Introduction

Venous thromboembolism (VTE) is a potential source of morbidity and mortality amongst hospital in-patients. Surgical patients are particularly at risk. It has previously been estimated that VTE may account for up to 25,000 deaths annually in the UK [1]. As a result, VTE prophylaxis measures have been recommended in a number of evidence-based national and international guidelines. Prophylaxis is recommended based on the silent nature of VTE and also the preventable nature of the disease.

The risk of DVT and fatal PE in general surgery patients not receiving prophylaxis is estimated to be between 15–30% and 0.2–0.9%, respectively, based on previous cohort studies [2]. Appropriate prophylactic measures, including low-molecular weight heparin (LMWH) and anti-embolic stockings (AES), however, have been estimated to reduce the risk of VTE in surgery patients by up to 60% [2]. Despite this obvious benefit, the agents may not be suitable for all patients and a risk of bleeding exists with LMWH use.

As a result, a risk assessment is commonly recommended as part of the hospital admission process to document and clarify the best prophylactic protocol. This recommendation is present in both the National Institute of Clinical Excellence (NICE) and Scottish Intercollegiate Guideline Network (SIGN) guidelines [3, 4]. As yet, no Irish national guideline on VTE prophylaxis in surgical patients has been published. Our centre uses an evidence-based institutional guideline that makes recommendations similar to the above UK guidelines [5].

In the UK, VTE measures, including assessment, are commonly linked to tariffs for Hospital Trusts leading to high rates of compliance (95%) [6]. The situation in the Irish setting is, however, less certain. We aimed to assess venous thromboembolism prophylaxis prescribing practices and compliance with a kardex-based risk assessment tool in our general surgery population.

Methods

Two audit cycles were completed. As part of the initial baseline audit cycle (Audit A) data on all general surgery in-patients, on two separate surgical wards on two separate days in October 2015, was collected. All data collection was performed by the lead author. Drug kardexes were assessed for VTE prophylaxis measures and use of the risk assessment tool contained in the drug kardex. The adequacy of VTE prophylactic measure compliance was based on the risk assessment tool recommendations. A pro-forma modelled on the SIGN audit tool was used for data collection. The NICE, SIGN, and institutional guidelines were adopted as a gold standard for audit against [3–5]. All state that patients should undergo a risk assessment for VTE and bleeding risk at admission and that appropriate mechanical and pharmacological prophylaxis should be prescribed based on this.

The initial audit results and information on the risk assessment tool were presented as an educational intervention at two separate departmental teaching sessions. All grades of doctor from the various general surgical teams as well as medical students routinely attend these sessions. Approximately 25 members of the general surgical teams attended these meetings. A second follow-up audit (Audit B) to close the audit loop and assess response was completed at 3 months in January 2016. Statistical analysis was performed using GraphPad QuickCalc (Graphpad Inc., California, USA). Analysis assessing any change post-intervention was performed using the fisher exact test with a p value of 0.05 or less considered to be statistically significant.

Results

In Audit A, 74 patients were assessed. Emergency admissions accounted for 70% of patients assessed. Elective admissions included a range of patients from the various surgical teams, including colorectal, breast, and vascular surgery. The new kardex-based risk assessment tool was completed in 2.7%. In relation to VTE prophylactic measures, AES were advised in 55 patients (74%). It was prescribed in 75% of these cases. LMWH was advised in 65 patients and was prescribed in 97% of cases.

In Audit B, a smaller sample of 30 patients were included to assess the result of the intervention. In this group, 56% were emergency admissions. 66% had a risk assessment performed, a statistically significant improvement (p < 0.0001). AES were advised in 22 patients (73%); however, prescribing rates were lower at 36%. LMWH was advised in 27 patients (90%) and prescribed in 96% of patients. Table 1 provides a summary of the results.

Table 1 Summary of audit findings
Full size table

Discussion

Given the potential risks associated with VTE, there is an understandable requirement for adequate prophylactic measures. Unfortunately, despite recognition of these risks adequate prophylaxis is often lacking. One large multinational study, ENDORSE, assessed prophylaxis in a large sample of both medical and surgical patients [7]. This study included patients from 358 hospitals in 32 different countries with a sample of 30,827 surgical patients assessed. Patients from most surgical specialities, including orthopaedics, were included. In this surgical group, 19,842 (64.4%) patients were deemed to be at risk based on the 2004 American College of Chest Physician Guidelines [2], but only 58.5% received adequate prophylaxis. A lower proportion, 31%, received some form of mechanical prophylaxis. The Irish component of this study examined 552 patients, including 175, at risk surgical patients. In this group, 64% received the recommended prophylaxis, including 54% placed in AES [8].

Audit plays a key role in the quality improvement process and the data collected during our study showed some interesting findings. We demonstrated a marked improvement in use of a new kardex-based risk assessment tool following the second audit cycle. It is likely the educational intervention accounted for some of this improvement. The initial audit highlighted very poor compliance with the risk assessment tool at baseline. However, as the tool had only recently been introduced in the hospital kardex, a lack of familiarity or awareness of it may account for this low initial rate. It would be expected that continuing education and familiarity with the risk assessment tool will lead to improved compliance. The intervention in this study was also limited to doctors. Given the involvement of nursing staff to VTE delivery, it might be expected that expansion to include nursing and allied health staff would improve compliance and results at further audits.

Our results also demonstrate very high rates of appropriate LMWH prescription in general surgical patients. Although our sample size is relatively small, this appears to be an improvement on previously reported rates and was evident in both audit cycles [7, 8]. Rates of AES use were also high in the initial audit; however, prescription rates varied between the two cycles, suggesting less consistency with this measure. Further educational intervention may need to highlight the importance and benefit of mechanical prophylaxis given this marked variation.

There are some limitations to our audit. Our sample size overall is small making the generalisability of our findings difficult to confirm. There is also a potential bias given the limited time period. Only one group of doctors rotating through the department at the time of the audit was assessed. Prescribing practices in this group could potentially be misrepresentative of other junior doctors. Despite this, we feel our findings that are reflective of current practice.

Conclusion

Our results demonstrate that rates of LMWH prescribing are high in this general surgical population although AES prescribing rates vary. The use of the VTE risk assessment tool, although low initially, increased following the initial audit and a simple educational intervention. This highlights the importance of regular audit and education to improve compliance with clinical standards.