2022 exceptional surveillance of venous thromboembolism in over 16s (NICE guideline NG89)

The GAPS trial

The GAPS trial (Shalhoub et al 2020) was identified through NIHR. This study was first published in BMJ in May 2020, and a full report was available in December 2020 (Shalhoub et al 2020). The GAPS trial (Shalhoub et al 2020) was a multicentre, non-blinded non-inferiority trial conducted in 7 centres across the UK between May 2016 and January 2019. The non-inferiority margin was defined as a 3.5% VTE event reduction in the LMWH only arm.

In this trial, patients aged ≥ 18 years due for elective surgery and eligible for thromboprophylaxis (at least 1 "thrombosis risk" on the Department of Health (DOH) risk assessment tool) were recruited. Patients with 1 risk factor were considered as "moderate risk", and those with more than 1 risk factor were considered as "high risk". A minimisation algorithm that incorporated centre, sex and risk of VTE ("moderate" vs "high" risk) was used. Patients were not stratified by the types of surgical procedures at randomisation.

The trial excluded patients who required extended prophylaxis according to the NICE VTE prevention guideline (such as major abdominal cancer surgery patients, or orthopaedic surgery patients), or had certain clinical factors which put them at a higher risk of VTE (such as pregnancy, requiring braces or casts, known thrombophilia or thrombogenicity, previous VTE, fitted with an inferior vena cava filter) or required IPCD (Intermittent pneumatic compression device) beyond the post anaesthesia recovery area. Therefore, most patients who had orthopaedic surgery, or were at a higher risk of VTE, were excluded.

The primary outcome measured was VTE occurring within 90 days of surgery, defined as either a deep vein thrombosis (DVT) detected by whole leg ultrasound at 14-21 days post-surgery, or symptomatic pulmonary embolism (PE) or DVT confirmed by imaging and verified by blinded experts up to 90 days post-surgery. The other outcomes included all-cause mortality, major bleeding, and adverse effects from AES.

The type and brand of LMWH and AES used in this trial varied according to local practice. LMWH were given for the period of admission according to manufacturers' recommended doses. Patients who were randomised to the combination prophylaxis group also received AES providing 18 mmHg compression at the ankle level and were asked to wear AES for the duration of hospital admission. The trial report stated that the AES "…could be fitted at the time of admission or immediately post-operatively in accordance to local practice and to facilitate pragmatism," but there is no information provided about the percentage of patients fitted with AES on admission (as recommended by NICE VTE prevention guideline) versus those who only received it post-surgery.

GAPS trial study results

Of 11679 patients screened, 1905 (16.3%) participants were randomised into the trial, and 1858 were included in the intention-to-treat (ITT) analysis. 6690 (57%) patients were ineligible. The most common reason was day surgery (20.8%). Slightly over one-quarter of all screened participants were ineligible due to characteristics related to increased VTE risks. See Table 1.

Table 1: Reasons for exclusions from trial

Of the patients included in the ITT analysis, 70% (n=1292) received surgery in the abdominal area. See Table 2.

Table 2: Types of surgical procedures received

More than 80% of participants received the interventions allocated (80.0% in the LMWH alone group and 79.8% in the LMWH plus AES group). More than 95% AES used were below-knee stockings. Patients were considered "fully compliant" for LMWH if they received all prescribed LMWH doses. Compliance with AES was defined as "good" if participants wore stockings for at least 75% of their hospital admission. About 80% of participants were rated as having "full" compliance with LMWH, or "good compliance" with GCS.

VTE within 90 days occurred in 1.4% (13 out of 921) of patients in the LMWH plus AES arm compared to 1.7% (16 out of 937) of patients in the LMWH arm: risk difference of −0.30% (95% confidence interval −0.65% to 1.26%). The risks of bleeding complications were 9/1566 (0.57%) among participants who had received LMWH, and 1/160 (0.6%) among participants who did not receive any LMWH. 6.4% (50/787) of participants who received LMWH, and AES had "related complications"; 82% (41 participants) of these were "discomfort".

Implications and limitations of GAPS trial study findings

GAPS is the first major randomised controlled trial (RCT) involving AES in surgical patients in recent years and conducted within the NHS (National Health Service). The study's findings suggest that there is little advantage in offering combination prophylaxis, compared to only LMWH to patients undergoing elective surgery. While some commentators had observed that the number of VTE events (29/1958 or 1.5%) in this study is relatively low, this has to be interpreted with caution; only 16.3% of screened participants were recruited into this trial, and more than a quarter of screened participants were not eligible due to reasons related to a higher risk of VTE. Participants with additional risks of VTE (such as known thrombophilia or previous VTE) who may benefit from combination prophylaxis were excluded. Furthermore, both intervention arms received an effective pharmacological prophylaxis.

The current NICE VTE prevention guideline was published while the GAPS trial was underway and recommendations about the choice of prophylaxis have changed. In this guideline, combination prophylaxis is no longer recommended for most types of non-orthopaedic elective surgery procedures, and only remains an option for certain types of procedures if the risk of VTE is high and outweigh the risk of bleeding in an individual patient. Therefore, a significant portion of participants in this trial would no longer be considered for combination prophylaxis according to the current NICE VTE prevention guideline. Without a stratification according to the types of surgery and relatively low event rates, it is difficult to extrapolate the study's findings according to surgical procedure types where combination prophylaxis remains an option for higher risk patients.

Although compliance was considered as "full" for LMWH and "good" for AES for more than 80% of participants, it is not possible to eliminate this as a factor affecting the observed effectiveness. The AES compliance report is self-reported. Furthermore, participants who received AES in this trial could have it fitted either on admission or after surgery (the NICE VTE prevention guideline recommended "on admission") and there was no information on how many only initiated AES after surgery.

Focused literature search

Views of topic experts

In this exceptional review, we contacted 11 topic experts who were members of the guideline committee for the NICE guidelines or recruited to the NICE Centre for Guidelines Expert Advisers Panel to represent their speciality and had an interest in VTE. The topic experts completed a questionnaire about the impact of the GAPS trial on the guideline and current clinical practice. Seven topic experts responded: 2 vascular surgeons, 1 haematologist, 1 vascular consultant nurse, 1 thrombosis and haemostasis consultant nurse, 1 colorectal surgeon and 1 bariatric surgeon. Two of the respondents were also investigators in the GAPS trial.

All topic experts responded that, based on the new evidence reported in the GAPS trial, the recommendations in the NICE VTE prevention guideline (NG89) should be updated. One of the main considerations is the costs of AES to the NHS if it is not effective.

However, as there is a mix of patients with different surgical procedures included in the GAPS trials, the experts seemed to differ in their interpretation of the trial and how it should be implemented. Some seemed to have misinterpreted the GAPS trial widely across all elective surgery patients and the effectiveness of AES when used without LMWH. The GAPS trial did not suggest that LMWH is superior to AES as monotherapy. Nevertheless, there were suggestions of "stopping the use of stockings" (in all patients) or only limiting AES to patients with LMWH contraindication or other clinical factors which would have excluded them from the GAPS trial. Two of the respondents (who were involved in the GAPS trial) were aware that some trusts have changed policies and "stopped stockings" in surgical patients inappropriately whereas the others noted that there is no change in policies in their local areas. The rationale the topic experts provided includes concerns about the costs of AES to the NHS the lower baseline risk of VTE shown in recent reviews (of orthopaedic patients only). The experts also pointed out that a survey also suggested varied interpretations and implementation of the GAPS trial by individual trusts (Lawton et al 2022). In addition, there seem to be some variations in practice, as some of the responses implied that AES is often used in combination with LMWH in most elective surgery patients, and AES is regularly offered to most elective surgery patients, regardless of risk factors.