Abdominal surgery (excluding bariatric surgery)

National Guideline Centre (UK)

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Guideline Centre (UK). Venous thromboembolism in over 16s: Reducing the risk of hospital-acquired deep vein thrombosis or pulmonary embolism. London: National Institute for Health and Care Excellence (NICE); 2018 Mar. (NICE Guideline, No. 89.)

See Partial Update

December 2019: In recommendation 1.3.5 the British Standards for anti-embolism hosiery were updated because BS 6612 and BS 7672 have been withdrawn. August 2019: Recommendation 1.12.11 (1.5.30 in this document) was amended to clarify when anti-embolism stockings can be used for VTE prophylaxis for people with spinal injury.

December 2019: In recommendation 1.3.5 the British Standards for anti-embolism hosiery were updated because BS 6612 and BS 7672 have been withdrawn. August 2019: Recommendation 1.12.11 (1.5.30 in this document) was amended to clarify when anti-embolism stockings can be used for VTE prophylaxis for people with spinal injury.

Cover of Venous thromboembolism in over 16s

Venous thromboembolism in over 16s: Reducing the risk of hospital-acquired deep vein thrombosis or pulmonary embolism.

Show details

Contents

< Prev Next >

35Abdominal surgery (excluding bariatric surgery)

35.1. Introduction

This section covers major abdominal surgery, including both open and laparoscopic surgery. Major abdominal surgery covers inpatients undergoing gastrointestinal, gynaecological and urological surgery.

Gastrointestinal surgery by its nature is heterogeneous in terms of the age of patients, the pathological conditions being dealt with and organs and systems operated upon. There remain a variety of procedures retained within this category that are specialisations in themselves. These include upper gastrointestinal surgery and lower intestinal surgery (or coloproctology). Factors that may alter the risk of VTE:

Patients having surgery for cancer will have an increased risk of developing a DVT or pulmonary embolism.
Patients having emergency procedures are often elderly and will consequently be at higher risk of developing a DVT or pulmonary embolism.
Some patients having emergency procedures may already be using anticoagulation or antiplatelet therapy. This needs to be considered when deciding on the method of VTE prophylaxis.

Open gynaecological surgery includes abdominal and vaginal surgery, excluding caesarean section. Factors that may alter the risk of VTE:

Patients may be using hormonal contraception and hormone replacement therapy, which will increase their risk of developing a DVT or pulmonary embolism.
Patients having surgery for cancer will have an increased risk of developing a DVT or pulmonary embolism.

Open urological surgery is divided into two major groups: pelvic cancer surgery and renal surgery. Patients undergoing these procedures are usually between the ages of 65 and 75.

Factors that may alter the risk of VTE:

Many urological surgery patients have spinal and epidural anaesthesia. This may reduce the risk of developing a DVT.
Renal surgery procedures may involve division of the renal vein where it drains into the inferior vena cava. This could potentially increase the risk of VTE.

There are no specific factors that increase the risk of bleeding or the hazard associated with it in open gastrointestinal, gynaecological or urological surgery. There are no other special factors that would affect the choice of, and use of, specific methods of VTE prophylaxis in these surgeries.

Laparoscopic surgery is used in gastrointestinal, gynaecological and urological surgery. Specific considerations apply to it in all these specialities. Factors that may alter the risk of VTE:

There is some concern that the increased pressure in the peritoneal cavity during laparoscopic surgery causes venous stasis which may increase VTE risk.
Some laparoscopic procedures tend to last longer than open procedures.
Being less invasive, most people will make a quicker return to mobility following laparoscopic procedures compared to open procedures.

Factors that may alter the risk of bleeding:

Laparoscopic procedures may be associated with less bleeding than open surgery.
Bleeding may make laparoscopic surgery difficult or impossible and result in the need for conversion to open surgery.

There are no other special factors that may affect the choice, and use of, specific methods of VTE prophylaxis in laparoscopic surgery.

35.2. Review question: What is the effectiveness of different pharmacological and mechanical prophylaxis strategies (alone or in combination) for people undergoing abdominal surgery (gastrointestinal, gynaecological, urological)?

For full details see review protocol in appendix C.

Table 167

PICO characteristics of review question.

35.3. Clinical evidence

Sixty-seven studies in 69 papers were included in the review these are summarised in Table 168 below. Sixty-two studies were previously included in the previous guideline (CG92);⁵^, ³¹⁷^, ³¹⁶^, ²⁹²^, ²⁹³^, ⁶^, ¹⁹^, ³⁷^, ³⁸^, ⁴⁴^, ²⁹^, ³⁰^, ²⁸^, ²⁵ ^,²⁶^, ²⁴^, ²²^, ⁴²^, ⁵⁰^, ⁵⁴^, ⁵⁶^, ⁵⁵ ^,⁵⁷ ^,⁵⁸^, ⁹²^, ¹⁰²^, ¹⁰⁹ ^,¹¹⁰^, ¹¹⁸^, ¹²⁰ ^,¹³¹ ^,¹³⁸^, ¹³⁶^, ¹⁵⁶^, ¹⁶⁰^, ¹⁵⁹^, ¹⁶⁹^, ¹⁷⁵^–¹⁷⁷^, ¹⁹³^, ²⁰²^, ²⁰⁴^, ²¹⁰^, ²³²^, ²³⁵^, ²³⁸^, ²³⁹^, ²³⁶^, ²⁴⁵^, ²⁵⁰^, ²⁵¹^, ²⁶⁸^, ²⁷²^, ²⁸⁴^, ²⁸⁶^, ²⁹¹^, ²⁹⁰^, ²⁹⁴^, ³⁰³^, ¹³⁷ ^,³⁰² and five studies were added to the update;¹¹¹^, ²⁶⁰^, ²²³^, ¹⁵⁸^, ²⁷³^, ¹³⁷. Evidence from these studies is summarised in the clinical evidence summary tables below (Table 169, Table 170, Table 171, Table 172, Table 173, Table 174, Table 175, Table 176, Table 177, Table 178, Table 179, Table 180, Table 181, Table 182, Table 183, Table 184, Table 185, Table 186, Table 187, Table 188, Table 189, Table 190, Table 191, Table 192, Table 193, Table 194, Table 195, Table 196, Table 197, Table 198, Table 199, Table 200, Table 201, Table 202, Table 203, Table 204, Table 205, Table 206, Table 207). See also the study selection flow chart in appendix E, forest plots in appendix L, study evidence tables in appendix H, GRADE tables in appendix K and excluded studies list in appendix N.

Based on the current review protocol, six systematic reviews that were included in CG92 were excluded but checked for references. The studies from all of one systematic review¹¹ were excluded due to having the incorrect intervention. Some of the studies from five systematic reviews⁷^, ⁶¹^, ¹⁶⁷^, ²¹⁷^, ²⁵⁶ were excluded due to having incorrect population, intervention or comparisons. For this update, data from the original papers, rather than systematic review data, was used.

A large amount of people undergo major abdominal surgery, and where evidence for other populations relating to torso surgery (e.g. thoracic surgery and cardiac surgery) is lacking, the committee agreed to consider major abdominal surgery as indirect evidence. Therefore in order to compare the clinical effectiveness data of multiple possible interventions, it was proposed that a network meta-analysis be carried out on the outcome data for DVT, PE and major bleeding in this population. These analyses provide estimates of effect (with 95% credible intervals) for each intervention compared to one another and compared to a single baseline risk (in this case the baseline treatment was no prophylaxis or in the case of the major bleeding outcome a combination of no prophylaxis and mechanical prophylaxis). These estimates provide a useful clinical summary of the results and facilitate the formation of recommendations based on the best available evidence. For full details on the NMA methodology and results, please see appendix M.

Table 168

Summary of systematic reviews included in the review.

Table 169

Clinical evidence summary: AES (above knee) versus no prophylaxis.

Table 170

Clinical evidence summary: AES (below knee) versus no prophylaxis.

Table 171

Clinical evidence summary: AES (undefined) versus no prophylaxis.

Table 172

Clinical evidence summary: AES (above knee) versus UFH.

Table 173

Clinical evidence summary: AES (below knee) versus UFH.

Table 174

Clinical evidence summary: AES (above knee) versus AES (below knee).

Table 175

Clinical evidence summary: AES (below knee) + UFH versus AES (below knee).

Table 176

Clinical evidence summary: AES (above knee) + UFH versus UFH.

Table 177

Clinical evidence summary: AES (below knee) + UFH versus UFH.

Table 178

Clinical evidence summary: AES (above knee) + IPCD (full leg) versus AES (above knee).

Table 179

Clinical evidence summary: AES (undefined) + IPCD (full leg) versus AES (undefined).

Table 180

Clinical evidence summary: AES (undefined) + IPCD (full leg) versus UFH.

Table 181

Clinical evidence summary: AES (undefined) + IPCD (full leg) versus electrical stimulation.

Table 182

Clinical evidence summary: Electrical stimulation versus UFH.

Table 183

Clinical evidence summary: Foot pump versus no prophylaxis.

Table 184

Clinical evidence summary: FID + IPCD (below knee) + LMWH (low dose) versus FID + IPCD (below knee).

Table 185

Clinical evidence summary: IPCD (below knee) versus no prophylaxis.

Table 186

Clinical evidence summary: IPCD (full leg) versus IPCD (below knee).

Table 187

Clinical evidence summary: IPCD (full leg) versus VKA.

Table 188

Clinical evidence summary: ICPD (undefined) + LMWH (standard prophylactic dose) versus IPCD (undefined).

Table 189

Clinical evidence summary: UFH versus no prophylaxis/mechanical.

Table 190

Clinical evidence summary: UFH versus IPCD (below knee).

Table 191

Clinical evidence summary: UFH versus VKA.

Table 192

Clinical evidence summary: LMWH (low dose) versus no prophylaxis.

Table 193

Clinical evidence summary: LMWH (low dose; standard duration) versus UFH.

Table 194

Clinical evidence summary: LMWH (standard dose; standard duration) versus no prophylaxis/mechanical.

Table 195

Clinical evidence summary: LMWH (standard dose; standard duration) versus IPCD (undefined).

Table 196

Clinical evidence summary: LMWH (standard dose; standard duration) versus UFH.

Table 197

Clinical evidence summary: LMWH (high dose; standard duration) versus no prophylaxis.

Table 198

Clinical evidence summary: LMWH (high dose; standard duration) versus UFH.

Table 199

Clinical evidence summary: LMWH (low dose; standard duration) versus LMWH (standard dose; standard duration).

Table 200

Clinical evidence summary: LMWH (standard dose; extended duration) versus LMWH (standard dose; standard duration).

Table 201

Clinical evidence summary: LMWH (high dose; extended duration) versus LMWH (high dose; standard duration).

Table 202

Clinical evidence summary: LMWH (standard dose; extended duration) + AES (undefined) versus LMWH (standard dose; standard duration) + AES (undefined).

Table 203

Clinical evidence summary: Fondaparinux versus LMWH (standard dose; standard duration).

Table 204

Clinical evidence summary: Fondaparinux + IPCD (undefined) versus IPCD (undefined).

Table 205

Fondaparinux versus no prophylaxis/mechanical.

Table 206

Fondaparinux + UFH + mechanical (AES + IPCD) versus LMWH + UFH + mechanical (AES + IPCD).

Table 207

VKA versus no prophylaxis.

35.4. Economic evidence

Published literature

Two original economic models were developed for this population in CG92.²²⁴ Additionally, one health economic study was also identified with the relevant comparison and has been included in this review.³⁰⁵ These are summarised in the health economic evidence profiles below (Table 208, Table 209 and Table 210) and the health economic evidence tables in appendix J.

An economic model was developed for this population in CG46; for both standard duration and post-discharge prophylaxis. Both these models were selectively excluded due to the availability of the more applicable model from CG92.²²⁴ Additionally, three economic studies relating to this review question were previously included in CG46,²²⁶ but one was excluded due to methodological limitations,²¹⁹ and the other two were selectively excluded due to the availability of more applicable evidence.¹²¹ ^,²⁵³ These are listed in appendix O, with reasons for exclusion given.

See also the health economic study selection flow chart in appendix F.

Table 208

Health economic evidence profile: LMWH (standard dose, standard duration) + AES (knee-length) vs LMWH (standard dose, standard duration) + AEs (thigh-length) vs LMWH (standard dose, standard duration).

Table 209

Health economic evidence profile: pharmacological, mechanical or combination of prophylaxis strategies vs each other.

Table 210

Health economic evidence profile: LMWH (post-discharge) vs no post-discharge prophylaxis.

35.5. Evidence statements

Clinical

Pairwise meta-analysis statements

Mechanical prophylaxis versus mechanical prophylaxis

AES

Two studies (n=291) evaluated the use of above knee AES compared to no prophylaxis. A clinical benefit of AES was found for DVT, and a possible clinical benefit was found for PE, although for this outcome there was very serious imprecision around the estimate. No clinical difference was found for all-cause mortality. The evidence ranged from very low to moderate quality due to risk of bias and imprecision.

One study (n=95) compared below knee AES to no prophylaxis and found a possible clinical benefit of stockings in terms of DVT. However there was very serious imprecision, and therefore the estimate is also consistent with no difference and clinical harm. The evidence was very low quality due to risk of bias and imprecision.

One study compared AES at an undefined length to no VTE prophylaxis. The evidence showed that for the outcome of DVT, there was a clinical benefit of AES. Evidence for this comparison was of moderate quality due to risk of bias.

One study (n=114) compared above knee AES with below knee AES. For the only reported outcome of DVT, there was a possible clinical harm of above knee AES, however there was very serious imprecision around the estimate and therefore was also consistent with no difference and clinical benefit. The evidence for this comparison was of very low quality due to risk of bias and imprecision.

Foot pump

One study of 66 participants evaluated the use of foot pumps compared to no prophylaxis. The evidence demonstrated a possible clinical benefit of foot pumps in terms of both all-cause mortality and DVT, however imprecision around these estimates was also consistent with no difference and in the case of mortality, also possible harm as well. The quality of evidence for this comparison ranged from low to very low due to risk of bias and imprecision.

IPCD

Four studies evaluated IPCD (below knee) versus no prophylaxis. A possible clinical benefit of IPCD was found for both DVT and fatal PE, however for both of these outcomes there was very serious imprecision around the estimate, and therefore was also consistent with no difference and clinical harm. No clinical difference was found for all-cause mortality, and there was a suggested clinical harm of IPCD in terms of PE. Again, both of these outcomes had serious imprecision around the estimate. The evidence for this comparison was very low due to risk of bias, imprecision, and for the DVT outcome, inconsistency.

One study (n=90) evaluated the use of IPCD (full leg) compared to IPCD (below knee). The evidence showed a possible clinical benefit of full leg IPCD in terms DVT and fatal PE, but a suggested clinical harm for full leg IPCD in terms of PE. Quality was very low due to risk of bias and imprecision.

Pharmacological prophylaxis versus pharmacological prophylaxis

UFH

Two studies evaluated the use UFH versus VKA in terms of DVT (n=197). A possible clinical benefit was found for UFH, however there was serious imprecision around the estimate and therefore evidence was also consistent with no difference. One study reported the outcome of major bleeding (n=100). No clinical difference was found between UFH and VKA, however there was very serious imprecision which meant that this was also consistent with clinical benefit and clinical harm. The evidence quality ranged from low to very low due to risk of bias and imprecision.

LMWH (low dose)

One study compared LMWH at a low dose with no prophylaxis (n=183). There was a suggested clinical benefit for LMWH for all-cause mortality, DVT and PE. There was no clinical difference for major bleeding and thrombocytopaenia. Quality ranged from very low to low due to risk of bias, imprecision and for one outcome, indirectness.

LMWH at a low dose was compared to UFH. Seven studies reported the outcomes all-cause mortality, PE and major bleeding (n=6694–7018). The evidence demonstrated a possible clinical harm of LMWH for all-cause mortality, and a possible clinical harm for major bleeding. Both outcomes had serious imprecision around the estimate, and therefore were also consistent with no difference. There was no clinical difference between LWMH and UFH in terms of PE, with very serious imprecision consistent with clinical benefit and clinical harm. Five studies reported the outcomes DVT and fatal PE (n=3045–5848). Evidence from these studies showed a possible clinical harm for both outcomes, however there was serious and very serious imprecision around the estimates. The quality of the evidence ranged from very low to low due to risk of bias, imprecision and inconsistency.

LMWH at a low dose was compared to LMWH at a standard dose. Two studies reported the outcome all-cause mortality (n=2931). The evidence demonstrated a possible clinical harm of low dose LMWH, however there was very serious imprecision consistent with no difference and benefit. Three studies reported the outcomes DVT, PE and major bleeding (n=2853–2966). There was a possible clinical harm of low dose LMWH in terms of DVT, no clinical difference in terms of PE, and a possible clinical benefit of low dose LMWH in terms of major bleeding. All outcomes had very serious imprecision. One study reported the outcome fatal PE (n=35). This study demonstrated no clinical difference between the two doses of LMWH, however there was very serious imprecision consistent with both harm and benefit. Evidence for the comparison ranged from very low to moderate quality, due to risk of bias, imprecision and, for the major bleeding outcome, indirectness and inconsistency.

LMWH (standard dose)

For the comparison of LWMH (standard dose) versus UFH, eight studies reported the outcomes DVT, PE and major bleeding. There was a possible clinical benefit of LMWH for PE, no clinical difference for DVT, and a suggested clinical harm of LMWH for major bleeding. The DVT outcome had serious imprecision around the estimate consistent with benefit, whereas the major bleeding outcome demonstrated serious imprecision consistent with no difference. Five studies reported the outcome all-cause mortality. No clinical difference between LMWH and UFH was found, however there was very serious imprecision around the estimate, and therefore was consistent with clinical harm and clinical benefit. One study reported fatal PE, and found a possible clinical benefit of LMWH, however this outcome had very serious imprecision consistent with no difference and clinical harm. The evidence ranged from low to very low quality due to risk of bias, imprecision, and inconsistency.

Standard dose LMWH at an extended duration was compared to standard dose LMWH at a standard duration. One study reported the outcomes all-cause mortality, DVT, PE and fatal PE (n=332–501). A possible clinical benefit of extended duration LMWH was found for all-cause mortality, DVT, PE and fatal PE, however all outcomes had either serious or very serious imprecision around the estimate. Two studies reported the outcome major bleeding (n=928). There was no clinical difference for major bleeding, however there was very serious imprecision around the estimate consistent with both benefit and harm. The evidence ranged from very low to low quality due to risk of bias and imprecision.

LMWH (high dose)

One study evaluated LMWH at a high dose versus no prophylaxis. The evidence demonstrated a possible clinical benefit for LWMH was found for DVT. However there was serious imprecision around the estimate, and therefore evidence was also consistent with no difference. No clinical difference was found between LMWH and no prophylaxis in terms of all-cause mortality, however again there was very serious imprecision around the estimate. The evidence was of low quality due to risk of bias and imprecision.

For the comparison of LMWH at a high dose versus UFH, one study of 43 participants reported the outcomes all-cause mortality, DVT and major bleeding. There was no clinical difference between the two pharmacological prophylaxis methods for the all-cause mortality and DVT outcomes, although there was very serious imprecision around the estimate for both outcomes, which therefore were also consistent with benefit and harm. There was a possible clinical harm of LMWH in terms of major bleeding, with very serious imprecision around the estimate. The quality of the evidence was very low for all outcomes due to risk of bias and imprecision.

One study compared high dose LMWH at an extended duration versus high dose LMWH at a standard duration (n=488–625). A possible clinical benefit of extended duration LMWH was found for DVT, however there was serious imprecision around the estimate and therefore was also consistent with no difference. A possible clinical harm was found for all-cause mortality and major bleeding however there was very serious imprecision consistent with no difference and benefit. There was no clinical difference for PE, with very serious imprecision consistent with both benefit and harm. The evidence ranged from very low to low quality due to risk of bias and imprecision.

Fondaparinux

One study compared fondaparinux to LMWH at a standard dose (n=2042–2927). A possible clinical benefit was found for fondaparinux in terms of all-cause mortality, and DVT. Both outcomes had serious imprecision around the estimate and so were also consistent with no difference. A possible clinical harm was found for PE and major bleeding. Very serious imprecision around the estimate for PE meant that it is also consistent with no difference and benefit, and serious imprecision around the estimate for major bleeding meant that the outcome is also consistent with no difference. No clinical difference was found for fatal PE, with very serious imprecision. The evidence ranged from low to very low quality due to risk o bias and imprecision.

VKA

One study compared VKA with no prophylaxis (n=96). For the outcome of DVT, there was a possible clinical benefit of VKA, however there was serious imprecision around the estimate and therefore this was also consistent with no difference. The evidence was low quality due to risk of bias and imprecision.

Mechanical prophylaxis versus pharmacological prophylaxis

One study compared above knee AES with UFH (n=97). There was a possible clinical benefit of AES in terms of fatal PE, however there was very serious imprecision around the estimate consistent with no difference and harm. The evidence was very low quality due to risk of bias and imprecision.

One study compared below knee AES with UFH (n=159). No clinical difference was found for both all-cause mortality and PE, with very serious imprecision consistent with both benefit and harm. The evidence was of very low quality due to risk of bias, imprecision and, for the PE outcome, indirectness.

One study of 100 participants compared electrical stimulation with UFH. There was a possible clinical harm of electrical stimulation in terms of DVT, however there was very serious imprecision consistent with benefit and no difference. The evidence was of very low quality due to risk of bias and imprecision.

One study compared full leg IPCD versus VKA (n=100). A possible clinical harm of ICPD was found for DVT and PE. For both outcomes there was very serious imprecision around the estimate consistent with benefit and no difference. There was no clinical difference for all-cause mortality, again with very serious imprecision. The evidence was very low quality due to risk of bias and imprecision.

Pharmacological prophylaxis versus mechanical prophylaxis

UFH was compared to no prophylaxis/mechanical prophylaxis. Twelve studies reported the outcome DVT (n=1991), and the evidence demonstrated a clinical benefit for UFH. Ten studies reported the outcome PE (n=897). There was a possible clinical benefit of UFH, however there was serious imprecision, and was therefore also consistent with no clinical difference. Seven studies reported the outcome major bleeding (n=725). This demonstrated a possible clinical harm of UFH, with serious imprecision consistent with no difference. Four studies reported the outcomes all-cause mortality and fatal PE (n=393–506). There was a possible clinical benefit of UFH for both outcomes, however both outcomes also had very serious imprecision around the estimate and were consistent with no difference and clinical harm. The evidence ranged from very low to moderate quality due to risk of bias and imprecision.

Standard dose LMWH was compared to no prophylaxis/mechanical prophylaxis. One study reported the outcome all-cause mortality (n=80). There was a possible clinical benefit of LMWH for this outcome, however there was very serious imprecision around the estimate and so this was also consistent with harm and no difference. Two studies reported DVT and PE (n=130). There was a possible clinical benefit of LMWH for both outcomes, however there was serious and very serious imprecision around the estimates, consistent with no difference, and no difference and clinical harm. Five studies reported the outcome major bleeding (n=527). The evidence demonstrated a possible clinical harm of LMWH for this outcome, however there was serious imprecision which was also consistent with no difference. The evidence was very low to low quality due to risk of bias and imprecision.

One study compared fondaparinux to no prophylaxis/mechanical prophylaxis (n=1285). There was a clinical harm of fondaparinux in terms of DVT. No other outcomes were reported. The evidence was high quality.

Two studies compared UFH and below knee IPCD (n=265). A possible clinical harm was found for UFH in terms of DVT, however there was serious imprecision around the estimate and therefore was also consistent with no difference. No clinical difference was found for PE, however there was very serious imprecision around the estimate consistent with both benefit and harm. The evidence ranged from very low to low quality due to risk of bias and imprecision.

One study compared standard dose LMWH to IPCD at an undefined length (n=211). The evidence demonstrated a possible clinical harm of LMWH in terms of DVT, however there was very serious imprecision around the estimate consistent with no difference and benefit. There was no clinical difference in terms of PE, with very serious imprecision consistent with both benefit and harm. For the outcome of thrombocytopaenia, a possible clinical benefit of LWMH was found, however there was also very serious imprecision consistent with no difference and harm. The evidence was very low quality due to risk of bias and imprecision.

Combination prophylaxis versus combination prophylaxis or single-prophylaxis agents

AES

One study compared below knee AES in combination with UFH to below knee AES alone (n=163). There was no clinical difference between the interventions for both all-cause mortality and PE, however there was very serious imprecision for both outcomes consistent with both benefit and harm. The evidence was very low quality due to risk of bias and inconsistency.

Above knee AES in combination with UFH was compared to UFH alone. One study reported the outcomes all-cause mortality and fatal PE (n=160–176). A possible clinical harm was found for the combination intervention in terms of all-cause mortality, however there was very serious imprecision around the estimate, and therefore this was also consistent with no difference and benefit. A possible clinical benefit was seen for the combination in terms of fatal PE, however again there was very serious imprecision consistent with no difference and harm. Two studies reported the outcomes DVT and PE (n=336). There was a clinical benefit of the combination intervention in terms of DVT, and a possible clinical benefit in terms of PE, although this outcome estimate had very serious imprecision and was consistent with no difference and harm. The evidence ranged from very low to moderate quality due to risk of bias and imprecision.

One study compared below knee AES in combination with UFH to UFH alone (n=174). The evidence showed no clinical difference for all-cause mortality or PE. Both outcomes had very serious imprecision around the estimate and therefore were also consistent with both benefit and harm. The evidence was very low quality due to risk of bias, imprecision and, for the PE outcome, indirectness.

One study compared the combination of above knee AES and full leg IPCD with above knee AES alone (n=77). There was a possible clinical benefit of the combined interventions for DVT, however there was very serious imprecision around the estimate and this was therefore also consistent with no difference and harm. There was no clinical difference in terms of PE, however there was very serious imprecision consistent with both benefit and harm. The evidence was very low quality due to risk of bias and imprecision.

One study compared AES at an undefined length in combination with full leg IPCD to AES alone (n=108). There was a possible clinical benefit of the combined interventions in terms of DVT, however there was serious imprecision consistent with no difference. There was no clinical difference in terms of PE, with very serious imprecision around the estimate, consistent with both harm and benefit. The evidence ranged from very low to low quality due to risk of bias and imprecision.

One study compared AES at an undefined length in combination with full leg IPCD to UFH alone (n=100). There was a possible clinical benefit of the combined intervention in terms of DVT, however there was very serious imprecision around the estimate and therefore was also consistent with no difference and harm. No other outcomes were reported. The evidence was very low quality due to risk of bias and imprecision.

One study compared AES at an undefined length in combination with full leg IPCD to electrical stimulation alone (n=100). There was a possible clinical benefit of the combined intervention in terms of DVT, however there was serious imprecision around the estimate consistent with no difference. No other outcomes were reported. The evidence was low quality due to risk of bias and imprecision.

Foot impulse device

One study compared the combination of FID, below knee IPCD and low dose LMWH to the combination of FID and below knee IPCD. A possible clinical benefit was found for both DVT and PE, however with very serious and serious imprecision around the estimates. No clinical difference was found in terms of thrombocytopaenia, however there was very serious imprecision consistent with both benefit and harm. The evidence was very low to low quality due to risk of bias, imprecision and, for the DVT outcome, indirectness.

IPCD

Two studies compared IPCD at an undefined length in combination with standard dose LMWH with IPCD at an undefined length alone (n=334). The evidence showed a clinical benefit of the combination intervention in terms of DVT. There was no clinical difference in terms of PE, however there was very serious imprecision around the estimate for this outcome, and therefore was consistent with both benefit and harm. The evidence ranged from very low to low quality due to risk of bias, imprecision, and for the PE outcome, indirectness.

LMWH

One study compared standard dose and extended duration LMWH in combination with AES at an undefined length, to standard dose and standard duration LMWH in combination with AES at an undefined length (n=343–427). There was a possible clinical harm of the extended duration LMWH combination in terms of all-cause mortality, however there was very serious imprecision around the estimate and so this was also consistent with benefit and no difference. There was a possible clinical benefit for both DVT and PE. Both outcomes also had serious and very serious imprecision around the estimate. There was no clinical difference in terms of fatal PE. This outcome had very serious imprecision around the estimate consistent with both harm and benefit. The evidence ranged from very low to low quality due to risk of bias and imprecision.

Fondaparinux

One large study compared fondaparinux in combination with IPCD at an undefined length, to IPCD at an undefined length alone (n=842–1285). There was a possible clinical harm of the fondaparinux + IPCD combination in terms of all-cause mortality, however there was very serious imprecision around the estimate and therefore this was also consistent with benefit and no difference. There was a clinical benefit of the combined intervention in terms of DVT, and a possible benefit in terms of PE, although this was also consistent with no difference and clinical harm. There was no clinical difference in terms of fatal PE, although due to very serious imprecision around the estimate this was also consistent with both benefit and harm. The evidence ranged from very low to moderate quality due to risk of bias and imprecision.

One study compared fondaparinux in combination with UFH and mechanical prophylaxis (AES and IPCD), to standard dose LMWH in combination with UFH and mechanical prophylaxis (AES and IPCD) (n=258–298). There was a possible clinical benefit of the fondaparinux combination intervention in terms of PE, however there was very serious imprecision consistent with no difference and clinical harm. There was a possible clinical harm in terms of major bleeding, however there was very serious imprecision around the estimate, and therefore was also consistent with no difference and benefit. The evidence was very low quality due to risk of bias and imprecision.

Network meta-analysis statements

DVT (symptomatic and asymptomatic)

48 studies were included in the network meta-analysis (NMA) for the outcome of DVT (symptomatic and asymptomatic), involving 22 treatments. Treatments included no VTE prophylaxis, pharmacological and mechanical interventions as single agents as well as combination interventions of both pharmacological and mechanical interventions. Results from the network meta-analysis presented LMWH at a standard dose for a standard duration initiated post-operatively in combination with IPCD, fondaparinux in combination with IPCD, and AES (above-knee) in combination with IPCD (full leg) as the most clinically effective interventions in terms of the outcome of DVT (symptomatic and asymptomatic). The least clinically effective interventions were no prophylaxis, VKA and LMWH at a low dose for a standard duration initiated pre-operatively. One inconsistency was identified when relative risk values from pairwise meta-analyses were compared with relative risk values from the NMA. There was also a considerable amount of uncertainty around the rank-point estimates with considerably wide credible intervals.

PE

26 studies were included in the NMA for the outcome of PE, involving 13 treatments. Treatments included no VTE prophylaxis, pharmacological and mechanical interventions as single agents as well as combination interventions of both pharmacological and mechanical interventions. Results from the network meta-analysis presented LMWH at a standard dose for an extended duration initiated pre-operatively, AES (above knee), LMWH at a standard dose for a standard duration initiated by post-operatively as the most clinically effective interventions in terms of the outcome of PE. The least clinically effective interventions were IPCD (full leg), fondaparinux and IPCD (below knee). No inconsistencies were identified when relative risk values from pairwise meta-analyses were compared with relative risk values from the NMA. There was also a high amount of uncertainty around the rank-point estimates with very wide credible intervals.

Major bleeding

24 studies were included in the NMA for the outcome of major bleeding, involving 15 treatments. Treatments included no VTE prophylaxis and pharmacological interventions (mechanical interventions were combined with no prophylaxis as the assumption was made that these interventions do not contribute to bleeding risk). Results from the network meta-analysis presented no prophylaxis, LMWH at a low dose for a standard duration initiated pre-operatively and UFH as the most clinically effective interventions in terms of major bleeding. The least clinically effective interventions were LMWH at a high dose for a standard duration initiated pre-operatively, fondaparinux and LMWH at a standard dose for a standard duration initiated post-operatively. One inconsistency was identified when relative risk values from pairwise meta-analyses were compared with relative risk values from the NMA. There was also a high amount of uncertainty around the rank-point estimates with considerably wide credible intervals across a majority of the interventions.

Economic

One cost-utility analysis found that for VTE prophylaxis:
- In low risk general surgery patients, LMWH (standard dose, standard duration) + thigh-length AES was cost effective compared to LMWH (standard dose, standard duration) alone (ICER: £2,632 per QALY gained)
- In intermediate and high risk general surgery patients, LMWH (standard dose, standard duration) + thigh-length AES was dominant (less costly and more effective) compared to LMWH (standard dose, standard duration) alone
This analysis was assessed as directly applicable with potentially serious limitations
One cost-utility analysis found that in people admitted for general surgery AES was the most cost-effective intervention (having the highest incremental net monetary benefit [INMB]) compared to no prophylaxis (INMB: £488). This analysis was assessed as partially applicable with potentially serious limitations.
One cost-utility analysis found that post-discharge LMWH (standard dose) was cost effective (INMB: £49) compared to no post-discharge prophylaxis in patients admitted for general surgery. This analysis was assessed as directly applicable with potentially serious limitations.

35.6. Recommendations and link to evidence

Recommendations	1.5.37. Offer VTE prophylaxis to people undergoing abdominal (gastrointestinal, gynaecological, urological) surgery who are at increased risk of VTE. For people undergoing bariatric surgery, follow recommendations 1.5.41–1.5.43.[2018] 1.5.38. Start mechanical VTE prophylaxis on admission for people undergoing abdominal surgery. Choose either: anti-embolism stockings or intermittent pneumatic compression. [2018] Continue until the person no longer has significantly reduced mobility relative to their normal or anticipated mobility. [2018] 1.5.39. Add pharmacological VTE prophylaxis for a minimum of 7 days for people undergoing abdominal surgery whose risk of VTE outweighs their risk of bleeding, taking into account individual patient factors and according to clinical judgement. Choose either: LMWH^aa or fondaparinux sodium^bb. [2018] 1.5.40. Consider extending pharmacological VTE prophylaxis to 28 days postoperatively for people who have had major cancer surgery in the abdomen. [2018]
Research recommendation	None
Relative values of different outcomes	The committee considered all-cause mortality (up to 90 days from hospital discharge), deep vein thrombosis (symptomatic and asymptomatic) (up to 90 days from hospital discharge), pulmonary embolism (symptomatic and asymptomatic) (up to 90 days from hospital discharge), fatal PE (up to 90 days from hospital discharge), and major bleeding (up to 45 days from hospital discharge) as critical outcomes. The committee considered clinically relevant non-major bleeding (up to 45 days from hospital discharge), health-related quality of life (up to 90 days from hospital discharge), heparin-induced thrombocytopaenia (duration of study), and technical complications of mechanical interventions (duration of study) as important outcomes. Please see section 4.4.3 in the methods chapter for further detail on prioritisation of the critical outcomes.
Quality of the clinical evidence	Sixty-seven randomised controlled trials were included in this review. Sixty-two of these were included in the previous guideline (CG92). Five new studies were added to the review. A total of thirty-nine comparisons were included in this review, evaluating the use of pharmacological (UFH, LMWH, VKA and fondaparinux) and mechanical (AES, IPCD, foot pump, FID and electrical stimulation) interventions for VTE prophylaxis. For the majority of evidence in this review, the quality ranged from a GRADE rating of moderate to very low. This was due to a lack of blinding, presence of selection bias, incomplete outcome reporting due to the high number of drop outs in some included studies, and use of inadequate or unreported method of measurement, resulting in a high or very high risk of bias rating. Furthermore, much of the evidence in the review had serious or very serious imprecision, leading to further downgrading to the quality of evidence. A high quality GRADE rating was seen for one outcome, in the fondaparinux versus no prophylaxis/mechanical prophylaxis comparison, for the DVT outcome.
Trade-off between clinical benefits and harms	The committee noted that the review contains both open and laparoscopic surgery populations, and that these populations were likely to have different mobilisation times and associated risks. The committee discussed creating separate recommendations for these populations but recognised that it would be difficult to align a distinction in recommendations in line with the risk assessment recommendations, given that not all laparoscopic procedures are under 90 minutes, and given the fact that many of the included studies did not separate the two populations as they either used a mix of laparoscopic and open surgery procedures, or did not specify the type of procedure used. Mechanical prophylaxis The committee noted that there was no evidence for foot impulse devices as a standalone intervention and therefore a positive recommendation for the use of this intervention for VTE prophylaxis could not be made. The committee also discussed the evidence for the use of AES. It was considered that while there was no convincing evidence that above knee AES was better than below knee, the economic evidence suggested a slight benefit for above knee AES. Therefore, the committee agreed there was insufficient evidence to specify one particular option of above or below knee AES in the recommendations. In terms of IPCD the committee discussed the practical considerations that need to be taken into account with respect to mobilising the patient. IPCD are usually used only during the surgery. Mechanical prophylaxis is recommended until the patient is back to normal mobility as the committee believe that mechanical prophylaxis offers little benefit once a patient is mobile. Pharmacological prophylaxis The committee considered the evidence for pharmacological prophylaxis. The committee noted that there was evidence to support LMWH and fondaparinux as being better than no prophylaxis. However there was not sufficient evidence to determine whether LMWH was better than fondaparinux. For prevention of DVT the evidence suggested that pharmacological prophylaxis (LMWH or fondaparinux) in combination with IPCD may be of most clinical benefit. The network meta-analysis (NMA) conducted showed that combination prophylaxis strategies with pharmacological and mechanical interventions are more clinically beneficial in terms of reducing DVT. These combination strategies had higher rankings compared to pharmacological or mechanical interventions as standalone interventions, particularly LMWH at a standard dose for a standard duration initiated post-operatively in combination with IPCD which was ranked as the most clinically effective prophylaxis in the NMA for DVT. Pharmacological prophylaxis is recommended for a minimum of 7 days because the average duration of trials was between 7 and 10 days. The committee agreed this should be extended to 28 days for cancer surgery because the evidence identified was for this duration.
Trade-off between net clinical effects and costs	Two economic studies were included in this review. One is an economic evaluation recently published as part of an HTA funded study. This was assessed as directly applicable with minor limitations. The other was the economic model previously developed for CG92 which covered two comparisons; one for standard duration prophylaxis options and the second for post-discharge prophylaxis. The model comparing standard duration prophylaxis options was assessed as partially applicable with potentially serious limitations. The model for post-discharge prophylaxis was assessed as directly applicable with potentially serious limitations. Additionally, four studies were selectively excluded; one was excluded due to methodological limitations, three (including the model developed for CG46) were selectively excluded due to the availability of the more applicable included studies. The first of the two included studies was an economic model that compared above and below knee AES; each combined with LMWH (standard dose and standard duration), vs LMWH alone. The results were presented for three levels of baseline risk of VTE: high, intermediate and low. For people at high or intermediate risk of VTE, LMWH + thigh-length AES was the dominant option. For people at low risk, LMWH + thigh-length AES was the cost effective option with an ICER of £2,632 per QALY gained compared to LMWH alone. Two models were developed in CG92. The first was for standard duration prophylaxis and included the following interventions: AES, IPCD-FID, UFH (standard dose)+AES, LMWH (standard dose)+ AES, LMWH (standard dose), Aspirin (high dose), UFH (standard dose), Fondaparinux+ IPCD-FID, Fondaparinux, VKA (variable dose), UFH (standard dose) + Aspirin (high dose), and no prophylaxis. The committee noted that not all of these interventions are still relevant to current practice (for example aspirin [high dose] and VKA). Mechanical prophylaxis with either AES or IPCD were the most cost effective options in the base case analysis with INMB of £488 and £464 respectively. However in a two-way sensitivity analysis that varied the baseline risk of PE and MB, combined prophylaxis of LMWH+ stocking was the most cost-effective option for high baseline risk of PE and low risk of major bleeding. The second model compared post-discharge prophylaxis with LMWH with no prophylaxis. The results showed that extending the duration of LMWH prophylaxis to continue post-discharge was cost effective compared to no prophylaxis with an INMB of £49. The committee considered the economic evidence presented, alongside the clinical evidence. The committee noted that, in line with CG92 recommendation, combined prophylaxis for people at high risk of VTE is the most cost effective option. This was supported by the newly published HTA report that stratified surgical patients according to their level of VTE risk; where combined prophylaxis was the most cost effective option. The committee considered the recent clinical evidence and determined that both LMWH and fondaparinux were better compared to no prophylaxis; however, no clear conclusion could be made in terms of superiority of one over the other. However, as low quality clinical evidence for the DVT outcome suggested superiority of fondaparinux, the committee considered that this would justify the increased cost, and the choice of either as pharmacological prophylaxis options should be made based on the baseline bleeding risk. The committee discussed whether the evidence was enough to recommend either knee or thigh length AES. The economic evidence supported the cost effectiveness of combined prophylaxis that includes thigh length AES, however the committee noted that thigh length AES are less convenient for people to wear and are more difficult to fit. Hence, the committee agreed that the choice of the length of stocking should be made taking into account the preference of the individual and his/her ability to adhere to wearing them. No studies were identified that compared thigh versus knee length for IPCD, so the committee considered that, similar to AES, the choice of the length should be based on preference, likelihood of adherence and ease of fitting. The committee also discussed the duration of prophylaxis and noted that the economic model developed for CG92 supported extending the duration of prophylaxis for those who are at increased risk of VTE. These were primarily people undergoing surgeries for cancer. For this population, continuing LMWH post discharge was found to be more cost effective than no post-discharge prophylaxis.
Other considerations	None.

Footnotes

aa: At the time of publication (March 2018), LMWH did not have a UK marketing authorisation for use in young people under 18 for this indication. The prescriber should follow relevant professional guidance, taking full responsibility for the decision. Informed consent should be obtained and documented. See the General Medical Council’s Prescribing guidance: prescribing unlicensed medicines for further information.
bb: At the time of publication (March 2018), fondaparinux sodium did not have a UK marketing authorisation for use in young people under 18 for this indication. The prescriber should follow relevant professional guidance, taking full responsibility for the decision. Informed consent should be obtained and documented. See the General Medical Council’s Prescribing guidance: prescribing unlicensed medicines for further information.

Bookshelf ID: NBK561781

Contents

< Prev Next >

PubReader
Print View
Cite this Page
National Guideline Centre (UK). Venous thromboembolism in over 16s: Reducing the risk of hospital-acquired deep vein thrombosis or pulmonary embolism. London: National Institute for Health and Care Excellence (NICE); 2018 Mar. (NICE Guideline, No. 89.) 35, Abdominal surgery (excluding bariatric surgery)
PDF version of this page (1.2M)
Disable Glossary Links

Abdominal surgery (excluding bariatric surgery) - Venous thromboembolism in over...
Abdominal surgery (excluding bariatric surgery) - Venous thromboembolism in over 16s

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

Venous thromboembolism in over 16s: Reducing the risk of hospital-acquired deep vein thrombosis or pulmonary embolism.

35Abdominal surgery (excluding bariatric surgery)

35.1. Introduction

35.2. Review question: What is the effectiveness of different pharmacological and mechanical prophylaxis strategies (alone or in combination) for people undergoing abdominal surgery (gastrointestinal, gynaecological, urological)?

35.3. Clinical evidence

35.4. Economic evidence

Published literature

35.5. Evidence statements

Clinical

Pairwise meta-analysis statements

Mechanical prophylaxis versus mechanical prophylaxis

AES

Foot pump

IPCD

Pharmacological prophylaxis versus pharmacological prophylaxis

UFH

LMWH (low dose)

LMWH (standard dose)

LMWH (high dose)

Fondaparinux

VKA

Mechanical prophylaxis versus pharmacological prophylaxis

Pharmacological prophylaxis versus mechanical prophylaxis

Combination prophylaxis versus combination prophylaxis or single-prophylaxis agents

AES

Foot impulse device

IPCD

LMWH

Fondaparinux

Network meta-analysis statements

DVT (symptomatic and asymptomatic)

PE

Major bleeding

Economic

35.6. Recommendations and link to evidence

Footnotes

Views

In this Page

Other titles in this collection

Related NICE guidance and evidence

Recent Activity