Article Text


Study protocol: first nationwide comparative audit of acute lower gastrointestinal bleeding in the UK
  1. Kathryn Oakland1,
  2. Richard Guy2,
  3. Raman Uberoi3,
  4. Frances Seeney1,
  5. Gary Collins4,
  6. John Grant-Casey1,
  7. Neil Mortensen2,
  8. Mike Murphy1,
  9. Vipul Jairath5,6
  10. on behalf of the National Comparative Audit in Blood Transfusion Lower Gastrointestinal Bleeding Working Group
  1. 1NHS Blood and Transplant, UK
  2. 2Department of Colorectal Surgery, Oxford University Hospitals, Oxford, UK
  3. 3Department of Interventional Radiology, Oxford University Hospitals, Oxford, UK
  4. 4Nuffield Department of Orthopaedics, Centre for Statistics in Medicine, Botnar Research Centre, University of Oxford, Oxford, UK
  5. 5Department of Medicine, Division of Gastroenterology, Western University and London Health Sciences Centre, Ontario, Canada
  6. 6Nuffield Department of Medicine, University of Oxford, Oxford, UK
  1. Correspondence to Kathryn Oakland; Kathryn.Oakland{at}


Introduction Acute lower gastrointestinal bleeding (LGIB) is a common indication for emergency hospitalisation worldwide. In contrast to upper GIB, patient characteristics, modes of investigation, transfusion, treatment and outcomes are poorly described. There are minimal clinical guidelines to inform care pathways and the use of endoscopy, including (diagnostic and therapeutic yields), interventional radiology and surgery are poorly defined. As a result, there is potential for wide variation in practice and clinical outcomes.

Methods and analysis The UK Lower Gastrointestinal Bleeding Audit is a large nationwide audit of adult patients acutely admitted with LGIB or those who develop LGIB while hospitalised for another reason. Consecutive, unselected presentations with LGIB will be enrolled prospectively over a 2-month period at the end of 2015 and detailed data will be collected on patient characteristics, comorbidities, use of anticoagulants, transfusion, timing and modalities of diagnostic and therapeutic procedures, clinical outcome, length of stay and mortality. These will be audited against predefined minimum standards of care for LGIB. It is anticipated that over 80% of all acute hospitals in England and some hospitals in Scotland, Wales and Northern Ireland will participate. Data will be collected on the availability and organisation of care, provision of diagnostic and therapeutic GI endoscopy, interventional radiology, surgery and transfusion protocols.

Ethics and dissemination This audit will be conducted as part of the national comparative audit programme of blood transfusion through collaboration with specialists in gastroenterology, surgery and interventional radiology. Individual reports will be provided to each participant site as well as an overall report and disseminated through specialist societies. Results will also be published in peer-reviewed journals. The study has been funded by National Health Services (NHS) Blood and Transplant and the Bowel Disease Research Foundation and endorsed by the Association of Coloproctology of Great Britain and Ireland.

Statistics from

Strengths and limitations of this study

  • This is the first nationwide audit of lower gastrointestinal bleeding (LGIB) and is likely to be the largest prospective observational study of LGIB of its kind to date.

  • All aspects of care throughout the patient journey will be described and audited, allowing detailed evaluation of many components of care.

  • Inclusion of hospitals based on routine admission of LGIB patients as opposed to size or location makes this audit representative of care in the UK as a whole, and therefore the results are widely applicable.

  • Although case ascertainment and data collection are prospective, this study relies on accurate record-keeping in patients' notes and electronic records, which may be unreliable.


Acute lower gastrointestinal bleeding (LGIB) is traditionally defined as bleeding arising distal to the ligament of Treitz, accounts for 20% of all hospitalisations for GI haemorrhage in the UK1 and has a crude incidence of 87/100 000.2 While the source of bleeding is not always apparent after presentation, it can further be considered to arise from either the mid-GI tract (between the Treitz angle and the ileocaecal valve) or from the colon (between the ileocaecal valve and the rectum). Population-based data from Europe suggest the incidence is rising and mortality rates may be as high as those for upper GIB (UGIB).3 Bleeding can arise from multiple sources such as diverticula, haemorrhoids, polyps, colorectal cancer, intestinal ischaemia, colitis and angiodysplasia.4 Risk factors for bleeding include increasing age,4 as well as the use of antiplatelet medications, anticoagulants3 and non-steroidal anti-inflammatory drugs (NSAIDs).5

The spectrum of disease leading to hospitalisation can range from trivial and self-limiting bleeding through to catastrophic, life-threatening haemorrhage requiring emergency intervention with mesenteric embolisation or surgery. There are few studies reporting mortality. In a population-based study, the mortality was found to be 1.2%.2 A sample of an American national hospitalisation database estimated in-hospital mortality at 3.9%4, whereas a sample of Spanish hospitals estimated mortality from any lower GI event to be 8.8%.6

LGIB is also a common indication for the transfusion of red blood cells (RBCs). A multicentre study from the North of England suggested that 17% of RBCs were transfused for GIB.7 This is relevant given the recent randomised evidence that the liberal use of RBCs after UGIB may be associated with harm.8

Unlike UGIB, there are few large studies providing detailed information on patient characteristics, transfusion and pathways of care in LGIB. The approach to diagnosis and intervention in terms of the use of endoscopy or radiology is uncertain and there is likely to be considerable variation in practice. This is reflected in the almost complete absence of national or international guidelines for LGIB, compared with at least four high-profile guidelines for UGIB.9–12

Identifying the source of bleeding following presentation with LGIB poses a diagnostic challenge. Flexible sigmoidoscopy and colonoscopy may enable direct visualisation of the bleeding point, but this may be limited by poor bowel preparation in the acute setting. Although urgent lower GI endoscopy (within 12 hours) may be more likely to identify a source, there may be little associated beneficial impact on clinical outcome or length of stay.13 ,14 Endoscopic therapy using chemical or mechanical haemostatic agents is becoming increasingly sophisticated, but it is not known whether these are routinely used for LGIB, as they are for UGIB or their effectiveness.

Increasingly a bleeding source may also be identified using computerised tomographic angiography (CTA) or mesenteric angiography (MA). If active extravasation of contrast is visualised on angiography, mesenteric embolisation offers a minimally invasive method to control haemorrhage avoiding the need for surgery. Although there is potential risk of developing associated colonic ischaemia after embolisation, the development of super-selective embolisation may to reduce this.15 ,16 Whether this has resulted in a reduction in the requirement for major abdominal surgery and its associated complications is not known.

In 2015, the National Confidential Enquiry into Patient Outcome and Death (NCEPOD) in the UK conducted a national audit of all hospitalised patients with severe GIB (defined as those that received ≥4 units red cells).17 Significant opportunities to improve care were identified, and recommendations to end the traditional separation of UGIB and LGIB were made.17 It also highlighted the need for research in LGIB and endorsed the development of risk stratification methods relevant to all GIB.

Providing a comprehensive interventional radiology or endoscopic therapy service poses a significant demand on resources. Many units in the UK are still not able to provide 24/7 emergency care,18 a problem that has been exacerbated by the recent vascular configuration.

This may mean that patients are being transferred between hospitals for definitive treatment, when indicated. There are no contemporary data on the number of acute hospitals providing access to emergency interventional radiology and lower GI endoscopy. The associated impact on patient access to these services is unknown.


The overall objective of this nationwide audit is to characterise the clinical characteristics, management strategies and outcomes of patients with acute LGIB presenting to UK hospitals. Specific objectives include the following:

  1. Description of the use of inpatient investigations (lower GI endoscopy, CT, interventional radiology, nuclear medicine and surgery) and their associated diagnostic yield (including factors associated with failed investigation), comparing in and out of hours availability and demand, complications and effect on length of stay, readmissions, morbidity and mortality.

  2. Evaluation of therapeutic modalities (endoscopic haemostasis, embolisation and surgery) focusing on indication, availability and therapeutic yield with regard to rebleeding, need for further procedures and the associated impact on outcomes.

  3. Quantification of blood product transfusion in comparison with established national guidelines and protocols.19 ,20

  4. Description of the management and current treatment strategies for patients on long-term anticoagulants who develop LGIB.

  5. Identification of institutional and patient-specific risk factors for poor outcome to aid the triage of patients presenting with LGIB.

Reporting contemporaneous data on presenting characteristics, requirement of inpatient investigation and success of treatment will allow the future development of guidelines on the optimal management of LGIB with the aim of improving patient care, reducing variation in practice and ultimately improving outcomes.

Methods and analysis

This is a UK-wide, prospective audit of all admissions presenting with, or developing LGIB while an established inpatient. Hospitals will be recruited from September 2015. Case ascertainment will last for 2 months and all data must be submitted by the end of January 2016.

National Health Services (NHS) Blood and Transplant has an established audit programme that regularly conducts national projects examining the use of blood products within the UK. These audits are used to examine current practice in comparison with established guidelines and have led to many successful projects across therapeutic areas. As well as comparing blood management with national protocols these audits present an opportunity to compare practice in other aspects of clinical care, such as best practice in perioperative and medicines management.

The cases

The audit will include all unselected patients that present with LGIB that results in an admission to hospital or develops while patients are admitted for another reason. Cases will be identified using presenting symptoms as opposed to examination findings or discharge diagnoses, and thus will include mid-GIB as well as bleeding distal to this, since presenting signs and symptoms will be similar. Cases will be eligible if they fulfil the following criteria: age ≥16 years, history of bright or dark blood per rectum, maroon coloured stool or blood mixed in with stool, clots per rectum or passage of melaena without haematemesis.

Melaena without haematemesis is included so that cases of small bowel bleeding are unlikely to be missed. Previous reports have shown that it can be difficult to distinguish upper from lower GI sources of bleeding;16 hence, to optimise the identification of LGIB, the inclusion criteria are deliberately broad. This means that a small number of patients with UGIB may be captured in the data set, but this is reflective of the uncertainties that may exist in routine clinical care. There are two opportunities in the patient questionnaire to indicate that a UGIB case has been included; if the patient has an endoscopy that identifies the source of bleeding to be proximal to the ligament of Treitz the data collector can select that the source of LGIB was from the upper GI tract, or can indicate that there is not enough data to determine whether the case is a true case of LGIB. The data from these patients will be collected centrally and will undergo the same cleaning protocol as for LGIB patients, but will be excluded from any analysis specific to LGIB.

We aim to identify all cases of LGIB within a 2-month period, starting on 1 September 2015. Every identified case or potential case must be registered for inclusion. We are aiming to identify at least 1000 cases of acute LGIB. This estimate is based on the UK population incidence of LGIB and the benchmarked against the number of cases that were recruited in the 2007 national audit of UGIB and the use of blood.21

Data will be collected until discharge/transfer from hospital, death or up to day 28 (whichever occurs first). Readmission data will be collected until up to 28 days post discharge. This means that some follow-up data will continue to be collected after the ascertainment period.

Recruitment of sites

All NHS trusts in England admitting acute surgical and medical admissions will be contacted directly and invited to participate. Letters and emails explaining the rationale and aims will be sent to the medical director, chief executive, Clinical Audit Department and the haematologist with primary responsibility for transfusion, as well as transfusion practitioners within each acute hospital. Medical directors will be asked to give permission for their hospital to participate and to provide the contact details of their clinical lead for surgery. The clinical lead will then be provided with information about the methodology and timeline of the audit and asked to nominate a local audit lead to coordinate the project. Non-responders will be sent two further reminder letters. If there is no response after three formal requests, it will be assumed that the hospital will not be participating.

This study will be advertised to NHS hospitals in Scotland, Wales and Northern Ireland via their national blood services. Independent hospitals will not be invited to participate since GI bleeds are predominantly managed in the NHS. As indicated in June 2015, there were 140 eligible NHS trusts in England, and we aim to recruit 80% of these.

Data collection

Two broad categories of data will be collected; organisational and individual patient data.

Organisation data

Organisational data will record the availability of services for the investigation and treatment of LGIB. This will be available as a paper questionnaire and an electronic survey. Outcomes include the in-hour and out-of-hours availability of endoscopy, interventional radiology and surgery. Data on how patients access these investigations and treatment in hospitals without onsite services will be collected. The provision of massive transfusion protocols and GIB guidelines will be established (table 1). Each hospital will complete one copy of this questionnaire.

Table 1

Audit standards and associated specific outcomes within the organisational variables

Individual patient data

Patient data will include the clinical characteristics and outcomes of patients with acute LGIB. The data collection includes questions on clinical examination findings, the timing use and results of endoscopy, radiology and surgery, the prevalence of different aetiologies of LGIB and the use, timing and volume of blood products. Outcomes will include length of stay, in-hospital morbidity and mortality, readmission rates, rebleeding rates and transfusion requirements. Data on anticoagulation will be collected, looking at methods of reversal used, and whether national protocols have been followed (table 2). All data will be obtained prospectively from patient notes and electronic hospital records.

Table 2

Audit standards and specific outcomes within the patient variables

The clinical details for each patient identified will be entered into an online questionnaire, which is accessed by a site-specific, password-protected website. Entry of data from each case will take 20–40 min to complete depending on its complexity. Paper versions of the questionnaire will also be posted to sites to facilitate the collection of data for those sites with limited computer access. Cases and sites will be given a unique code to enable data entry without using any patient or hospital identifiers. Each participating hospital will be given a unique login and password to ensure data integrity. No patient identifiers will be collected at any time.

The website automatically downloads all data into a central database regardless of whether the site has indicated that the data are complete. This allows monitoring of the participants' progress and regular counts of the registered cases. Once the site is content that it has entered a complete data set, a tick box finalises the data set. This then alerts the central team that the data entry for the case is finished, and the data set will be checked for any missing mandatory data or nonsensical responses. Audit leads within each hospital will be contacted to provide additional or corrected data where necessary. This will happen on a daily basis throughout and after the study period to ensure data are as complete as possible. To ensure contemporaneous data collection, while the study is live, the project group will also review any cases that are incomplete but inactive for more than 1 week and contact the hospital lead to encourage their completion.

A team consisting of an audit lead, case identifier and several data enterers will collect the data in each NHS trust. The audit lead will ensure that cases are being identified and entered and that the data are complete and accurate. We expect that the leads will predominantly be colorectal or general surgical consultants or registrars, although they may be from any specialty. The audit lead will be responsible for coordinating the audit in their hospital, working with the case identifier and supporting the case enterers.

Questionnaire design

The questionnaires were piloted at 10 potentially eligible sites in the UK. Each site was asked to review the questionnaires and record feasibility of data collection for each question via a standardised grading system.

Seven sites returned the organisational questionnaire pilot and all but two questions were answered as expected. The questions found to be difficult to complete asked for a recording of the availability of guidelines, which were uniformly unanswered. On review, it was decided that the data collected by these questions were non-essential and time-consuming. These questions were removed from the data set.

Six hospitals were asked to identify and complete patient-specific questionnaires on five cases of LGIB. All mandatory questions were deemed feasible and accessible. The remainder of the questions were reviewed and clarified. No questions were excluded. Wording and phrasing was amended for questions deemed ambiguous based on the pilot exercise. Answers were reviewed to ensure data were interpretable and reproducible.

Case identification

There are no hospital diagnostic codes specific to LGIB. Methods aimed at identifying LGIB cases by mapping to ‘classification trees’ using codes such as the International Classification of Disease (ICD)30 have been shown to have varying performance.31 A previous large prospective audit on UGIB successfully identified cases by contacting clinical teams.21 As referral pathways may differ between hospitals, it can be difficult to create a standardised method that is reproducible nationally.

To establish a pattern of hospital admission locations for patients with acute LGIB, five hospitals (including a tertiary referral centre for interventional radiology and a small district general hospital) were asked to describe their referral pathways and pilot the process of case identification. Eleven potential departments and wards were identified as likely to accommodate patients with LGIB. Over a 2-week period, each hospital was instructed to contact each location multiple times to identify locations with the highest and lowest case yield (table 3).

Table 3

Pilot case identification tool

Feedback on ease of case identification, time spent and suggestions for other locations were collected. Of the five hospitals, only one site was able to provide data for the complete time period, identifying 28 cases of LGIB. The low response rate of the other hospitals indicates that this kind of case ascertainment is not reproducible or reliable. A recent national audit of severe GI haemorrhage demonstrated that unlike UGIB, which may present to a range of departments and specialities, LGIB presents to a more limited selection of locations,17 namely surgery, gastroenterology and general medicine wards. This was also demonstrated by the 28 cases identified here; all but 1 case was identified by daily contact with the admitting surgical team and acute medical admissions unit. To maximise case ascertainment in this national study, audit leads will be asked to have daily contact with surgical admission units and the surgical on-call team, daily contact with medical admission units and on-call team and visits to the gastroenterology wards three times per week.

Data analysis

Once all datasets are indicated as finished by the local site, checked for any missing data and incorrect entries amended they will be downloaded into one unifying database. Any duplicates will be removed. Variations in spelling of drug names, abbreviations and treatments will be standardised.

Although most questions require a single fixed response, there are several with an ‘other’ option. Where appropriate, these will be recoded as one of the other fixed responses or compiled into an appendix. The question asking for the documented cause of the bleeding is a free-text box. Where possible, this will be mapped to the ICD-10: classification of diseases of the digestive system.30 Any responses not fitting this classification will be compiled into an appendix. Any diagnoses that pertain to UGIB will be flagged.

Data will be collected on several baseline comorbidities, including those listed in the Deyo modification32 of the Charlson Comorbidity Index.33 The Charlson index has been used in administrative data sets, but its application to clinical data is more difficult as some of the definitions are subjective. To enable its use in a clinical setting, we made the following amendments on pragmatic clinical grounds: (1) mild and moderate liver disease was stratified into non-cirrhotic and cirrhotic, respectively, for ease of categorisation using medical notes; (2) congestive cardiac failure is usually classified by the New York Heart Association criteria34 but the criteria may not be reproducible in a review of surgical notes. This was changed to include patients on pharmacotherapy or with clinical examination findings consistent with heart failure; (3) peptic ulcer disease was classified by the use of pharmacological acid suppression; (4) renal disease was reclassified as chronic kidney disease stage 2–3 and stage 4 to represent moderate and severe respectively.35 A Charlson Comorbidity Index will be calculated for each case. A retrospective review of a national database showed that a Charlson index ≥2 was independently associated with in-hospital mortality in patients admitted with LGIB.4

The cases identified as UGIB will then be excluded from any further analysis. Audit standards applied to the remaining LGIB cohort, but cases will be grouped, where relevant, to allow comparative analysis particularly focusing on risk factors for poor outcome. Proposed subgroups include established inpatients and de novo presentations, transferred and non-transferred patients and groups stratified by Charlson comorbid status.

Calculating the hospital resources required by patients admitted with acute LGIB requires estimates of bed occupancy and frequency of inpatient and outpatient investigation and treatment. Hospital bed requirements will be described using data on length of stay, new discharge to a nursing home or rehabilitation facility and readmission rates. The type, frequency and waiting time for investigations will be calculated and comparisons by type of investigation will be made. Length of stay for patients who undergo inpatient treatment (as well as investigation) will be calculated in comparison with those that do not. The aim is to identify investigations and treatments associated with reduced length of stay, rebleeding rates and need for transfusion.

The draft tables for the analysis are included in online supplementary appendix 1.

Audit standards

The development of audit standards using existing guidelines is limited by the lack of national guidance. The most relevant guidelines that include LGIB are the Scottish Intercollegiate Guidelines Network (SIGN) guidelines.10 As there is no National Institute for Clinical Excellence (NICE) equivalent guideline for LGIB, these have been adopted where appropriate. The NCEPOD report on GIB17 also made recommendations on LGIB, and these have also been included. Where guidelines on specific aspects of the management of LGIB do not exist, British Society of Gastroenterology and NICE guidelines on the management of UGIB9 have been interchanged as the auditable standard, as appropriate. The British Committee for Standards in Haematology26 ,27 and NICE guidelines on the use of RBCs, platelets and fresh frozen plasma28 have been used as standards for transfusion. Recommendations made by the Association of Surgeons of Great Britain and Ireland29 and the National Emergency Laparotomy Audit24 on perioperative care have also been adopted where applicable. Recommendations on safe staffing have been taken from the British Society of Interventional Radiology statement.22 In areas where no guidelines exist, expert opinion has been sought. Organisation of services and principles of patient care will be audited against an amalgamation of these standards, as detailed in tables 1 and 2.

Ethics and dissemination

This audit is carried out as part of the National Comparative Audit of Blood Transfusion programme, which is supported by the National Blood Transfusion Committee in England. As this is an audit of established methods of care and it will not influence patient management while it is being conducted, it is not subject to ethical consideration by the NHS Research Ethics Committee.36 As stated in the NHS Code of Practice (2003), patient information may be collected for clinical audit without prior patient consent. No patient identifiers are collected as part of this audit.

A steering group made up of representatives from NHS Blood and Transplant, Association of Coloproctology of Great Britain and Ireland, British Society of Interventional Radiologists and the National Comparative Audit of Blood Transfusion programme will monitor progress of the study. Participating hospitals will have access to their own results via a site-specific report that will be submitted to the named contact in each participating hospital only. There will be no publication of the performance of individual hospitals.

We expect that the combined national results will be disseminated via two main publications; description of patient characteristics and outcomes, and evaluation of organisational services. These will be published on behalf of the UK Lower GI Bleeding Collaborative, which will be made up of the study leads and data enterers. The audit lead is responsible for the integrity of the data provided by their site. The steering group will act as guarantors of the publications.


Although LGIB is common, there is limited evidence on clinical presentations, use of resources and management outcomes. Many smaller studies14 ,15 have attempted to evaluate methods of investigation and treatment of LGIB but have been limited by numbers. This multinational audit in the UK is sufficiently large to capture infrequent outcomes such as complications related to infrequent investigations, interventions and report on overall mortality. It will provide a comprehensive commentary of the current management strategy of LGIB in the UK and identify areas for improvement. It will also facilitate geographical comparison of care to ensure standardisation of practice and will provide the basis for a unified approach to patient care. At the time of submission of this manuscript, data entry and data cleaning are ongoing and several queries are pending from sites. Once these are obtained, it is anticipated that the database will be locked in April 2016, after which the data will be analysed and presented according to the analysis plan. Dissemination of the audit report is expected in May 2016.


View Abstract


  • Twitter Follow Kathryn Oakland at @LGIBaudit

  • Contributors KO designed the study and wrote the manuscript. RG and NM developed audit standards related to surgery and critically reviewed the manuscript. RU developed audit standards related to radiology and critically reviewed the manuscript. FS and GC provided statistical support and critically reviewed the manuscript. JG-C designed the audit and electronic data tool and critically reviewed the manuscript. MM developed audit standards related to transfusion and critically reviewed the manuscript. VJ designed the study, developed audit standards related to gastroenterology and endoscopy and wrote the manuscript.

  • Funding This work was supported by funding from NHS Blood and Transplant, the Bowel Disease Research Foundation and the Royal College of Surgeons of England.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.