Pre-endoscopy (clinical) Rockall score vs the full Rockall score

What the pre-endoscopy Rockall score includes

The pre-endoscopy Rockall score, also called the clinical, initial or admission Rockall score, is the part that can be scored before any endoscopy. It has three items. Age scores 0 under 60 years, 1 from 60 to 79 and 2 at 80 or over. Shock scores 0 with a systolic blood pressure of 100 mmHg or more and a pulse below 100, 1 for a pulse of 100 or more with a systolic pressure of 100 or more, and 2 for a systolic pressure below 100. Comorbidity scores 0 with no major comorbidity, 2 for cardiac failure, ischaemic heart disease or any other major comorbidity, and 3 for renal failure, liver failure or disseminated malignancy. The total runs from 0 to 7. Rockall and colleagues derived these points in Gut in 1996 from 4,185 cases of acute upper gastrointestinal haemorrhage in the UK in 1993, with death as the outcome, and reported that the score picked out 15% of all cases at presentation as low risk.

The Rockall score calculator gives the pre-endoscopy score on its own when you choose “Not yet” for endoscopy.

What endoscopy adds in the full Rockall score

The full Rockall score, also called the complete or post-endoscopy score, adds two endoscopic items to the pre-endoscopy score and runs from 0 to 11. Diagnosis scores 0 for a Mallory-Weiss tear or when no lesion and no stigmata of recent haemorrhage are seen, 1 for all other diagnoses, such as a peptic ulcer, and 2 for a malignancy of the upper gastrointestinal tract. Major stigmata of recent haemorrhage score 0 when there are none or only a dark spot, and 2 for blood in the upper gastrointestinal tract, an adherent clot, or a visible or spurting vessel. Because these two items add 0 to 4 points, the full score is always at least as high as the pre-endoscopy score for the same patient. The Lancet paper that followed the Gut study applied the full score to the 2,531 patients in the 1993 UK national audit who had endoscopy after an acute admission, and found that 744 of them (29.4%) scored 2 or less.

Why the same Rockall number means a different risk

A pre-endoscopy Rockall score and a full Rockall score with the same number carry very different risks of death, because the full score has already taken in what endoscopy found. In Rockall's 1996 paper, a pre-endoscopy score of 2 carried a mortality of 5.6%, while a full score of 2 carried 0.2%. A pre-endoscopy score of 5 carried 39.6% against 10.8% for a full score of 5, and the pre-endoscopy score reached 50.0% at 7, where the full score stood at 27.0%. The reason is that endoscopy spreads patients with the same pre-endoscopy score across the full scale: those with a Mallory-Weiss tear, or no lesion and no stigmata, keep their score, while those with other lesions, stigmata or cancer move up. So a high pre-endoscopy score is a warning, but after endoscopy the number to quote is the full score, read against the full-score figures. Reading one version against the other's table overstates or understates the risk.

Mortality as reported by Rockall and colleagues in 1996. The Rockall score calculator shows the right figure for whichever version you score, and charts both columns.

Low-risk and higher-risk cut-offs before and after endoscopy

The low-risk cut-off is a pre-endoscopy Rockall score of 0 or a full Rockall score of 2 or less, and the cut-offs for a higher risk of death are 4 or more before endoscopy and 5 or more after it. The low-risk groups come from Rockall's own papers: the Gut study described 15% of all cases at presentation and 26% after endoscopy as at low risk of rebleeding and negligible risk of death, and in the Lancet audit only 32 of 744 patients with a full score of 2 or less (4.3%) rebled and one (0.1%) died. The higher-risk thresholds come from Stanley and colleagues, who followed 3,012 consecutive patients at six hospitals in Europe, North America, Asia and Oceania (BMJ, 2017). A pre-endoscopy score of 4 or more predicted death with a sensitivity of 78.6% and a specificity of 65.0%, and a full score of 5 or more with 74.0% and 61.7%. Overall 30-day mortality in that study was 7%.

A pre-endoscopy score of 0 does not mean a patient will need no treatment. In Stanley's study, 436 patients (14.7%) scored 0 before endoscopy; 14% of them needed an intervention or died, and one of them died. In a 2015 New Zealand study, 15 of 42 patients scoring 0 (35.7%) needed transfusion or endoscopic treatment, and the authors concluded that the pre-endoscopy score should not be used to pick patients for outpatient endoscopy. For that decision, Stanley found a Glasgow-Blatchford score of 1 or less was the best threshold; score it with the Glasgow-Blatchford calculator.

Worked examples: scoring before and after endoscopy

Three patients show how the pre-endoscopy and full Rockall scores can stay level or move apart once endoscopy is done. A 45-year-old woman with no major illness vomits several times and then brings up fresh blood; her pulse is 88 and her systolic pressure 124 mmHg. Her pre-endoscopy score is 0 for age, 0 for shock and 0 for comorbidity, a total of 0. Endoscopy shows a Mallory-Weiss tear with no stigmata, so her full score stays at 0, where Rockall reported no deaths. An 83-year-old man with no major comorbidity has melaena, a pulse of 92 and a systolic pressure of 132 mmHg. Age alone gives him 2, so his pre-endoscopy score is 2, with a reported mortality of 5.6%. Endoscopy finds no lesion and no stigmata, so his full score is still 2, where mortality was 0.2%. The third patient, a 74-year-old woman with ischaemic heart disease, moves the other way.

Mortality at each score, in brackets, is from Rockall 1996. The third patient's pre-endoscopy score of 4 already meets Stanley's threshold for predicting death, and her full score of 7 carried a rebleeding rate of 43.8% in the 1996 paper. Enter each patient in the Rockall score calculator, with endoscopy set to “Not yet” and then “Yes”, to see both scores.