This was a cross-sectional observational study and the data was collected from medical records of patients. Records of all patients who were admitted to medical ICU in a tertiary care hospital from October 2019 to December 2019 were included in the study. Ethical clearance was obtained for the present study from Institutional Ethical Committee (IEC No. IEC/2019/267). Patients with an ICU stay of less than 48 hours or already on anti-coagulation were excluded from the study. The details of each patient were taken from the medical records. This included their age, sex, current illness, other co-morbidities and whether risk assessment was done for VTE. Based on medical records risk assessment was done using the Padua score [6] and IMPROVE VTE RAM [7]. The bleeding risk was calculated using the IMPROVE Bleeding risk score [8]. The data required for calculating the above scores were taken from medical records and based on the validated risk assessment scoring the respective scores were calculated. The prophylaxis given to patients was scrutinised for appropriateness based on ASH guidelines [10].

A total of 212 admissions took place in medical ICU of the hospital from 01^st October 2019 to 31^st December 2019. Amongst these 36 were excluded from the study, 32 patients either died within 48 hours of admission or were transferred to other wards in view of improved general condition. A total of 04 patients were already on anti-coagulation. Total 176 medical records were included in this study [Table/Fig-1].

[Table/Fig-1]:

Flow chart of recruitment of patients.

The results were computed using Microsoft Excel. The continuous variables were described using mean and Standard Deviation (SD). The categorical variables were analysed using number and percentages with 95% confidence interval. The statistical analysis for difference between risk assessments based on Padua score and IMPROVE VTE RAM was calculated by Chi-square test and p-value <0.05 were considered significant.

Patient Characteristics

The mean age of patients was 61.65 (±16.8) years. There were 115 male patients and 61 female patients (M:F ratio of 1.89). Only 3 patients (1.7%) had Body Mass Index (BMI) recorded out of 176 patients. The major contributors to ICU admission were Cerebrovascular Accident (CVA) (22.73%; 95% CI: 28.91%-16.53%) and Pneumonia (10.23%; 95% CI: 14.70%-05.75 [Table/Fig-2]. Primary Hypertension (53.40%; 95% CI: 60.77%-40.03%) and Type 2 Diabetes Mellitus (36.36%; 95% CI: 43.47%-29.25%) were the most common co-morbidities in these patients [Table/Fig-3].

Risk Assessment for VTE

Out of the 176 patients appropriate risk assessment was not done in any patient.

Padua Score

On calculating the Padua score 149 patients (84.66%; 95% CI: 89.98%-79.33%) were in high risk for VTE (Padua Score ≥4) and 27 patients were (15.34%; 95 % CI: 10.01%-20.66%) in low risk for VTE.

IMPROVE VTE RAM

Based on IMPROVE VTE RAM 39 patients (22.16%; 95% CI: 28.29%-16.02%) were in low risk (Score of 0 or 1), 94 (53.41%; 95% CI: 60.77%-46.03%) in moderate risk (Score of 2 or 3) and 43 (24.43%; 95% CI: 30.77%-18.08%) in high risk (Score equal to or more than 4) for VTE thus prophylaxis was indicated in 137 patients (Moderate and high risk) (77.84%; 95% CI: 83.97%-71.70%).

VTE Prophylaxis

Prophylaxis was given in a total of 76 patients out of which 09 received unfractionated heparin in view of renal involvement and rest all patients received LMWH. Mechanical prophylaxis was not given to any patient.

The sensitivity of clinician’s judgement as compared to Padua score in assessing the requirement of thromboprophylaxis in a patient was 51% and the specificity was 100%. Positive Predictive Value and Negative Predictive Value (NPV) of clinician’s judgement as compared to Padua score in assessing the requirement of thromboprophylaxis in patients were 100% and 27%, respectively. Accuracy of clinician’s judgement was 58.52% as compared to Padua score in assessing requirement of thromboprophylaxis in patients [Table/Fig-4].

The sensitivity of clinician’s judgement as compared to IMPROVE VTE RAM in assessing the requirement of thromboprophylaxis in a patient was 55.47%. PPV and NPV of clinician’s judgement as compared to IMPROVE VTE RAM in assessing the requirement of thromboprophylaxis in patients were 97.43% and 37.75%, respectively [Table/Fig-5].

Difference in number of patients in whom thromboprophylaxis was indicated as per IMPROVE VTE RAM and as per Padua RAM was not significant (p-value=0.10) [Table/Fig-6].

IMPROVE Bleeding Risk

There was high risk of bleeding in 91 patients (51.7%; 95% CI: 59.08%-44.32%) as assessed by IMPROVE bleeding risk score and incidence of major bleeding was 05 in high risk patients and 02 in low risk patients however none of the bleeding episode was related to anticoagulation. No episode of VTE was recorded in any patient.

Bleeding Episodes

Seven (3.97%; 95% CI: 06.86%-01.09%) patients had bleeding episodes but none was related to anticoagulation rather they were part of patient’s primary diagnosis (Four patients with intracerebral haemorrhage due to CVA, one patient with Upper Gastrointestinal (UGI) bled due to duodenal ulcer, one patient with UGI bleed due to Immune Thrombocytopenic Purpura (ITP) and one patient with UGI bleed due to esophageal varices).

Majority of critical medical patients are elderly; in present study the mean age was 61.65 (±16.8) years. These patients are at a higher risk of thromboembolism as well as bleed thus these patients must be assessed appropriately for the need of thromboprophylaxis. Only three patients had BMI recorded in medical documents likely due to inability to take weight in bed bound patients in a critical care environment.

In this study, none of the patients have been assessed for risk of VTE using any validated tool. There were 149 (84.65%; 95% CI: 89.98%-79.33%) patients in high risk as per Padua score however only 76 out of the 149 patients in high risk (51%; 95% CI: 59.03%-42.98%) were given thromboprophylaxis based on overall clinical impression of treating physician. A total of 73 patients who were in high risk of VTE as per Padua score and 61 patients who were either moderate or high risk for VTE as per IMPROVE VTE RAM score did not receive any thromboprophylaxis. The clinician has however not prescribed thromboprophylaxis to any patient in low risk of VTE as per Padua score. This reflects the fact that we have missed many high-risk patients who required thromboprophylaxis. The accuracy of clinicians judgement as compared to validated risk assessment tools have been just above 50% {58.52% (95% CI: 65.80%-51.24%) as per Padua score and 64.20% (95% CI: 71.28%-57.12%) as per IMPROVE VTE RAM} which is grossly inadequate, emphasising the need to assess each medical patient using a validated tool for risk of VTE as early as possible after admission to a critical care setting. Similar findings were seen in ENDORSE study in which only 39.5% of total medical patients in high risk for VTE received any thromboprophylaxis. In the same study, Indian data showed that only 17.4% of the total high-risk patients received thromboprophylaxis [3]. Another study by Parikh KC et al., revealed that 20% of patients requiring thromboprophylaxis did not receive it [12]. These findings emphasise the fact that risk assessment is of utmost importance and must be practiced by all critical care physicians.

There is limited data on comparison between IMPROVE VTE RAM and Padua score for VTE risk assessment however in this study difference in number of patients in whom thromboprophylaxis was indicated as per IMPROVE VTE RAM and as per Padua RAM was not significant (p-value=0.1). This emphasises the fact that both risk assessment tools are equivalent in assessing risk for VTE. The consensus on management of deep vein thrombosis by association of physicians of India recommends the use of Padua score for risk assessment in medical patients [13]. Irrespective of the risk assessment tool used, the practice of thromboprophylaxis will improve if we follow any validated risk assessment tool to decide whether a patient requires thromboprophylaxis for VTE.

The bleeding risk was high in 51.7% (95% CI: 59.08%-44.32%) patients in this study as compared to 13% patients in the STIME study which is likely due the clinical profile of ICU medical patients who have multiple co-morbidities and are on multiple drugs making them vulnerable to bleeding episodes. In STIME study, all patients admitted to a medical unit were assessed and not only ICU patients [14].

Obesity is one of the risk factors for VTE [15] and this factor could not be added in present study patients in view of non-availability of data; if the BMI would have been available than the number of high-risk cases might have been higher using the Padua score. No VTE episodes were recorded in present study however in view of limited follow-up of patients due to paucity of medical records and relatively small sample size this should not be interpreted as reduced risk of VTE in these critically ill patients. This data was taken from a single centre and thus large scale data from multiple hospitals will help us understand the problem better.

Critically ill medical patients are at high risk of VTE as well as bleeding episodes and thus appropriate prophylaxis is of utmost importance which is not being practiced in critical care settings. There is a need to sensitise the healthcare providers about the importance of risk assessment for VTE and bleeding in all critical care patients using any validated risk assessment tool. This will go a long way in preventing morbidity and mortality in the patients.