Use of Institutional Clinical Care Guidelines to Determine HA-VTE Risk Level for Consideration of Thromboprophylaxis in Hospitalized Children

Background – Hospital-acquired venous thromboembolism (HA-VTE), comprised of deep vein thrombosis and pulmonary embolism, is associated with significant mortality and morbidity and the overall incidence is rising, leading to its recognition as a leading quality and patient safety concern in pediatric hospitals. Risk-based prophylactic strategies are used in adults but a paucity of high quality evidence exists regarding the safety and efficacy of these interventions in the pediatric population.  Defining an objective assessment of VTE risk to guide standard approaches to the use of mechanical and/or pharmacologic prophylaxis (in the absence of contraindications) is a desirable strategy to decrease HA-VTE incidence while minimizing potential risks associated with prophylaxis.

Methods- With Institutional Review Board approval, we retrospectively evaluated electronic health record data from 39 patients age 0-21 years (inclusive) who developed VTE more than 48 hours after admission to Children’s Hospital Colorado (Aurora, CO, USA) from January 1, 2014 – December 31, 2014. We identified purported HA-VTE cases via ICD-9 discharge diagnosis codes, then two different reviewers cross-referenced and validated with pharmacy and radiology records.

We evaluated the HA-VTE patient’s VTE risk scores (high, moderate, or low) using our institutional VTE prevention clinical care guideline (CCG) at discrete time points prior to clot diagnosis, as well as whether the patient received mechanical and/or pharmacologic thromboprophylaxis.

Objectives: We sought to determine the ability of our CCG to accurately assess HA-VTE risk prior to clot development and to assess the utilization and timing of thromboprophylaxis.

Results:  Of the 39 patients who developed a HA-VTE during the study period, CCG-based risk assessment on the day before the event had identified four (10.3%) patients as low risk, two (5.1%) as moderate risk, and 33 (84.6%) as high risk, as seen in Table 1. This distribution represents an overall increase in risk score compared to those determined on admission when 13 (33.3%) were low risk, 4 (10.3%) were moderate risk, and 22 (56.4%) were high risk. Admission risk scores tended to hold steady or increase during the hospitalization (Figure 1) and only one patient experienced a decrease in risk level prior to HA-VTE diagnosis. HA-VTE patients often had high risk scores as early as seven or even 14 days prior to the event, and 20 (51%) of the patients who developed HA-VTE were high risk during the entire admission. 

Of the HA-VTE patients, 19 (48.7%) were receiving thromboprophylaxis on the day prior to the event. 4 (21.1%) received pharmaceutical type only, 12 (63.2%) received mechanical type only, and 3 (15.8%) received both types. 

Conclusion:  Institutional CCGs can be used to determine HA-VTE risk stratification in hospitalized children, thereby informing decisions regarding initiation of thromboprophylaxis as a preventative strategy in the absence of contraindications. Comparison to a group of age group- and risk level-matched control patients who did not develop HA-VTE is underway and will give appropriate context to the high risk designation as it pertains to HA-VTE incidence and the impact of thromboprophylaxis strategies.

 

Figure 1: Patient-level risk score tracking: Each colored line represents a unique HA-VTE patient’s risk score trajectory during hospitalization. Risk scores upon admission correspond to days represented by negative numbers denoting various days prior to event (Day 0), and all patients have evaluations from 14, 7, and 1 day prior to the event, if applicable.

Table 1: Risk scores prior to HA-VTE diagnosis: Risk scores at various time points during hospitalization

 

Risk Level	Risk Score on Admission	Risk Score 14 Days Prior	Risk Score 7 Days Prior	Risk Score 1 Day Prior	Risk Score Day of VTE
Low	13 (33.3%)	0 (0%)	2 (10.5%)	4 (10.3%)	4 (10.3%)
Moderate	4 (10.3%)	1 (9.1%)	1 (5.3%)	2 (5.1%)	2 (5.1%)
High	22 (56.4%)	10 (90.9%)	16 (84.2%)	33 (84.6%)	33 (84.6%)
Total	39	11	19	39	39