Analysis of Readmissions Due to VTE—Using Hospital Data to Improve VTE Prophylaxis Compliance: A Quality Improvement Project

Abstract

Venous thromboembolism (VTE) is a preventable cause of mortality and morbidity. We performed a retrospective analysis of patient records to identify those readmitted with a diagnosis of VTE within 6 months of the primary admission. The records were evaluated to see whether thromboprophylaxis had been provided to patients at high risk for VTE. A total of 360 hospital encounters between August 1, 2018, and August 31, 2019, with VTE, 57 (16%) encounters were readmissions with a primary diagnosis of deep vein thrombosis within 180 days of their primary stay. A high proportion (44%) of these readmissions were within the first 30 days. 3% (n = 9) of patients developed pulmonary embolism; 35 (61%) did not receive thromboprophylaxis on their primary stay. Thromboprophylaxis is often not utilized appropriately in healthcare settings. Our study showed substantial incidence of hospital readmissions due to VTE which is consistent with prior studies conducted globally. A more stringent adherence to the protocol along with risk stratification may lower rates of VTE admission and reduce associated morbidity and mortality.

Introduction

Venous thromboembolism (VTE) is a major preventable cause of mortality, morbidity, and cost particularly in an in-hospital setting. ¹ VTE is a condition where a thrombus forms in a vein which may embolize. It consists of deep vein thrombosis (DVT) that occurs in the deep veins of the body, most commonly of the lower limbs, and is a cause of more than 90% of VTE events. Venous thrombosis may also result in pulmonary embolism (PE) where part of the thrombus is displaced and lodged in a pulmonary artery. They may occur in situ or as a severe complication of DVT. PE accounts for long-term morbidity associated with VTE. Both DVT and PE are significant contributors to preventable in-hospital mortality. ¹

In tertiary care hospitals based in Pakistan studies have shown the incidence of VTE, DVT, or PE to range between 2.6% and 12.8%.^2-5 Despite carrying a high burden of cost as well as morbidity and mortality preventable through either pharmacological or mechanical methods, DVT/VTE is still often overlooked in our nation either due to a lack of knowledge or a lack of adherence to a defined thrombophylactic protocol. ⁶

The American College of Chest Physicians (ACCP) recommends thromboprophylaxis in at-risk medical patients as it has been shown to reduce the incidence of VTE. ⁷ The ACCP guidelines advocate that acutely ill medical patients admitted to the hospital with congestive heart failure, severe lung disease/chronic obstructive pulmonary disease, or those who are confined to bed and have one or more additional risk factors (including active cancer, previous VTE, sepsis, acute neurologic disease, or inflammatory bowel disease) receive pharmacological prophylaxis with low-molecular-weight heparin, low-dose unfractionated heparin (UFH), or fondaparinux. Similarly, guidelines are in place for surgical patients as well to reduce the risk of VTE post-surgery.

In Pakistan, regular use of thromboprophylaxis to prevent VTE is uncommon, and there are lack of data regarding the prevalence or incidence of VTE. ³ However, larger-scale tertiary hospitals in the country do have protocols in place for VTE prophylaxis but they are not standardized. For example, the current protocol at the tertiary care hospital where this study was carried out requires the intern and resident physicians to begin by assessing the risk factors for a VTE event in patients followed by risk factors for bleeding. After the initial risk assessment, pharmacological VTE prophylaxis that is consistent with the patient's overall risk status is to be offered.

We conducted a retrospective observational study with the aim of determining the incidence of patients who were readmitted due to a secondary VTE event after their primary admission. This was done to assess points where the current thromboprophylaxis protocol in place at the hospital could have failed and establish a baseline of readmission rates due to VTE.

With these data, we can identify barriers to administering VTE prophylaxis in both medical/surgical care, and any limitations in the current protocol and help disseminate, update, and improve current methods through Quality Improvement (QI) initiatives. One study by Restrepo et al showed a significant improvement in compliance, increasing the correct application of graduated compression stockings by nurses from 51% to 67%. ⁸ A more stringent adherence to the protocol would predictably result in lower rates of readmission due to DVT and reduce the preventable increase in morbidity and mortality.

Methods

Data Source

This study was a retrospective cohort conducted on a subset of patients admitted to Aga Khan University Hospital (AKUH), Karachi, Pakistan, between August 1, 2018, and August 31, 2019. The patient population was restricted to those 18 years or older of both male and female genders, were surgical or nonsurgical patients and developed VTE during their stay, or were later readmitted within the next 6 months with a VTE. The medical record number of eligible patients was used to access their clinical information. The study was conducted after the approval of the institutional review board.

The exclusion criteria were patients who were admitted primarily with VTE with no previous stay in the hospital, patients who presented with a thromboembolic event after more than 6 months of their initial stay, and patients who had a previous history of VTE. These patients were excluded on the basis that they would not represent the failure of the VTE prophylaxis protocol.

Data Collection and Analysis

Patient data from the AKUH's internal database were manually reviewed, and relevant information was extracted by 2 of the authors in a nonblinded manner. To ensure accuracy, any uncertainty regarding whether a patient met the inclusion criteria was thoroughly examined and discussed with a third author prior to the finalization of the data.

AKUH primarily has a paper-based health record system, however, through an online portal patient information such as investigations, radiology, hospital encounters, complete medication, and physician orders during their stay and discharge summaries can be accessed. Discharge summaries, medication orders, patient's medical records, and radiological reports were reviewed if they were readmitted within 1 month, and within 2 to 6 months of their previous admission. In the evaluation of VTE hospital readmission rates, it was required that the primary hospital admission be followed by a VTE hospital readmission and that a diagnosis of VTE was made at that hospital admission. Demographic information was collected for all patients. Additional data collected included patient age, sex, length of stay, initial diagnosis, comorbidities, patient mortality, administration of VTE prophylaxis, type of VTE prophylaxis, and start and end dates of prophylaxis. VTE prophylaxis was considered appropriate if it was administered immediately after the patient's admission and risk assessment until the day of their discharge unless they had any contraindications. Data were manually entered into a secure excel sheet, which only the study's investigators had access to. The whole process is summarized in Figure 1. All personal identifiers were removed from the data collected.

Figure 1.

Flow diagram of analysis for patients admitted with venous thromboembolism (VTE).

Results

A total of 360 patients who were admitted with VTE during our study timeline were analyzed and based on the inclusion and the exclusion criteria, 16% of the individual patients (n = 57) met the study criteria.

Patient Demographics

Characteristics of patients readmitted with a VTE event are reported in Table 1. The largest proportion of patients who were readmitted to the hospital for VTE by age group were more than 65 years old (39%). Hypertension was the most prevalent comorbidity (53%). An underlying malignancy was found in 29% of patients in this cohort.

Table 1.

Baseline Demographics of Patients Who Were Readmitted With VTE.

Patients n	57
Mean length of stay days	9
Age group, n (%)
18-34	7 (12%)
35-44	5 (9%)
44-54	8 (14%)
55-64	15 (26%)
65+	22 (39%)
Gender, n (%)
Male	26 (46%)
Female	31 (54%)
Associated comorbidities, n (%)
Hypertension	31 (53%)
Diabetes	20 (35%)
Malignancy	17 (29%)
Chronic Kidney Disease	14 (24%)
Ischemic Heart Disease	6 (10%)
Chronic Atrial Fibrillation	4 (7%)
Gout	1 (2%)
Cardiomyopathy	1 (2%)

Abbreviation: VTE, venous thromboembolism.

Hospital Readmission Rate

Of the total selected patients with post-discharge VTE, 44% (n = 25) were readmitted within 1 month of the primary admission. Only 28% (n = 7) of them had received VTE pharmacological prophylaxis during their primary inpatient stay. Of these patients, 5 (71%) were administered 5000 units of subcutaneous heparin every 8 to 12 h and 2 (29%) were administered 40 mg of daily enoxaparin dose through a subcutaneous route; 72% (n = 18) of the total patients readmitted within 1 month received no pharmacological DVT prophylaxis.

It was observed that 30% (n = 17) of patients were readmitted within a period of 2 to 6 months after their primary hospitalization. Within this category, 35% (n = 6) received pharmacological prophylaxis; with 67% (n = 4) receiving heparin and 33% (n = 2) receiving enoxaparin; 65% (n = 11) did not receive any pharmacological VTE prophylaxis.

Additionally, during their initial stay at the hospital, 26% (n = 15) of patients developed a thromboembolic event. Of these patients, 60% (n = 9) had received pharmacological prophylaxis; with 78% (n = 7) receiving heparin while 22% received enoxaparin; 40% (n = 6) of these patients received no pharmacological thromboprophylaxis. Figure 2 shows the data for pharmacological prophylaxis and no prophylaxis for patients during their primary inpatient stay.

Figure 2.

Pharmacological prophylaxis given to patients during their primary inpatient stay stratified by when deep vein thrombosis (DVT) was developed.

Out of all the individual patients who were diagnosed with DVT, 21% (n = 12) of the patients died and 16% (n = 9) of the patients were further complicated by the development of a PE. Due to the limitations of the data collected, the exact cause of death could not be determined.

Discussion

Data analysis from AKUH's inpatient database between August 1, 2018, and August 31, 2019, revealed that 16% (n = 57) of patients diagnosed with VTE were those readmitted within 180 days of their primary stay, with a majority (43.9%) readmitted within the first 30 days. These results align with similar published international studies. The RIETE (Registro Informatizado de la Enfermedad TromboEmbólica) study followed a large prospective cohort of confirmed VTE patients and found that 3.0% to 3.8% experienced a recurrent DVT within 3 months. ⁹ Greenberg et al reported that 8.2% of all readmissions after aneurysmal subarachnoid hemorrhage were for thromboembolic events. ¹⁰ Other studies also showed that a large proportion of readmissions after surgical interventions were due to VTEs.^11-14 These results are not unexpected, as prevention of VTE requires proper inpatient prophylaxis and potential outpatient therapy and monitoring. In this study, a large proportion of patients (61%) were not given any pharmacological prophylaxis.

The hospital this study took place does have a set protocol for the prophylaxis of VTE. It has multiple components that should be followed every time a patient is admitted. Patients are assessed for factors that are known to increase the risk for VTE such as prolonged immobility, obesity, age over 60, family history of VTE, underlying malignancy, use of hormone replacement therapy or oral contraceptives, and known thrombophilia to name common examples.^15,16 Risk of bleeding is also assessed before starting any pharmacological therapy. After this risk stratification, methods are employed to reduce the risk of VTE, such as mechanical (compression stockings) and pharmacological prophylaxis (UFH or enoxaparin).

Even with this protocol in place, the readmission numbers due to VTE suggest that the protocol may not be properly utilized or implemented. Although evidence-based thromboprophylaxis protocols are available, compliance is poor worldwide.^17,18 This could be the result of healthcare practitioners neglecting VTE prophylaxis due to other acute medical conditions that may have been deteriorating at the time among other reasons. A study done in Pakistan found that the utilization of thromboprophylaxis for at-risk patients is low (33%), and some doctors believed that thromboprophylaxis was not relevant to their practice or they “do not feel it is important.” ⁶ The study may reveal why compliance is low globally; the lack of knowledge among healthcare workers regarding the seriousness of VTE and its burden on the healthcare system. Considering the bigger picture, the morbidity and mortality of such condition is not to be ignored as VTE is a significant cause of death in hospitalized patients. ¹⁹

One possible way of reducing readmissions due to VTE would be a mandatory risk assessment of all patients when they are admitted. The need for anticoagulation in hospitalized patients could be determined by using algorithms that consider various risk factors for VTE such as the Padua Prediction Score ²⁰ or the Caprini Risk Assessment Model. ²¹ Both have been shown to effectively stratify the VTE risk in inpatients, ²² which can then be used to prescribe suitable anticoagulation for VTE prophylaxis. Previous studies have shed light on strategies such as embedding a VTE prevention protocol in multiple stages of hospital care, admission, transfer, and preoperative stage if any. This ensures the revision of prophylaxis status for each patient at different stages during their stay. ²³

A large proportion of patients in this study developed VTE within 30 days after discharge. Although at AKUH post-discharge prophylaxis is usually not prescribed unless the patient has a specific medical indication such as atrial fibrillation, those patients who are at a higher risk can be prescribed medication post-discharge to prevent DVT from developing at home. Literature on this has been conflicting, however, as trials such as Apex or Magellan show a decreased risk in developing VTE after prescribing Betrixaban and Rivaroxaban, respectively, post-discharge, while the Mariner trial showed no statistical benefit in administering Rivaroxaban. ²⁴

An observation made during this study was that of those patients who developed VTE after their primary admission, 29% had an underlying malignancy. Malignancy is a known risk factor for developing VTE. ¹⁶ With the risk stratification algorithms mentioned above, the incidence of such events can be decreased. Another finding of note was that 16% of patients who had DVT developed PE, a potentially fatal complication that has a mortality rate of 30% if left untreated and up to 8% if treated. ²⁵ With proper prophylaxis given at the right time, the development of this complication can be reduced, and the associated mortality can be avoided.

Future Direction

It is imperative that steps be taken to reduce the readmission of patients because of the development of VTE. For this, an assessment of the baseline knowledge, attitudes, and practices of all practicing physicians in the hospital is required to identify any gaps and understand why they are there. QI measures can then be taken such as active training for thromboprophylaxis, adding checks in place that ensure proper risk stratification of every patient that is admitted and proper follow-up on VTE, DVT, and PE rates. At present, documenting the reasons for not administering VTE prophylaxis is not mandatory and was absent in patient record files. Including such information may enhance the existing protocol. As the risk of VTE development and subsequent hospitalization can be reduced with proper evidence-based prophylaxis, physicians should adhere to the implemented VTE prophylaxis protocol for the betterment of health and well-being of their patients.

Limitations

In most retrospective analysis studies, the principal limitation is the risk of missing or skewed data. The sample size for this study was only taken from one tertiary care hospital in Pakistan and may not be representative of other hospitals. Some of the information was collected from discharge summaries of patients which did not have a standardized format and as such may have missing data. There is also always the risk of human error when collecting data; however, data were double-checked by the study's authors. We did not include patients who presented to outpatient clinics with VTE. One of our biggest limitations was that we had no method of following up with patients who decided not to return to the hospital or decided to go to another hospital. Due to not having a universal electronic health record system in Pakistan and the majority of hospitals having a paper health record, it was not possible to determine whether patients had been admitted to other hospitals and extract that data. As such development of VTE after primary admission may be underestimated. In some patients, it was difficult to differentiate whether the patient required readmission because of symptoms of VTE or worsening of symptoms related to the primary admission.

Conclusion

The results of our study show a high rate of readmission due to VTE and justify the need for prophylaxis. More stringent adherence to the protocol would allow for lower rates of VTE admission and reduce the preventable increase in morbidity and mortality. This burden on the healthcare system can be substantially reduced by the introduction and implementation of QI algorithms such as appropriate risk stratification and incorporation of such modules in health care worker's training programs.

Along with risk stratification, appropriate pharmacological therapy, and continuity of care transition processes, the frequency of hospital readmissions can be reduced, enabling better distribution of health resources and increased patient satisfaction rates.

Written informed consent was obtained from a legally authorized representative(s) for anonymized patient information to be published in this article.