Observatoire Regional Breton sur l'Infarctus (ORBI) risk score

Clinical background

Cardiogenic shock develops in 2 to 10 per cent of patients with acute coronary syndrome and remains the leading cause of in-hospital death after STEMI, with a mortality that despite modern reperfusion therapy has plateaued at around 50 per cent [2]. Early identification of patients at risk is decisive: the ability to plan the level of monitoring proactively, to transfer the patient to coronary intensive care and to discuss mechanical circulatory support early all presuppose that the risk is known before the patient decompensates.

The challenge is that many patients who go on to develop shock are not in shock on arrival. They may be haemodynamically stable prehospitally and on arrival at the catheterisation laboratory, but deteriorate in association with reperfusion injury, progressive left ventricular dysfunction or mechanical complications over the following hours. Conventional risk markers such as Killip class and blood pressure on arrival capture only part of this, and lack the angiographic and procedural variables that strongly predict the development of shock. The ORBI risk score was developed to fill this gap and to integrate prehospital, clinical, biochemical and angiographic data into a single risk assessment [1].

Calculating the ORBI risk score

The ORBI risk score is the sum of eleven variables, ten of which are binary and one (Killip class) ordinal with three levels:

$$\text{ORBI} = 2,\mathbb{1}$${\text{age}>70} + 2,\mathbb{1}{\text{stroke/TIA}} + 3,\mathbb{1}{\text{cardiac arrest}} + 1,\mathbb{1}{\text{anterior}} + 2,\mathbb{1}{\text{delay}>90} + K + 3,\mathbb{1}{\text{HR}>90} + 4,\mathbb{1}{\text{SBP}<125 ,\wedge, \text{PP}<45} + 3,\mathbb{1}{\text{glucose}>10} + 5,\mathbb{1}{\text{LMS}} + 5,\mathbb{1}{\text{TIMI}<3}

where $K = 0$ for Killip class I, $K = 2$ for class II and $K = 6$ for class III. The score ranges from 0 to 36.

The derivation cohort consisted of 6,838 STEMI patients without cardiogenic shock on arrival, treated with primary PCI and included in the regional French ORBI registry (Observatoire Régional Breton sur l'Infarctus, Brittany) [1]. The variables were identified by stepwise multivariable logistic regression with the development of cardiogenic shock during the admission as the outcome. External validation was carried out in the RICO registry (observatoire des Infarctus de Côte-d'Or, Dijon) with 2,208 patients [1].

Four risk categories were defined from the score: low (0 to 7), low to intermediate (8 to 10), intermediate to high (11 to 12) and high (13 or more). In the derivation cohort the observed proportion with cardiogenic shock was 1.3 per cent, 6.6 per cent, 11.7 per cent and 31.8 per cent in the four categories. In the RICO validation cohort the corresponding proportions were 3.1 per cent, 10.6 per cent, 18.1 per cent and 34.1 per cent [1].

Interpretation in practice

The score should be read as support for the intensity of monitoring and for logistical planning, not as a decision tool for whether reperfusion should be performed. Several components, particularly the TIMI flow grade after PCI and the culprit vessel, are known only after the procedure is complete.

A patient in the high risk category has almost three times the risk of one in the intermediate to high category, and more than twenty times the risk of one in the low category. This warrants not only a higher level of monitoring but also proactive planning: early echocardiography to assess left and right ventricular function, securing central venous access, and discussion with a cardiac surgical centre about the possible need for mechanical circulatory support can be considered even before signs of hypoperfusion appear.

Validation and performance

In the derivation cohort a c-statistic of 0.84 was achieved, and in the RICO validation cohort 0.80, with adequate calibration in both [1].

A large multinational study including 53,537 patients with acute coronary syndrome treated with PCI evaluated the sex-specific performance of the ORBI score [2]. In the Swiss AMIS-Plus cohort the AUC was 0.78 (95 per cent CI 0.76 to 0.81) for women and 0.81 (0.79 to 0.83) for men, a statistically significant difference (P = 0.048). In the French RICO cohort the discrepancy was larger: an AUC of 0.77 (0.74 to 0.81) for women compared with 0.84 (0.81 to 0.86) for men (P = 0.002). The study developed a successor, SEX-SHOCK, which included ST depression, creatinine, CRP and the left ventricular ejection fraction and which outperformed ORBI in both sexes [2]. This indicates that the performance of ORBI is poorer in women, which is clinically relevant since women with ACS are older, have a greater burden of comorbidity and longer prehospital delays.

A prospective validation at Rigshospitalet in Copenhagen has reported good discrimination for in-hospital cardiogenic shock, but has pointed to a fundamental issue of timing: a substantial proportion of shock cases arise periprocedurally, and when the score is used to predict post-procedural shock specifically, it markedly overestimates the risk. This is because several components of the score (TIMI flow after PCI, the culprit vessel) are by definition post-procedural and therefore reflect events that have already occurred rather than future risk [3].

The concept of phase-specific risk assessment has been highlighted in more recent literature: machine learning models incorporating both pre- and post-procedural variables have shown that the relevant predictors shift between the interventional phase and the intensive care phase, and that a static score calculated at one point in time does not fully capture the dynamic evolution of risk after reperfusion [4].

Limitations

The score applies only to patients with STEMI treated with primary PCI and without cardiogenic shock on arrival. Patients already in shock on arrival were excluded from the derivation cohort and the score is not validated in this population [1]. Nor does it apply to NSTEMI patients, although some validation studies have included broader ACS populations [2].

The central problem is timing: the score cannot be calculated in full until after PCI, since the TIMI flow grade after the procedure and the culprit vessel are two of the variables (together up to 10 of the 36 points). The score therefore characterises risk rather than guiding the initial reperfusion decision. This is a deliberate design, but it means that the tool is primarily useful for post-procedural planning of monitoring and resource allocation, not for prehospital or emergency department decisions.

Killip class and blood pressure on arrival are unstable variables that can change rapidly in the early phase. A patient initially classified as Killip I may progress to Killip II or III within minutes, which changes the score. Likewise, heart rate and blood pressure can vary with pain, agitation and initial medication.

The poorer performance in women is an important limitation [2]. The derivation cohort was predominantly male, and sex-specific differences in ACS pathobiology, including the older age and different comorbidity profile of women, do not appear to be fully captured by the score's variables.

Finally, the score was derived in a French regional registry over a specific period. Reperfusion strategies, time delays and patient populations vary between countries and have evolved since the cohort was collected, which may affect transferability to contemporary populations.

References

1. Auffret V, Cottin Y, Leurent G, et al. Predicting the development of in-hospital cardiogenic shock in patients with ST-segment elevation myocardial infarction treated by primary percutaneous coronary intervention: the ORBI risk score. Eur Heart J. 2018;39(22):2090-2102. PMID: 29554243
2. Wang Y, Zeller M, Auffret V, et al. Sex-specific prediction of cardiogenic shock after acute coronary syndromes: the SEX-SHOCK score. Eur Heart J. 2024;45(43):4564-4578. PMID: 39217456
3. Holle SLD, Søholm H, Frydland M, et al. A Risk Score without a Window: The Clinical Timing Problem of Observatoire Régional Breton sur l'Infarctus Risk Score. Cardiology. 2026. PMID: 42149788
4. Stamate E, Culea-Florescu AL, Miron M, et al. Dynamic Predictive Models of Cardiogenic Shock in STEMI: Focus on Interventional and Critical Care Phases. J Clin Med. 2025;14(10):3503. PMID: 40429500