatic Brain Injury, Lower Extremity Surgery, Scoliosis Repair/Spine Surgery, Solid Organ Transplant
RBCs = Red Blood Cells
Grouped Anticoagulant = Aspirin, Low Molecular Weight Heparin, Intravenous + Subcutaneous Heparin, Vitamin K
Note: Extracorporeal Membrane Oxygenation, Continuous Renal Replacement therapy, Ventricular Assist Device, Allergy to Heparin, Heparin-induced Thrombocytopenia, Granulocyte Transfusion, Diabetic Ketoacidosis and Pregnancy/Lactation had 0-1 cases in study with no use of LDHI
Patients with diagnoses of heart disease, craniotomy, and sickle-cell disease, had increased LDHI use (54% with OR=2.63, 52% with OR=1.75, and 83% with OR=7.46 respectively) while those with pulmonary hypertension and cancer had decreased use (28% with OR=0.60 and 27% with OR=0.56 respectively). Patients with bleeding tendencies, including stroke, non-intracranial hemorrhage, or intracranial hemorrhage, had no significant association with LDHI use. Patients receiving prophylactic anticoagulation (aspirin, low molecular weight heparin, intravenous heparin, or vitamin K antagonists) were more likely to receive LDHI (49.3% with OR=1.70).
Patients who participated in the study with a history of multiple cardiac catheterizations, complex surgical cardiac repair, and/or cavopulmonary anastomosis, had increased LDHI use (51% with OR=1.70, 58% with OR=2.96 and 65% with OR=3.17 respectively) while patients on home parenteral nutrition showed decreased LDHI use (10.3% with OR=0.18). Personal or family history of thrombosis did not influence the use of LDHI.
LDHI use varied by CVC anatomic site and type, with transthoracic and umbilical cord locations as the most common CVC sites associated with LDHI use (85% with OR=17.73 and 69% with OR=7.26 respectively). Atrial and Umbilical CVC types were associated with increased LDHI use (85% with OR=5.39 and 69% with OR=2.21 respectively). Patients with more than one CVC were more likely to receive LDHI compared to those with one CVC (53% vs. 37%, p<0.001) (Table 2).
Table 2.
Central Venous Catheter Related Predictors of Low Dose Heparin Use (Univariate)
| CVC Site and Type | Total N | LDHIa N (%) | OR | 95% CI (OR) | P |
|---|---|---|---|---|---|
| CVCb Site | |||||
| Transthoracic | 13 | 11 (84.6%) | 17.73 | 3.86-81.37 | 0.0002 |
| Umbilical cord | 26 | 18 (69.2%) | 7.26 | 3.06-17.20 | <0.0001 |
| Subclavian vein | 205 | 64 (31.2%) | 1.46 | 1.01-2.12 | 0.04 |
| Internal jugular vein | 319 | 138 (43.3%) | 2.46 | 1.79-3.38 | <0.0001 |
| Femoral vein | 326 | 173 (53.1%) | 3.65 | 2.66-4.99 | <0.0001 |
| Peripheral = Ref | 414 | 98 (23.7%) | |||
| CVCb Type | |||||
| Hemofiltration/dialysis | 6 | 3 (50%) | 0.98 | 0.20-4.90 | 0.98 |
| Atrial | 13 | 11 (84.6%) | 5.39 | 1.18-24.53 | 0.03 |
| Umbilical | 26 | 18 (69.2%) | 2.21 | 0.94-5.15 | 0.07 |
| Tunneled | 166 | 41 (24.7%) | 0.32 | 0.22-0.47 | <0.0001 |
| Peripherally inserted | 493 | 128 (26%) | 0.34 | 0.27-0.44 | <0.0001 |
| Untunneled = Ref | 606 | 306 (50.5%) | |||
LDHI = Low Dose Heparin Infusion
CVC = Central Venous Catheter
Patients in cardiac ICUs had increased LDHI use (74% with OR=3.92) in comparison to those in medical/surgical, and mixed ICUs (20% with OR=0.34 and 42% respectively). Smaller ICUs had greater use of LDHI than larger ICUs (70% vs. 27%, p<0.001) (Table 3).
Table 3.
Hospital Related Predictors of Low Dose Heparin Use (Univariate)
| Total N | LDHIa N (%) | OR | 95%% CI | P | |
|---|---|---|---|---|---|
| Hospital Location | |||||
| Asia, Australia or New Zealand | 168 | 100 (59.5%) | 2.79 | 1.99-3.92 | <0.0001 |
| Canada | 72 | 14 (19.4%) | 0.46 | 0.25-0.84 | <0.02 |
| Spain or Portugal | 173 | 83 (48%) | 1.75 | 1.26-2.43 | 0.0008 |
| United States = Ref | 899 | 310 (34.5%) | |||
| Hospital/Institution Type | |||||
| Community w/ Academic Affiliation | 261 | 130 (49.8%) | 1.80 | 1.36-2.36 | <0.0001 |
| Community without Academic Affiliation | 48 | 20 (41.7%) | 1.29 | 0.72-2.33 | 0.39 |
| Academic = Ref | 1003 | 357 (35.6%) | |||
| Children's Hospital/Institution | |||||
| Yes | 1090 | 387 (35.5%) | 0.47 | 0.35-0.63 | <0.0001 |
| No = Ref | 222 | 120 (54%) | |||
| Pediatric ICUb Type | |||||
| Medical or Surgical | 397 | 79 (19.9%) | 0.34 | 0.26-0.46 | <0.0001 |
| Cardiac | 138 | 102 (73.9%) | 3.92 | 2.61-5.88 | <0.0001 |
| Mixed medical or Surgical and Cardiac=Ref for all 3 | 777 | 326 (42%) | |||
| Pediatric ICUb Size | |||||
| ≤10 beds | 95 | 67 (70.5%) | 9.34 | 5.45-15.99 | <0.0001 |
| 11-20 beds | 478 | 191 (40%) | 2.60 | 1.82-3.71 | <0.0001 |
| 21-30 beds | 489 | 198 (40.5%) | 2.66 | 1.86-3.79 | <0.0001 |
| >30 beds=Ref | 250 | 51 (20.4%) | |||
| Pediatric ICUb Annual Admissions | |||||
| ≤500 | 164 | 84 (51.2%) | 2.86 | 1.97-4.14 | <0.0001 |
| 501-1000 | 355 | 182 (51.3%) | 2.86 | 2.13-3.85 | <0.0001 |
| 1001-1500 | 350 | 122 (34.9%) | 1.46 | 1.08-1.98 | 0.02 |
| >1500 =Ref | 443 | 119 (26.9%) | |||
LDHI = Low Dose Heparin Infusion
ICU = Intensive Care Unit
Geographic location influenced LDHI use, with the Asia-Pacific region having the highest usage of LDHI (60% with OR=2.79) and Canada having the least use (19% with OR=0.46) (Table 3).
Multivariable Analysis
In the multivariable analysis of patient related predictors, age was not a significant factor for the use of LDHI. Asian and Black race, and Hispanic/Latino ethnicity remained significant for increased use of LDHI (OR= 2.44, 1.61 and 1.52 respectively). Patients with active pulmonary hypertension remained significant with decreased LDHI use (OR= 0.54). Patients with a diagnosis of surgical trauma continued to be associated with increased LDHI use (OR= 1.58). Other factors associated with increased LDHI use were cryoprecipitate transfusion and the use of vasoactive agents (OR=4.65 and 1.48, respectively) (Table 4). Patient history of complex surgical cardiac repair remained a significant factor for increased LDHI use (OR=2.23) while patients on home parenteral nutrition continued to be a significant factor for decreased LDHI use (OR=0.29). Pre-study history of sickle-cell disease, multiple cardiac catheterizations, and cavopulmonary anastomosis were no longer significant patient factors.
Table 4.
Multivariate Analysis of Patient, Central Venous Catheter and Hospital Factors
| Total N | LDHIa (yes)% | OR | 95% CI (OR) | p | |
|---|---|---|---|---|---|
| Patient related predictors | |||||
| Race | |||||
| Black | 193 | 78 (40.4%) | 1.61 | 1.07-2.43 | 0.02 |
| Asian | 61 | 33 (54.1%) | 2.44 | 1.29-4.60 | 0.04 |
| Other | 135 | 57 (42.2%) | 1.21 | 0.72-2.03 | 0.46 |
| White = Ref | 923 | 339 (36.7%) | |||
| Ethnicity | |||||
| Hispanic/Latino | 221 | 100 (45.2%) | 1.52 | 1.00-2.32 | 0.05 |
| Unknown | 109 | 35 (32.1%) | 0.78 | 0.43-1.42 | 0.42 |
| Non-Hispanic/Latino=Ref | 982 | 372 (37.9%) | |||
| Active Diagnosis Yes vs No=Ref | |||||
| Pulmonary Hypertension | 117 | 33 (28.2%) | 0.54 | 0.31-0.92 | 0.02 |
| Grouped Diagnoses: Surgical/Traumab | 179 | 69 (38.6%) | 1.58 | 1.06-2.36 | 0.02 |
| Interventions (yes/no = Ref) | |||||
| Cryoprecipitate transfusion | 20 | 16 (80%) | 4.65 | 1.20-18.13 | 0.03 |
| Vasoactive agent | 411 | 233 (56.7%) | 1.48 | 1.08-2.04 | 0.01 |
| Patient History (yes/no=Ref) | |||||
| Home Parenteral Nutrition | 29 | 3 (10.3%) | 0.29 | 0.08-1.11 | 0.07 |
| Complex Surgical Cardiac Repair | 345 | 200 (58%) | 2.23 | 1.55-3.20 | <0.0001 |
| CVCc Related Predictors: Site and Type | |||||
| CVCc Site | |||||
| Subclavian vein | 205 | 64 (31.2%) | 1.42 | 0.72-2.82 | 0.31 |
| Internal jugular vein | 319 | 138 (43.3%) | 1.12 | 0.56-2.23 | 0.74 |
| Femoral vein | 326 | 173 (53.1%) | 2.09 | 1.11-3.93 | 0.02 |
| Peripheral = Ref | 414 | 98 (23.7%) | |||
| CVCc Type | |||||
| Tunneled | 166 | 41 (24.7%) | 0.59 | 0.36-0.96 | 0.04 |
| Peripherally inserted | 493 | 128 (26%) | 0.62 | 0.35-1.12 | 0.49 |
| Untunneled = Ref | 606 | 306 (50.5%) | |||
| Hospital related predictors | |||||
| Hospital Location | |||||
| Asia, Australia or New Zealand | 168 | 100 (59.5%) | 3.31 | 1.86-5.89 | <0.0001 |
| Canada | 72 | 14 (19.4%) | 0.52 | 0.23-1.14 | 0.10 |
| Spain or Portugal | 173 | 83 (48%) | 1.34 | 0.69-2.58 | 0.38 |
| United States = Ref | 899 | 310 (34.5%) | |||
| Hospital/Institution | |||||
| Community w/ Academic Affiliation | 261 | 130 (49.8%) | 1.84 | 1.20-2.83 | 0.005 |
| Community without Academic Affiliation | 48 | 20 (41.7%) | 2.78 | 1.18-6.53 | 0.02 |
| Academic = Ref | 1003 | 357 (35.6%) | |||
| Children's Hospital/Institution | |||||
| Yes | 1090 | 387 (35.5%) | 0.40 | 0.25-0.65 | 0.0002 |
| No = Ref | 222 | 120 (54%) | |||
| Pediatric ICUd Type | |||||
| Medical or Surgical | 397 | 79 (19.9%) | 0.62 | 0.42-0.91 | 0.02 |
| Cardiac | 138 | 102 (73.9%) | 4.40 | 2.52-7.67 | <0.0001 |
| Mixed medical or Surgical and Cardiac=Ref for all 3 | 777 | 326 (42%) | |||
| Pediatric ICUd Size | |||||
| ≤10 beds | 95 | 67 (70.5%) | 3.40 | 1.42-8.10 | 0.006 |
| 11-20 beds | 478 | 191 (40%) | 0.92 | 0.53-1.60 | 0.78 |
| 21-30 beds | 489 | 198 (40.5%) | 1.39 | 0.86-2.26 | 0.18 |
| >30 beds=Ref | 250 | 51 (20.4%) | |||
LDHI=Low Dose Heparin Infusion
Surgical Trauma= Burns, Spinal Cord Injury, Lower Extremity Fracture, Craniotomy, Traumatic Brain Injury, Lower Extremity Surgery, Scoliosis Repair/Spine Surgery, Solid Organ Transplant
CVC=Central Venous Catheter
ICU=Intensive Care Unit
Note: Some of the CVC data (transthoracic, umbilical vein CVCs) were excluded for ensuring multivariable model convergence
Amongst the hospital related predictors, the Asia-Pacific region remained significant for increased LDHI use (OR=3.31) (Table 4). Hospital and institution type also showed significance; community hospitals (both with and without academic affiliation) were more likely to use LDHI compared to the academic institutions (OR= 1.84 and 2.78 respectively) and children's hospitals were less likely to use LDHI (OR= 0.40) (Table 4). When compared to mixed ICUs, medical or surgical pediatric ICU types were less likely to use LDHI (OR= 0.62), whereas cardiac ICUs were more likely to use LDHI (OR=4.40) (Table 4). Using an ICU size of >30 beds as a reference, smaller ICUs (<10 beds) showed a greater likelihood of using LDHI (OR=3.40) (Table 4).
An alternative model (not shown) which included CVC insertion site, but not CVC type (due to the small numbers in some of the types) allowed inclusion of transthoracic and umbilical vein categories of CVC site. After this analysis, all CVC sites were more likely to use LDHI when compared to peripherally located CVCs (OR= 8.27-transthoracic, 4.43-umbilical, 3.81-femoral, 2.24-internal jugular vein, 1.7-subclavian).
Discussion
In this multi-national post-hoc analysis of the PROTRACT study, the prevalence of LDHI was 38%, with varying patient, CVC and institutional factors associated with its use. The regional and geographic variation in practice is not unexpected, and is reflective of the inconsistent literature on the efficacy of LDHI. Some neonatal studies support LDHI for prolonging CVC patency (2) (3 (4) (5). This contrasts with adult evidence which conveys that there is no benefit of heparin prophylaxis vs. saline (17). The limited pediatric evidence is varied, with most studies unable to draw distinct conclusions or showing no benefit (6) (7) (8) (9) (10) (11) (12) (13) (14). The small volume of patients at any institution, along with the technical and financial difficulties of carrying out large-scale clinical trials, forces practitioners to make patient-management decisions in spite of limited evidence. This decision-making appears to be influenced as much by hospital and geographic culture as by patient factors. This is consistent with a previous survey of pediatric intensivists, where the second most common criterion to initiate heparin prophylaxis was attending preference, 42% (18).
Multiple co-morbidities were collected and analyzed for impact on LDHI use as part of the PROTRACT study. These included chronic disease, thrombosis history, obesity, TPN use, as well as current diagnoses and interventions. These variables are listed in full in Table 1. After multivariable analysis, most patient co-morbidities including chronic disease and thrombosis history did not impact LDHI use, while hospital and geographic factors retained significance. The presence and use of hospital or unit-wide protocols for LDHI use was not collected. The existence and compliance with such protocols would have the potential to minimize the role of factors such as ethnicity, age or pathology as indications for LDHI use.
Patient age was not associated with LDHI after multivariable analysis despite the neonatal evidence for LDHI to prolong catheter patency (2) (3) (4) (5). This could be reflective of some intensivists weighting the adult study conclusions over the neonatal studies, or a preference to standardize treatment practice despite varying patient age. This is also consistent with survey results, where only 38% of intensivists used patient age as a criterion for the initiation of heparin prophylaxis (18).
The divergence between adult and neonatal evidence regarding the efficacy of LDHI could be due to difference in CVC and blood vessel sizes. Accordingly, mechanical obstruction in the smaller vessels of neonates with adverse effects on blood flow is more probable, and may explain the reported efficacy of LDHI for CVC patency in this age group.
Patients with increased LDHI use through transthoracic and umbilical vein sites may reflect a comfort with neonatal evidence, which has showed prolonged patency of CVCs with LDHI. As the neonatal studies provide the only evidence for efficacy of LDHI, it is reasonable to speculate that this may influence providers caring for patients at the highest risk for compromise if CVC patency is lost. The increased use of LDHI in patients with complex surgical cardiac repair, may reflect a heightened specialty-specific (cardiovascular surgery, cardiac intensivist) concern for thrombosis and sequelae such as CVC occlusion.
Multivariable analysis showed independent association of LDHI use by geographic location, with increased use in the Asia-Pacific region. These effects may be due to differences in institutional practice, policy and training techniques. However, increased LDHI in certain races (Black, Asian) remained significant after multivariable analysis, suggesting this patient factor still plays a role in LDHI use.
The amount and details of medication and fluids delivered were not collected in this study, which may have influenced the use of LDHI. For instance, patients requiring continuously infused medications or central venous pressure monitoring may be more likely to receive LDHI than those with CVC-administered intermittent medications. This is supported by the increased association of LDHI with those on vasopressor support. We can speculate that practitioners did not feel the need for LDHI in patients at ‘baseline’ risk, as patients on home TPN had decreased use of LDHI.
This study is limited by its post-hoc analysis of an original study. The patients’ need for the CVC and the rationale for LDHI, were not known. In a previous survey by Clarke, intensivists often prescribed LDHI without a clear rationale. As multiple benefits including CVC patency, line infection and patient thrombosis were postulated (18), indications for LDHI may be difficult to assess even in future prospective study. Other limitations include the a priori definition of LDHI (heparin of <10 U/hr) that may not reflect all practices. The PROTRACT study found that pharmacologic thromboprophylaxis was used in 43% of the children with cyanotic heart disease which included heparin infusion with rates >10U/hr. However, this definition may not entirely distinguish the dosing of LDHI for CVC patency from heparin infusion for systemic prophylaxis, as some providers may have overlapping indications and various subtleties to their use of LDHI. Additionally, outcomes including CVC patency, thrombosis, bleeding, and hospital length of stay were not collected.
The conflicting neonatal and adult evidence concerning the efficacy of LDHI along with the scarcity of data within the pediatric population, corroborates the need for further research, such as a randomized controlled study to determine if there is a specific age that LDHI is no longer beneficial. The inclusion of outcomes such as the number of CVC occlusions, the days of line patency, incidence of venous thromboembolisms, and the need for CVC replacements, would help to determine the efficacy of LDHI use and to clarify why it has mixed implementation in PICUs.
Conclusion
LDHI remains a common method for the maintenance of CVC patency in critically-ill children despite the lack of substantial data supporting its efficacy. In our analysis of the PROTRACT study, we showed that LDHI use has global prevalence, as 38% of patients with CVCs in this study received LDHI. Various patient, hospital and CVC factors are associated with LDHI use. With randomized controlled trials, we hope to shed more light on the effectiveness of LDHI in comparison to other methods of CVC patency such as saline infusion.
Acknowledgement
We thank the investigators and research teams at the many sites who participated and provided data in the PROTRACT study: Akron Children's Hospital – Ann-Marie Brown, MSN; Alberta Children's Hospital – Leena Desai, MSc, Elaine Gilfoyle, MD, MMEd; Baylor College of Medicine – Michelle Goldsworthy, BN, Nancy Jaimon, BSN, MSN, Laura Loftis, MD; Baystate Medical Center – Michael Canarie, MD; Boston Children's Hospital– Daniel Kelly, MD, Adrienne Randolph, MD, MSc; Centre Hospitalier Universitaire de Sherbrooke– Miriam Santschi, MD, MSc; Centre mere-enfant Soleil du CHU de Quebec – Marc-Andre Dugas, MD, MSc, Louise Gosselin, BN; Centro Hospitalar Lisboa Norte – Joana Rios, MD; Children's Hospital and Medical Center, Omaha and Nebraska Medical Center – Edward Truemper, MD, MS, Brenda Weidner, BSN; Children's Hospital and Research Center Oakland – Heidi Flori, MD, Julie Simon, RN; Children's Hospital at Westmead – Marino Festa, MD, Karen Walker, PhD, Nicola Watts, PhD; Children's Hospital of Philadelphia – Daniela Davis, MD, MSCE, Mary Ann DiLiberto, BS, William Kamens, BA, Rebecca McIntosh, BA, Brooke Park, BSN, Janice Prodell, RN; Children's Hospital of Wisconsin – Sheila Hanson, MD, MS, Kathleen Murkowski, BS, David Triscari, BS; Children's Hospitals and Clinics of Minnesota – Chris Leonard, BA, Jeffrey Nowak, MD, Alison Overman, BA; CHU Sainte-Justine University of Montreal – Mariana Dumitrascu, MD, Jacques Lacroix, MD, Marisa Tucci, MD; Cohen Children's Medical Center of New York – Aaron Kessel, MD, James Schneider, MD, Todd Sweberg, MD; Complexo Hospitalario Universitario A Coruna – Angela Ferrer Barba, MD, Carmen Ramil, MD; Connecticut Children's Medical Center – Christopher Carroll, MD, MS; Dell Children's Medical Center of Central Texas – LeeAnn Christie, MSN, Renee Higgerson, MD; Doernbecher Children's Hospital – Aileen Kirby, MD; Duke University Medical Center – Ira Cheifetz, MD, Kyle Rehder, MD, Samantha Wrenn, BS; Gregorio Maranon General University Hospital – Jesus Lopez-Herce, MD, PhD; Helen DeVos Children's Hospital – Nabil Hassan, MD, Akunne Ndika, MBBS, MPH; Hospital Clinico Universitario de Santiago – Antonio Rodriguez-Nunez, MD, PhD, Sara Trabazo-Rodriguez, MD; Hospital Dona Estefania – Maria Ventura, MD; Hospital Infantil Universitario Miguel Servet – Juan Pablo Garcia Íñiguez, MD, Paula Madurga Revilla, MD; Hospital Infantil Universitario Niño Jesus – Maria Isabel Iglesias Bouzas, MD; Hospital Pediatrico Coimbra – Maria Dionisio, MD; Hospital Regional Universitario Materno Infantil Carlos Haya – Patricia García Soler, MD; Hospital Sao Joao – Miguel Fonte, MD; Hospital Universitario Materno Infantil Las Palmas de Gran Canaria – Antonio Jimenez, MD; Joseph M Sanzani Children's Hospital at Hackensack University Medical Center – Shira Gertz, MD; KK Women's and Children's Hospital – Loh Tsee Foong, MBBS, MMed, Anuradha Menon, MBBS; Maria Fareri Children's Hospital – Simon Li, MD, MPH; Mater Children's Hospital – Sara Mayfield, BHSc, Andreas Schibler, MD; Montreal Children's Hospital – Yasser Kazzaz, MBBS, Samara Zavalkoff, MDCM; Nationwide Children's Hospital – Sue Cunningham, BSN, Kristin Greathouse, MS, BSN, Dianna Hidalgo, BS, Sarah O'Brien, MD, MSc, Kami Perdue, BS, Lisa Steele, BSN; Nuestra Senora de Candelaria Hospital – Jose Sebastian Leon Gonzalez, MD; Penn State Children's Hospital – Debra Spear, RN, Robert Tamburro, MD; Princess Margaret Hospital – Simon Erickson, MBBS; Riley Hospital for Children – Stephanie Fritz, RN, Christi Rider, LPN, Mark Rigby, MD, PhD; Royal Children's Hospital Brisbane – Debbie Long, PhD, Anthony Slater, MBBS, Tara Williams, BSN; Royal Children's Hospital Melbourne – Warwick Butt, MBBS, Carmel Delzoppo, BSc, Sophie Syddall, PhD; Sant Joan de Deu Hospital – Iolanda Jordan, PhD, Lluisa Hernandez, MD; St. Louis Children's Hospital – Rachel Jacobs, BA, Philip C. Spinella, MD, Cindy Terrill, CCRP; Starship Children's Hospital – John Beca, MBChB, Tracey Bushell, BA, Miriam Rea, BN, Claire Sherring, RN; Stony Brook University Medical Center – Kathleen Culver, DNP, Margaret Parker, MD; Sydney Children's Hospital – Gary Williams, MBBS, Janelle Young, MApSc, MPH; University Hospital of Salamanca – Mirella Gaboli, MD, PhD, Pedro Gomez de Quero, MD; University of California at San Francisco – Victoria Lo, BS, Anil Sapru, MD, MAS; University of Rochester Medical Center – Jill M. Cholette, MD, L. Eugene Daugherty, MD; University of Virginia – Robin Kelly, RN, Douglas Willson, MD; Women's and Children's Hospital, Adelaide – Georgia Letton, BA, Michael Yung, MD; Yale-New Haven Children's Hospital – Edward Vincent S. Faustino, MD, Kim Marohn, MD, Veronika Northrup, MPH, Li Qin, PhD, Sree Pemira, MD, Joana Tala, MD, Elyor Vidal, MPH
Funding Source: National Heart Lung Blood Institute Education Grant R25 HL096240 (Onyeama); National Institutes of Health CTSA Grants Numbers UL1 TR000142 and KL2 TR000140 (Faustino) and American Heart Association Award Number 14CRP20490002 (Faustino)
Footnotes
Conflicts of Interest: None declared by all authors.
Copyright form disclosures:
Dr. Hanson received support for travel from the American Society of Hematology (Travel expenses to meetings for planning Guidelines for Pediatric Thrombosis). Her institution received grant support from Bristol Meyers Squibb (Pharm Phase 2 trial for Apixiban, oral anticoagulant) and the National Institutes of Health (NIH) (AGE OF BLOOD IN CHILDREN IN PEDIATRIC INTENSIVE CARE UNITS). Ms. Onyeama received payment for Summer of Jun-Aug 2014 for preparation of poster and abstract. Ms. Dasgupta disclosed that she does not have any potential conflicts of interest. Dr. Hoffmann received support for article research from the NIH, Wellcome Trust, and Research Councils UK (RCUK). Dr. Faustino received support for article research from the NIH. His institution received grant support from the NIH and from the American Heart Association.
References
- 1.Monagle P, Chalmers E, Chan A, et al. Antithrombotic therapy in neonates and children: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. (8th Edition) 2008;133(6 Suppl):887S–968S. doi: 10.1378/chest.08-0762. [DOI] [PubMed] [Google Scholar]
- 2.Shah P, Shah V. Continuous Heparin Infusion to Prevent Thrombosis and Catheter Occlusion in Neonates with Peripherally Placed Percutaneous Central Venous Catheters. Cochrane Database Systematic Review. 2008;(2) doi: 10.1002/14651858.CD002772.pub3. Art. No. CD002772. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Shah P, Kalyn A, Satodia P, et al. A Randomized, Controlled Trial of Heparin Versus Placebo Infusion to Prolong the Usability of Peripherally Placed Percutaneous Central Venous Catheters (PCVCs) in Neonates. The HIP (Heparin Infusion for PCVC) Study American Academy of Pediatrics; 2007. pp. e284–e291. [DOI] [PubMed] [Google Scholar]
- 4.Uslu S, Ozdemir H, Comert S, et al. The effect of Low-dose Heparin on Maintaining Peripherally Inserted Percutaneous Central Venous Catheters in Neonates. Journal of Perinatology. 2010;30:794–799. doi: 10.1038/jp.2010.46. [DOI] [PubMed] [Google Scholar]
- 5.Birch P, Ogden S, Hewson M. A Randomised,Controlled Trial of Heparin in Total Parenteral Nutrition to Prevent Sepsis Associated with Neonatal Long lines: the Heparin in Long Line Total Parenteral Nutrition (HILLTOP) trial. Arch Dis Child Fetal Neonatal Ed. 2010;95:F252–F257. doi: 10.1136/adc.2009.167403. [DOI] [PubMed] [Google Scholar]
- 6.Skinner R, Koller K, McIntosh N, et al. Prevention and Management of Central Venous Catheter Occlusion and Thrombosis in Children with Cancer. Pediatr Blood Cancer. 2008;50:826–830. doi: 10.1002/pbc.21332. [DOI] [PubMed] [Google Scholar]
- 7.Lee OKE, Johnston L. A Systematic Review for Effective Management of Central Venous Catheters and Catheter Sites in Acute Care Paediatric Patients. Worldviews on Evidence-Based Nursing. 2005;2:4–13. doi: 10.1111/j.1524-475X.2005.04085.x. [DOI] [PubMed] [Google Scholar]
- 8.Monagle P, Chan AKC, Goldenberg NA, et al. Antithrombotic Therapy in Neonates and Children: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141:e737S–e801S. doi: 10.1378/chest.11-2308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shah PS, Shah N. Heparin-bonded Catheters for Prolonging the Patency of Central Venous Catheters in Children. Cochrane Database of Systematic Reviews. 2014:CD005983. doi: 10.1002/14651858.CD005983.pub2. [DOI] [PubMed] [Google Scholar]
- 10.Veldman A, Nold MF, Michel-Behnke I. Thrombosis in the Critically Ill neonate: Incidence, Diagnosis, and Management. Vascular Health and Risk Management. 2008;4(6):1337–1348. doi: 10.2147/vhrm.s4274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.de Neef M, Heijboer H, van Woensel JBM, et al. The Efficacy of Heparinization in Prolonging Patency of Arterial and Central Venous Catheters in Children: a Randomized Double-Blind Trial. Pediatr Hematol Oncol. 2002;19:553–560. doi: 10.1080/08880010290097404. [DOI] [PubMed] [Google Scholar]
- 12.Cesaro S, Tridello G, Cavaliere M, et al. Prospective, Randomized Trial of Two Different Modalities of Flushing Central Venous Catheters in Pediatric Patients With Cancer. American Society of Clinical Oncology. 2009;12:2059–2065. doi: 10.1200/JCO.2008.19.4860. [DOI] [PubMed] [Google Scholar]
- 13.Randolph AG, Cook DJ, Gonzales CA, et al. Benefit of heparin in peripheral venous and arterial catheters: systematic review and meta-analysis of randomized controlled trials. Br Med J. 1998;316:969–975. doi: 10.1136/bmj.316.7136.969. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kamala F, Boo N, Cheah F, et al. Randomized Controlled Trial of Heparin for Prevention of Blockage of Peripherally Inserted Central Catheters in Neonates. Acta Paediatr. 2002;91:1350–1356. doi: 10.1111/j.1651-2227.2002.tb02833.x. [DOI] [PubMed] [Google Scholar]
- 15.Faustino EV, Hanson S, Spinella PC, et al. Prophylaxis Against Thrombosis Practice (PROTRACT) A Multinational Study of Thromboprophylaxis Practice in Critically Ill Children. Crit Care Med. 2014;42(5):1232–40. doi: 10.1097/CCM.0000000000000147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Faustino EV, Patel S, Thiagarajan RR, et al. Survey of Pharmacologic Thromboprophylaxis in Critically Ill Children. Crit Care Med. 2011;39(7):1773–8. doi: 10.1097/CCM.0b013e3182186ec0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.López-Briz E, Ruiz Garcia V, Cabello JB, et al. Heparin versus 0.9% Sodium Chloride Intermittent Flushing for Prevention of Occlusion in Central Venous Catheters in Adults. Cochrane Database of Systematic Reviews. 2014;(10) doi: 10.1002/14651858.CD008462.pub2. Art. No.CD008462. [DOI] [PubMed] [Google Scholar]
- 18.Clarke M, da Cruz E, Koehler J, et al. A Multicenter Survey of Heparin Prophylaxis Practice in Pediatric Critical Care. J Intensive Care Med. 2011;26(5):314–317. doi: 10.1177/0885066610392501. [DOI] [PubMed] [Google Scholar]