Hospital Acquired Venous Thromboembolism: A Preventability Assessment

Abstract

Background: The American Heart Association has a call to action to reduce hospital acquired venous thromboembolism (HA-VTE) by 20% by the year 2030. There is increasing recognition that quality improvement initiatives for VTE reduction should focus on reducing potentially preventable HA-VTE. The objective of our study was to determine what proportion of HA-VTE events are potentially preventable. Methods: This was a retrospective, single center pilot study of 50 patients with HA-VTE. Seven preventability factors were identified with a focus on VTE prescription and administration. Data were extracted through chart review using a systematic data collection form. The primary endpoint was the proportion of patients with potentially preventable HA-VTE. Descriptive statistics were used. Results: The median age was 66 years with an admission VTE risk level of moderate-high in 94%. Potentially preventable HA-VTE was found in 40% of cases. Missed doses occurred in 29.8% with a median of 2 missed doses and a range of 1 to 20. Patient refusal was the most common reason for missed doses in 71%. Delays in initiation occurred in 12.7%. Sixty percent of those on mechanical prophylaxis only had nonadherence. Conclusion: Forty percent of HA-VTE cases were potentially preventable. Missed doses was the most common preventability factor identified with patient refusal accounting for most missed doses.

Background

Hospital acquired venous thromboembolism (HA-VTE) encompasses both deep vein thrombosis (DVT) and pulmonary embolism (PE) and accounts for more than half of all VTE in the United States. ¹ HA-VTE is associated with an increase in hospital length of stay and mortality with pulmonary embolism being recognized as a leading cause of preventable death among hospitalized patients.^2,3 The American Heart Association (AHA) has recently identified HA-VTE as a national health priority with a call to action to decrease HA-VTE by 20% by 2030. ¹ There is increasing recognition of the need to focus VTE reduction efforts on potentially preventable HA-VTE rather than all HA-VTE. Potentially preventable HA-VTE has been reported to represent up to 70% of all cases of HA-VTE. ² A consensus definition and approach to measuring potentially preventable HA-VTE currently does not exist, Table 1.^4-8

Table 1.

Definition of Preventable Venous Thromboembolism.

Source	Definition
Arnold 7	● Objectively diagnosed deep vein thrombosis or pulmonary emboli that occurred in a setting in which thromboprophylaxis was indicated but was either administered inadequately or not administered at all.
Ma 8	● VTE that occurs in a setting in which thromboprophylaxis is indicated but it is administered inadequately or omitted. ● Adequate prophylaxis defined as consistent with guideline recommended agent, dose and frequency, given continuously for 7 d, until hospital discharge or until ambulatory and initiated within 24 h of clinical event associated with increased thrombosis (eg, surgery, myocardial infarction, or hospitalization).
Haut 5	● VTE occurring in those with suboptimal care defined as prescription failure or dose administration failure. ○ Patients with bleeding receiving sequential compression devices were considered to have defect free care. ○ Patients with upper extremity catheter associated DVT were excluded.
Lau 4	● VTE which develops in association with a VTE prevention process failure. ● Process failures could occur in VTE risk assessment, prescription of appropriate VTE prevention (including mechanical prophylaxis) or the administration of prescribed VTE prophylaxis.
Narayan 6	● VTE occurring in a patient with suboptimal prophylaxis without a contraindication. ● Suboptimal prophylaxis includes low dose, missed dose prior to VTE, suboptimal drug and delayed start defined as more than 24 h from admission.

Lau et al. describes effective VTE prevention as a multifactorial, complex process that addresses 4 key stakeholders ⁴ . Clinicians must ensure appropriate risk assessment is performed, providers must prescribe appropriate therapy considering patient specific factors, patients must accept the therapy and nurses must ensure therapy is administered. Using this framework Lau et al. defined potentially preventable VTE as VTE which develops in association with a VTE prevention process failure. Process failures could occur in VTE risk assessment, prescription of appropriate VTE prevention (including mechanical prophylaxis) or the administration of prescribed VTE prophylaxis. ⁴ The objective of this pilot study was to determine the proportion of HA-VTE events which were potentially preventable within the process steps of VTE prescription and administration.

Methods

Study Design

This was a retrospective cohort pilot study of 50 adult patients with HA-VTE conducted to assess the frequency of potentially preventable VTE. Patients were included if they were 18 years or older, admitted without ICD-10-CM code for DVT or PE and were subsequently discharged with a secondary ICD-10-CM diagnosis code for proximal DVT (DEEPVIB) or PE (PULMOID) as outlined in the Patient Safety Indicator 12 (PSI 12) Rate defined by the Agency for Healthcare Research and Quality. ⁹ Exclusion criteria included DVT or PE within 3 months prior to admission, within 24 hours after admission, and acute on chronic thrombosis. The study time period began in January of 2019 and continued until the goal of 50 patients were met. Pre-defined enrollment targets included limiting the inclusion of patients receiving only mechanical prophylaxis with intermittent pneumatic compression (IPC) boots to 10% of the study population and an even distribution of medical and surgical patients. The study was approved by the Institutional Review Board under expedited review and the requirement for informed consent was waived.

Data Collection and Analysis

Manual data extraction was performed using a systematic data collection instrument which was piloted prior to study initiation. Data extraction occurred for the index admission of the VTE event. Extracted data included patient demographics, provider documented admission VTE risk level, objective confirmation of DVT and/or PE events, intensive care unit status, pharmacologic prophylaxis ordered, time from admission (or from surgery) to administration of first dose of pharmacologic prophylaxis, frequency and assessment of pharmacologic prophylaxis within 7 days prior to VTE suspicion, and adherence to mechanical prophylaxis orders for those receiving IPC only. Extracted data were verified by a second investigator.

A preventability assessment for HA-VTE was created and applied to each patient case. Preventability factors included pharmacologic prophylaxis prescribed with an agent not recommended in the VTE prevention guidelines,^10,11 missed doses within 7 days preceding VTE suspicion not related to invasive procedures/surgical holds, interruption in therapeutic anticoagulation without the provision of VTE prevention, dose of pharmacologic prophylaxis too low based on renal function (enoxaparin only), delay in pharmacologic prophylaxis defined as more than 24 hours from admission or surgery in the absence of bleeding, pharmacologic prophylaxis addressed in progress notes but never ordered, and patients receiving only mechanical prophylaxis with IPC with devices off for more than 4 hours within the preceding 72 hours period. ¹² Patients with one or more preventability factors were considered to have potentially preventable HA-VTE. The primary endpoint of the study was the proportion of patients with potentially preventable HA-VTE.

Descriptive statistics were performed using Microsoft Excel and PRISM GraphPad. As data were not all normally distributed, non-parametric statistics were used.

Results

A total of 90 patients were screened from January 1st 2019 through August 31st 2019 to enroll 50 patients with HA-VTE in alignment with predefined enrollment goals. Of the 25 surgical patients, 40% were orthopedic surgery and 24% were cardiac surgery. Patient characteristics are outlined in Table 2. Forty-five patients received pharmacologic prophylaxis. Two patients of the 5 in the mechanical prophylaxis only group were prescribed but never received pharmacologic prophylaxis and are assessed for the processes of prescription and administration of pharmacologic prophylaxis (total patient population of 47). The most common initial VTE prevention agent prescribed was unfractionated heparin in 87.2% of patients.

Table 2.

Patient Characteristics.

Patient characteristic	N = 50 a
Male, n (%)	30 (60%)
Surgical, n (%)	25 (50%)
Median age, y (IQR)	66 (55.5-79)
Median actual body weight, kg (IQR)	85.5 (69.5-100.5)
Patients with BMI above 40 kg/m2, n (%)	5 (10%)
In ICU prior to VTE event, n (%)	26 (52%)
Initial VTE prevention regimen b , n (%)
Unfractionated Heparin 5000 units SQ q 12 h	4/47 (8.5%)
Unfractionated Heparin 5000 units SQ q 8 h	37/47 (78.7%)
Enoxaparin 40 mg Q 24 h	2/47 (4.3%)
Enoxaparin 30 mg Q 12 h	1/47 (2.1%)
Aspirin 81 mg BID	2/47 (4.3%)
Enoxaparin 1 mg/kg/d	1/47 (2.1%)
VTE event, n (%)
DVT only	17 (34%)
PE only	16 (32%)
DVT and PE	17 (34%)

^aUnless otherwise shown.

^bTwo of the 5 patients in the mechanical prophylaxis only group were prescribed heparin and refused all doses and a single patient was on therapeutic anticoagulation and missed a dose.

The provider admission VTE risk level determined using clinical assessment was low for 3 patients and moderate to high for 47/50 (94%). One medical patient was low risk on admission however was subsequently transferred to a progressive care area without a documented VTE risk assessment upon transfer. Two surgical patients with a documented low risk level included a knee arthroplasty and spinal decompression. There were no cases of patients with a moderate to high level VTE risk without a prescription of VTE preventative therapy.

Overall, the primary endpoint of potentially preventable HA-VTE occurred in 40% of patients with an even distribution between surgical and medical patients, Table 3. Four of 20 patients (20%) had 2 preventability factors. All patients were prescribed pharmacologic prophylaxis endorsed by guidelines. Missed dose(s) of pharmacologic prophylaxis occurred in 14/47 (29.8%) patients. The median number of missed doses was 2 with an interquartile range of 1 to 3 and a range of 1 to 20 doses within 7 days of VTE suspicion. Three of the 14 patients had holds for an invasive procedure/surgery which were not counted in the number of missed doses as these holds represent standard of care. Patient refusal was the most common reason for missed doses in 10 of 14 patients, 71.4%. Other reasons for missed doses included patient being off the floor, single dose hold orders without notation of reason, and absence of documentation/unable to determine.

Table 3.

Potentially Preventable HA-VTE.

Preventability factor	Number of patients	Comments
Prescribed agent not recommended within VTE prevention guidelines	0%
Missed doses of pharmacologic prophylaxis within 7 d prior to VTE suspicion	14/47, 29.8%	Includes two patients with mechanical prophylaxis only due to refusal of all injections
Interruption in therapeutic anticoagulation without VTE prophylaxis	0%
Dose of pharmacologic prophylaxis too low based on renal function (enoxaparin only)	0%
Delay in initiation of pharmacologic prophylaxis without presence of bleed; (more than 24 h from admission for medical or more than 24 h from surgery)	6/47, 12.8%	Includes 3 medical patients and 3 surgical patients, two of whom had SQ heparin noted in progress notes but not ordered until subsequent day
Pharmacologic prophylaxis discussed in notes but not ordered	1/47, 2.1%
Patients with mechanical prophylaxis only with intermittent pneumatic compression devices off for more than 4 h in the 72 h period preceding confirmed VTE	3/5, 60%	Two of these 3 patients also had refusal of pharmacologic prophylaxis

Six patients prescribed pharmacologic prophylaxis (12.8%) experienced a delay in initiation and were classified as potentially preventable HA-VTE. A delay occurred in 4 additional patients who were not classified as having preventable HA-VTE as bleeding was present and mechanical prophylaxis was provided. Three of 5 patients receiving only mechanical prophylaxis had IPC off for more than 4 hours in the 72-hour period preceding VTE suspicion.

Discussion

Potentially preventable VTE was assessed during the process steps of prescribing and administration of VTE preventative therapies. We found potentially preventable HA-VTE in 40% of patients. Previous studies have reported rates of potentially preventable HA-VTE of 10% to 70%.^2,5,6 The large variation in reported rates is related to the lack of a standardized definition and approach for measuring preventable HA-VTE. Missed doses of a pharmacologic agent for VTE prevention comprised the largest contributing factor for HA-VTE occurring in more than 25% of patients. Patient refusal was the primary reason for missed doses with one patient missing up to 20 doses in the 7 days preceding the VTE event. Missing a single dose of pharmacologic prevention has been associated with increased VTE risk.^5,13,14 Nursing beliefs and attitudes, gaps in knowledge and unit culture can all contribute to missed doses. ¹⁵ Evidence suggests nursing education, patient education and enhancing skills for the nurse-patient communication can significantly reduce missed doses of VTE prevention.^16-18Alternative therapies could also be considered. In certain patients, a low molecular weight heparin can be administered once daily decreasing the number of injections required and potentially reducing missed doses. Oral anticoagulants may also have a role in preventing missed doses. A survey of hospitalized patients on preferences for pharmacological VTE prevention found that 60% of patents had preference for an oral route of administration. ¹⁹

Delays in initiation of pharmacologic prophylaxis resulting in potentially preventable VTE occurred in both medical and surgical patients (12.8%). Four additional patients experienced delays in initiation however these patients were not classified has having potentially preventable HA-VTE as active bleeding was present. The exclusion of patients with contraindications to pharmacologic prophylaxis who are receiving mechanical prophylaxis is consistent with other studies and is representative of clinical practice. ⁵

As a consensus definition of preventable HA-VTE does not exist, this study included the assessment of factors not considered in prior studies.^5,6,8 Although not identified in this study, we reviewed patients for interruption in therapeutic anticoagulation without the provision of VTE prevention therapy. Cases of this were recognized during the height of the COVID pandemic and thus this factor was included. A single case of VTE prevention therapy discussed in progress notes without a corresponding medication order was found. This missed order in addition to delays in initiation in 2 additional patients (Table 3) suggest an opportunity for improving the alignment of stated care within progress notes to actual provider orders. Non-compliance with IPC was also identified as a preventable factor in those receiving mechanical prophylaxis alone. The AHA has identified the need for consensus on the definition of preventable VTE. ¹ We support the inclusion of interruption in therapeutic anticoagulation without VTE prevention, failure to order or delay in ordering of VTE prevention and noncompliance with mechanical prophylaxis as components of preventable HA-VTE.

The appropriate dosing of pharmacologic therapy for VTE prevention in obese and morbidly obese patients is an area which requires further study. In our pilot study, 10% were morbidly obese with BMIs ranging from 41.4 to 61.7 kg/m² and all received standard dose prophylactic therapy. Although guidance from professional organizations support increased dosing for VTE prevention in obesity this issue is not addressed within the VTE prevention guidelines.^10,11,20,21 The addition of this issue to the AHA research agenda for VTE prevention in hospitalized patients is recommended.

During the conduct of this pilot study the Joint Commission (JC) implemented a new optional quality measure, VTE-6, Hospital Acquired Potentially-Preventable Venous Thromboembolism. Preventable events are determined by evaluating patients not receiving pharmacological or mechanical prophylaxis prior to the suspicion of VTE. ²² The results of our study suggest VTE-6 will significantly underestimate the frequency of potentially preventable HA-VTE. None of the preventable VTE cases identified in this study would be identified by the VTE-6 measure as all patients received some form of VTE prevention during hospitalization prior to VTE suspicion. The JC VTE-6 measure excludes patients if there is documentation of patient or family member refusal of injections. Our study identified injection refusal as the most common reason for missed doses leading to potentially preventable VTE. Exclusion of this population will contribute to the under-recognition of preventable HA-VTE. Lau et al suggests the ideal VTE outcome measure would capture the smallest error at any stage of the VTE process including failure to receive a single dose of prescribed VTE prevention therapy. ⁴ As patient refusal of injection(s) can be modified with active intervention, we believe such patients should be included in the measure of potentially preventable HA-VTE.

This pilot study is limited by its retrospective nature, small sample size and lack of a control group. Despite these limitations our study provided important findings which contribute to the body of literature on potentially preventable HA-VTE. These new definitions will assist in capturing additional process failures. Another potential limitation of this study was the use of PSI-12 criteria for patient screening. The required chart review however ensured all included patients had confirmed DVT or PE or both.

Conclusion

In conclusion, in our retrospective study of patients with HA-VTE found 40% of patients had potentially preventable VTE. Missed doses of pharmacologic prophylaxis was the most common factor identified. We propose new factors for consideration in the definition of potentially preventable HA-VTE.