Elsevier

The Lancet

Volume 356, Issue 9238, 14 October 2000, Pages 1318-1321
The Lancet

Early Report
A risk score to predict need for treatment for uppergastrointestinal haemorrhage

https://doi.org/10.1016/S0140-6736(00)02816-6Get rights and content

Summary

Background

Current risk-stratification systems for patients with acute upper-gastrointestinal bleeding discriminate between patients at high or low risks of dying or rebleeding. We therefore developed and prospectively validated a risk score to identify a patient's need for treatment.

Methods

Our first study used data from 1748 patients admitted for upper-gastrointestinal haemorrhage. By logistic regression, we derived a risk score that predicts patients' risks of needing blood transfusion or intervention to control bleeding, rebleeding, or dying. From this score, we developed a simplified fast-track screen for use at initial presentation. In a second study, we prospectively validated this score using receiver operating characteristic (ROC) curves—a measure of the validity of a scoring system—and χ2 goodness-of-fit testing with data from 197 patients. We also validated the quicker screening tool.

Findings

We calculated risk scores from patients' admission haemoglobin, blood urea, pulse, and systolic blood pressure, as well as presentation with syncope or melaena, and evidence of hepatic disease or cardiac failure. The score discriminated well with a ROC curve area of 0.92 (95% CI 0·88–0·95). The score was well calibrated for patients needing treatment (p=0·84).

Interpretation

Our score identified patients at low or high risk of needing treatment to manage their bleeding. This score should assist the clinical management of patients presenting with upper-gastrointestinal haemorrhage, but requires external validation.

Introduction

Acute upper-gastrointestinal haemorrhage is a common reason for emergency hospital admission in the UK, with up to 172 admissions per 100 000 adults yearly.1 Patients who have a history of acute upper-gastrointestinal bleeding present with a wide range of clinical severity, ranging from insignificant bleeds to catastrophic exsanguination.2 Several systems have been designed to identify patients with high risks of adverse outcomes and to differentiate them from patients with lower risks.2, 3, 4, 5, 6, 7 The use of such scores is not unequivocally supported,8, 9 and none have been widely adopted. These measures have been developed from mathematical models of patients' risks of death or rebleeding. However, since clinical treatment aims to prevent patients from dying, we believe that it is more logical to identify which patients will require clinical intervention rather than which might die. We therefore developed and validated a risk score to assess whether patients presenting with acute upper-gastrointestinal bleeding will require admission for treatment to manage their bleeding.

Section snippets

Methods

In our first study, we developed the score with data collected in an audit of admissions for acute uppergastrointestinal haemorrhage in all 19 hospitals in west Scotland.1 Data from 1748 patients were used to build a logistic regression model with the need for treatment as a response variable. Patients were defined as needing treatment if they had had a blood transfusion or any operative or endoscopic intervention to control their haemorrhage, or if they had undergone no intervention but had

Results

Table 1 shows the risk markers from the score development study of 1748 patients who were included in the final logistic regression model, alongside their associated score component values. Patients' risk scores were calculated by adding the score components associated with each clinical risk marker at the initial presentation. Table 2 shows the score values for the development and validation groups, and the numbers of patients needing treatment. The stepwise logistic regression confirmed that,

Discussion

Our score discriminated well between patients who needed clinical intervention to control upper-gastrointestinal haemorrhage and those who did not. We modelled the score on the process rather than on the outcome of clinical treatment, thus it identifies which patients may need clinical treatment, rather than identifying those at risk of death. Since the aim of treatment for acute upper-gastrointestinal haemorrhage is to prevent death, calculation of patients' risk of death is analogous to

References (28)

  • W Pimpl et al.

    Estimation of the mortality rate of patients with severe gastroduodenal hemorrhage with the aid of a new scoring system

    Endoscopy

    (1987)
  • DR Bordley et al.

    Early clinical signs identify low-risk patients with acute upper gastrointestinal hemorrhage

    JAMA

    (1985)
  • B Katschinski et al.

    Prognostic factors in upper gastrointestinal bleeding

    Dig Dis Sci

    (1994)
  • EM Vreeburg et al.

    Validation of the Rockall risk scoring system in upper gastrointestinal bleeding

    Gut

    (1998)
  • Cited by (846)

    View all citing articles on Scopus
    View full text