Venous thromboembolism: exploring incidence and utility of screening in individuals with dual diagnosis

Abstract

Objectives

The development of venous thromboembolism (VTE) is a common complication following spinal cord injury (SCI) and brain injury (BI), leading to significant morbidity and mortality. The purpose of this study was to explore the incidence of VTE in patients with the dual diagnosis (DD) of SCI and concomitant BI using ultrasonography.

Design

Retrospective study.

Setting

Acute rehabilitation hospital.

Participants

Thirty-one individuals admitted for DD rehabilitation who were routinely screened for VTE with ultrasound on admission.

Interventions

Not applicable.

Main Outcome Measures

Retrospective chart review was performed to identify whether patients were found to have DVT or PE at the following three time points: in acute care prior to admission to rehabilitation, at time of admission diagnosed via screening examination, and after admission to rehabilitation during the inpatient stay via post screening examinations. Retrospective chart review was also performed to identify incidence of bleeding.

Results

67.7% of individuals were found to have DVTs at any timepoint. Of these DVTs, 22.6% were identified in acute care, 48.4% on admission to rehabilitation, and 16.1% during the course of rehabilitation stay. Of those who were placed on therapeutic anticoagulation due to admission diagnosis of VTE, 25% developed recurrent DVT and 12.5% had bleeding complications. No cases of PE were observed in this study population.

Conclusions

This study found a high incidence of DVT for the DD population at all three timepoints with a high proportion identified via screening ultrasonography on admission to rehabilitation. Further research is needed to investigate the incidence of VTE and utility of screening ultrasonography in this population.

Introduction

The development of venous thromboembolism (VTE) is a common complication of spinal cord injury (SCI) and brain injury (BI), leading to significant morbidity and mortality in these populations [1–4]. BI carries an increased risk for VTE primarily due to the long period of immobilization during comatose states and debilitating trauma, delays in starting VTE prophylaxis after intracranial hemorrhage, and a hypercoagulable state due to elevated levels of pro-coagulants [5]. Similarly, SCI patients are at risk for VTE due to intimal injury, hypercoagulability, and immobilization leading to stasis [3]. The development of VTE leads to the increased risk of complications including venous incompetence, pressure injuries, chronic leg swelling, thrombophlebitis, post thrombotic syndrome, and most importantly pulmonary embolism (PE) [2, 4]. In patients who have dual diagnosis (DD), having both SCI and BI, VTE may present as extensive deep vein thrombosis (DVT) or PE since leg swelling commonly follows SCI and patients often are unable to report leg pain due to impaired sensation from their SCI or impaired cognition from their BI [4].

The incidence of DVT in the acute stage of SCI (the first three months) has been reported between 9–100% with most occurring in the first 2–3 weeks post injury [3]. We have previously reported an incidence of 50.4% diagnosis of SCI patient diagnosed with DVT after the inception of injury through the acute rehabilitation course [6]. Of those diagnosed with DVT, 69.6% were identified on screening examination upon entrance into acute inpatient rehabilitation [6]. Previous studies of DVT incidence rates in the acute phase following BI have ranged from 20–55% [7]. Data published by this group of researchers identified that there was a 51.7% incidence of DVT in this cohort, with 69.1% of those DVT’s identified upon screening ultrasound upon entry into acute inpatient rehabilitation [8]. To date, there is no study evaluating the DVT rate in the DD population. The clinical experience of brain injury and SCI physicians would suggest that the rate of DVT in the DD population is high, and the utility of routine screening examination with ultrasound remains untested. Routine screening and diagnosis of VTE is achieved most commonly by duplex ultrasonography, however practice management guidelines for the prevention of VTE in individuals with SCI and trauma only recommend screening ultrasonography for individuals with SCI and BI who are symptomatic for VTE [4, 9, 10]. There are no established guidelines for the DD patient.

Given the high incidence of VTE in patients with SCI and BI, the current guidelines require further investigation. Early diagnosis is clinically important to reduce morbidity and mortality associated with VTE, which remains high. The purpose of this study was to assess the incidence of DVT at three different time points (in the acute care hospital, on admission to acute inpatient rehabilitation, and during inpatient rehabilitation). This study also investigated trends in incidence of PE events and the risk of bleeding complications in patients on therapeutic anticoagulation (AC) and explores the value of screening duplex ultrasonography on admission to rehabilitation in these two diagnostic categories.

Methods

Subjects

Patients who were consecutively admitted to a rehabilitation facility from January, 2016 through December, 2017 were screened for inclusion in the study. All 31 patients admitted during this period met the inclusion criteria of ages 15–99 years and primary diagnosis of DD. DD was defined as having a spinal cord injury and a moderate to severe brain injury as identified by international classification of diseases, ninth revision, coding. This was a retrospective chart review, which was approved by the Institutional Review Board.

Procedures

All individuals admitted with DD were screened with duplex ultrasonography for upper and lower extremity VTEs within three days of admission as part of standard admission practice for all patients at this rehabilitation hospital. Computerized tomography angiography (CTA) testing for PEs was obtained if clinically indicated by the medical team. Retrospective chart review was performed to identify whether patients were found to have DVT or PE at the following time points: (1) in the acute care hospital prior to admission to rehabilitation by chart review of available outside hospital records and/or history and physical on admission to inpatient rehabilitation; (2) at the time of admission to rehabilitation via screening examination by duplex ultrasonography; and (3) after admission to rehabilitation during the inpatient stay via clinically indicated testing by duplex ultrasonography for suspected DVTs or CTA for suspected PEs. Symptoms prompting workup with CTA included shortness of breath, tachycardia, and hypoxia. A new VTE was defined as a VTE not previously diagnosed at a prior time point. If a VTE was identified in the same location (extremity and/or major vein) as at a previous time point, the VTE was presumed to be the same VTE. If a VTE was identified in the same location as a prior time point but appeared larger and/or encompassed more veins, the VTE was presumed to be propagation of a prior VTE, and was not counted as a new VTE.

An additional retrospective chart review was performed on individuals with a diagnosed DVT on admission to rehabilitation. This chart review was performed to investigate bleeding complications of AC and to assess prophylactic versus treatment dose of AC at each time point. Prophylactic dose of AC was defined by thromboprophylaxis medication guidelines as was treatment dosing [11]. The guidelines were consistent for heparinoids, factor Xa inhibitors, and direct oral anticoagulants.

Measures

Data were collected from retrospective chart review of medical records and from the rehabilitation hospital’s electronic medical record. Demographic data included age at injury onset, admission body mass index (BMI), gender, race/ethnicity, and tobacco use in the past year. Injury characteristics included injury etiology, number of days from injury onset to rehabilitation hospital admission, length of stay (LOS) in rehabilitation, Rancho Score at admission, and initial Glasgow Coma Score (GCS). Comorbidity data included whether the individual had a surgery while at the rehabilitation hospital, a pre-injury history of cancer, pre-injury history of DVT, and pre-injury history of PE.

The acute care hospital course data included history of a neurosurgical procedure, history of hypercoagulable state, long bone and other large bone fracture, and inferior vena cava filter placement. A neurosurgical procedure was defined as placement of a drain for a cerebrospinal fluid leak, craniotomy, craniectomy, placement of a monitor, bolt, external ventricular device (EVD), ventriculoperitoneal (VP) shunt, spinal column surgery, or spinal cord related surgery. A long bone fracture was defined as fracture of the femur, tibia, and/or humerus. Fracture of the acetabulum and/or sacrum was included as supported by evidence from studies evaluating the incidence and co-variables associated with VTE [12].

To evaluate the incidence of VTEs and bleeding complications during the rehabilitation hospital course, the following were investigated: duplex ultrasonography screening results at time of admission, clinically indicated duplex ultrasonography results from suspected DVTs during the rehabilitation stay, and CTA or ventilation-perfusion scan for suspected PE. Surgery during rehabilitation was defined as (1) procedures for which the individual was taken to the operating room and (2) procedures that would have required AC to be held for over 24 h. Bleeding complications also were identified and included gastrointestinal bleeds and muscular hematomas identified on computed tomography scans.

Statistical analysis

All analyses were performed using SAS v9.4. The demographic and injury-related characteristics were summarized for the entire DD sample. Continuous variables were described using means and standard deviations, while medians and interquartile ranges were used for any skewed variables. Frequency counts and percentages were provided for categorical variables.

Incidence rates were calculated for all DVT and PE variables collected in this study. These rates were provided for those who were diagnosed with a DVT at any time point. Incidence rates also were determined for the data collected during the second chart review for those individuals who had a DVT identified at admission to rehabilitation, were not on a therapeutic dose of AC at an acute care hospital, and were placed on a therapeutic dose of AC on admission to rehabilitation.

Results

This sample had an average age of 41(SD = 15.8) years and an average BMI of 28 (SD = 5.7). The majority were male (77%), non-Hispanic white (87%), and did not use tobacco (76%). The average LOS in rehabilitation was 93 days (range 28 to 267). Days from injury to hospital admission had a median of 26 days (range 10 to 236). Most of the DD sample had a vehicular injury etiology (55%). More than 93% of the sample did not have a history of a DVT or PE prior to the injury. Twenty-three patients had undergone a neurological procedure at time of injury (74%) and nine had undergone a surgery during their stay at the rehabilitation hospital (29%). Table 1 illustrates all demographic, injury, and comorbidity characteristics for the sample.

Table 1.

Demographic and injury characteristics.

	Dual Diagnosis (N = 31)
	N	Mean (SD)
Age at Injury Onset (Years)	31	41.1 (15.8)
BMI (Admission)	31	27.6 (5.7)
Days from Injury Onset to Hospital Admissiona	31	26 (17–36)
LOS in Rehabilitation	31	92.5 (45.5)
Initial GCS	5	8.2 (5.9)
	N	%
Gender
Female	7	22.6%
Male	24	77.4%
Race/Ethnicity
Non-Hispanic White	27	87.1%
Other	4	12.9%
Tobacco Use in the Past Year
Yes	6	24.0%
No	19	76.0%
(Missing)	6
Rancho Score (Admission)
2	1	16.7%
3	0	0.0%
4	3	50.0%
5	0	0.0%
6	1	16.7%
7	1	16.7%
8	0	0.0%
9	0	0.0%
(Missing)	25
Motor Completeness		32.0%
C1-T1, AIS A or B	8	24.0%
T2-S2, AIS A or B	6	40.0%
C1-T1, AIS C or D	10	4.0%
T2-S2, AIS C or D	1
Missing	6
Injury Etiology
Vehicular	17	54.8%
Fall	5	16.1%
Sports	3	9.7%
Violence	4	12.9%
Medical	2	6.5%
Other	0	0.0%
History of Cancer
Yes	1	3.2%
No	30	96.8%
Pre-injury History of PE
Yes	1	3.2%
No	30	96.8%
Pre-Injury History of DVT
Yes	2	6.5%
No	29	93.5%
History of Hypercoaguable State
Yes	0	0.0%
No	31	100.0%
History of Neurosurgical Procedure
Yes	23	74.2%
No	8	25.8%
History of Monitor/Bolt/EVD/VP Shunt
Yes	3	9.7%
No	28	90.3%
History of Drain for CSF Leak
Yes	1	3.2%
No	30	96.8%
Craniotomy/Craniectomy
Yes	2	6.5%
No	29	93.5%
Spinal Column/Cord Related Surgery
Yes	20	64.5%
No	11	35.5%
History of IVC Filter Placement
Yes	7	22.6%
No	24	77.4%
(Missing)	0
Concomitant Long Bone Fracture
Yes	3	9.7%
No	28	90.3%
Surgery during stay at Rehab Hospital
Yes	9	29.0%
No	22	71.0%

SD Standard Deviation, BMI Body Mass Index, LOS Length of Stay, GCS Glasgow Coma Scale, DVT Deep Venous Thrombosis, PE Pulmonary Embolisms, EVD External Ventricular Drain, VP Ventriculoperitoneal, CSF Cerebrospinal Fluid, IVC Inferior Vena Cava.

^aMedian and interquartile range reported.

Twenty-one DD patients (68%) were found to have a DVT in either acute care, admission to rehabilitation, or after admission to rehabilitation. Five of these patients had multiple DVTs (16%) and the majority of DVTs were found on admission to rehabilitation (42%). For the entire DD sample, two (7%) were found to have had a PE at any time point. One patient had a PE in acute care and the other had a PE on admission to rehabilitation. All incidence rates can be seen in Table 2.

Table 2.

Incidence rates for DVT and PE.

	Dual Diagnosis (N = 31)
	N	%
DVT Identified at any Time Pointa	21	67.7%
Acute Care	7	22.6%
Admission to Rehabilitation	15	48.4%
After Admission to Rehabilitation	5	16.1%
Multiple DVTs	5	16.1%
First DVT Identified
Acute Care	7	22.6%
Admission to Rehabilitation	13	41.9%
After Admission to Rehabilitation	1	3.2%
PE Identified at any Time Pointa	2	6.5%
First in Acute Care	1	3.2%
First at Admission to Rehabilitation	1	3.2%
First after Admission to Rehabilitation	0	0.0%

DVT Deep Venous Thrombosis, PE Pulmonary Embolisms.

^aAcute Care, Admission to Rehabilitation, and after admission to Rehabilitation.

In the secondary chart review (Table 3), eight individuals were discovered to have had a DVT on admission to rehabilitation and were placed on a therapeutic dose of AC at rehabilitation. No DD patients had a PE after being started on therapeutic anticoagulation. Two of the eight individuals were found to have recurrent DVTs after being started on therapeutic anticoagulation. In case one, this patient had their anticoagulation paused due to a bleeding complication. This patient developed a thigh hematoma felt to be related to an ill-fitting knee-ankle-foot-orthosis, which resolved spontaneously and did not require a transfusion. In case two, the patient developed a new DVT despite being continued on therapeutic anticoagulation.

Table 3.

Incidence rates from second retrospective review^a.

	Dual Diagnosis (N = 8)
	N	%
PE after Rehab	0	0.0%
DVT after Rehab	2	25.0%
Bleeding Complication	1	12.5%

DVT Deep Venous Thrombosis, PE Pulmonary Embolisms.

^aPatients with a DVT identified on admission to rehabilitation, were not on a therapeutic dose of anticoagulation at acute care, and were placed on a therapeutic dose of anticoagulation at rehabilitation.

Discussion

We aimed to characterize the sample of individuals with DD and their incidence of VTE from acute hospitalization through initial acute rehabilitation, as well as the associated risk of PE and/or bleeding when placed on anti-coagulation. We describe an overall high incidence of VTEs with a high percentage diagnosed via screening duplex ultrasonography upon admission to acute rehabilitation. After treatment with therapeutic anticoagulation, individuals with DVT experienced minimal bleeding complications and, most importantly, no cases of PE were observed.

This is the first paper to describe the incidence of VTE in the dual diagnosis population, however prior SCI and BI literature has reported a wide range of overall incidence of DVTs. The varying in reported incident rate was dependent on various factors including injury characteristics, screening tools, and use of chemoprophylaxis^,[1, 3, 10, 13–15]. Compared to the authors’ prior work investigating the incidence of DVT in BI and SCI populations, there was a lower overall incidence of DVT in the BI (51%) and (50.1%) SCI samples compared with the DD sample (68%) [8]. While further research with a larger sample size would be needed to confirm these findings, this preliminary work supports a trend for an high incidence of DVT in the DD population. The high incidence rate of DVT in the DD population may be attributed to the unique contributions of SCI and BI, namely immobilization after SCI and the delay in starting chemo prophylaxis after intracranial hemorrhage in BI. Furthermore, both BI and SCI involve intimal injury and hypercoagulability that increases risk for VTE [1].

Our preliminary findings suggest that routine screening may identify a high incidence of DVTs in this population, however prospective studies with larger sample sizes are needed to further elucidate these trends. Given that there are no guidelines for the dual diagnosis population, a larger study might provide a more solid basis for clinical care in terms of screening duplex ultrasonography in this population. Screening duplex ultrasonography is a low-risk, cost-effective procedure that has the potential to reduce a major cause of morbidity and mortality in individuals with SCI and concomitant BI [16].

Limitations

This study has several limitations that should be considered. Most importantly, this was a retrospective review with a relatively small sample size. Given that this is the only report of VTE risk in the DD population, larger prospective studies are required. Limitations also included a high degree of missing International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) data to further classify the individuals’ spinal cord injuries. However, the lack of ISNSCI classification for some patients was likely due to the severity of their brain injuries, which impaired their ability to participate in the ISNSCI testing process.

Conclusion

This study found a high incidence of DVT for the DD population at all three timepoints with a high proportion found via screening ultrasonography on admission to rehabilitation. Given the trends identified in this preliminary study, further research is warranted to investigate the incidence of VTE and utility of screening ultrasonography in this population.

Author contributions

CD was responsible for study design, extracting the data, interpreting study results, and writing the report. AP was responsible for study design, interpreting study results, and writing the report. MS was responsible for analyzing data, interpreting results, and writing the report. SC was responsible for extracting the data and writing the report. MM was responsible for interpreting study results and writing the report. DC was responsible for interpreting study results and writing the report. RA was responsible for study design and writing the report. JF was responsible for study design and writing the report. JB was responsible for study design, extracting the data, interpreting study results, and writing the report.

Funding

This work was supported by the Craig Hospital Foundation under grant number 2643.

Data availability

Data are available upon request from the corresponding author.

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

This study was in accordance with the Declaration of Helsinki. This study was approved by the hospital institutional review board. Informed consent was not obtained as this was a retrospective review.