Shock index

Clinical background

The shock index serves a specific purpose: to detect circulatory compromise during the compensated phase of shock, when the heart rate and systolic blood pressure may each appear reassuring on their own. A patient who has lost up to 450 mL of blood may still have a heart rate and blood pressure within the normal reference ranges while the ratio between them is already shifted [2]. It is precisely in this window, before decompensation is obvious, that the shock index has its value.

The instrument is therefore not intended to replace established scoring systems such as SOFA or qSOFA, and it is not a laboratory result in itself. It is a rapid, cost-free measure of haemodynamic strain that can be calculated at the bedside and can flag the need for further investigation or early allocation of resources.

Calculating the shock index

The shock index is calculated as the heart rate divided by the systolic blood pressure:

$$\text{Shock index} = \frac{\text{Heart rate}}{\text{Systolic blood pressure}}$$

The variables are the two vital signs routinely measured at triage or on arrival. The index is dimensionless. A normal value lies around 0.5 to 0.7, although some data suggest that values up to about 0.9 may be accepted as normal in younger patients [2]. Values approaching 1.0 indicate impaired haemodynamics and an appreciable risk of shock.

The shock index was originally described by Allgöwer and Burri in 1967 as a measure of haemodynamic status in haemorrhage and hypovolaemic shock [1]. The concept has since been validated in a range of populations, from trauma to sepsis and obstetrics, but it is above all in trauma care that the thresholds have been calibrated against hard outcomes.

Interpretation in practice

In trauma populations, an SI >0.9 has been identified as the most widely accepted threshold for predicting the need for massive transfusion, even in patients who are relatively normotensive [2]. In a retrospective cohort of 8,111 patients with blunt trauma, the risk of activating the massive transfusion protocol was appreciably raised in patients with an SI >0.9 despite a systolic blood pressure >90 mmHg [2].

In sepsis, a lower threshold has been discussed. In a retrospective cohort of 2,524 adult patients, those with an SI >0.7 were three times as likely to have hyperlactataemia (lactate ≥4 mmol/L) as those with an SI <0.7, and the negative predictive value of a normal SI was 95% [2]. This means that in clinical work a normal shock index can be used to support setting aside appreciable shock for the time being, whereas a raised value prompts further assessment but does not on its own establish a diagnosis.

Validation and performance

The largest synthesis of the shock index in the emergency department setting is a review by Koch et al. from 2019, which examined the literature from trauma surgery, sepsis, myocardial infarction, pulmonary embolism and obstetrics [2]. In trauma populations the shock index has shown moderate to good discrimination. A retrospective study of 16,077 patients with blunt trauma found an AUC of 0.79 (95% CI 0.73–0.85) for the shock index in predicting 48-hour mortality in patients aged ≥55 years, and the age-adjusted shock index performed even better with an AUC of 0.83 (95% CI 0.78–0.88) [2]. Both outperformed the heart rate and systolic blood pressure individually.

In a broader emergency department population of 58,336 patients, those with an SI >1.2 were almost twelve times as likely to require admission as those with a normal SI (0.5–0.7) [2].

In sepsis, however, the shock index is less well studied and performs more variably. A comparative study by Akpinar et al. of 203 patients with sepsis or septic shock in an emergency department found that the shock index had an AUC of 0.67 (p <0.001) for predicting ward admission, but it was not statistically significant for intensive care admission or for one-month mortality [3]. Lactate (AUC 0.74) and SOFA (AUC 0.68) were superior for predicting mortality [3]. The shock index therefore added moderate information for triage decisions in sepsis but lacked sufficient discrimination for more serious outcomes.

In a subgroup of 295 patients with severe sepsis, 38.6% of those with a persistent SI >0.8 during at least 80% of emergency department measurements required a vasopressor within 72 hours, compared with 11.6% of those without persistent elevation [2]. This suggests that serial measurements may add value over a single measurement.

Limitations

The shock index applies to adult patients. For children there is a paediatric age-adjusted variant (SIPA) that has proved more reliable than the adult thresholds and should be preferred [2].

Several factors can distort the index. Beta blockers and calcium channel blockers may lower the heart rate and thereby cause the shock index to underestimate appreciable hypovolaemia. Patients with chronic hypertension may have an elevated baseline systolic blood pressure that makes the ratio appear lower than it would in a normotensive patient. Brain injury may cause bradycardia or tachycardia independently of volume status, and patients with autonomic neuropathy, for example in long-standing diabetes, may have a disturbed compensatory response.

The shock index rests on a single measurement on arrival and does not capture the dynamics of the patient's condition. In septic patients, data suggest that serial measurements during the emergency department stay may be more informative [2]. Moreover, a raised shock index should not be interpreted as a diagnosis in itself, but as a prompt to add lactate, a venous blood gas and targeted clinical assessment. The positive predictive value is low, particularly in sepsis, while the negative predictive value is high (95% in one sepsis cohort) [2].

A modified variant, the MSI (heart rate divided by mean arterial pressure), has been proposed in order to incorporate the diastolic blood pressure as well. A prospective study of 9,860 trauma patients found that both a low (<0.7) and a high (>1.3) MSI were associated with higher mortality, indicating a bimodal pattern in which a low MSI may reflect brain injury or hyperperfusion rather than hypovolaemia [2]. The MSI is not built into this calculator and requires the diastolic blood pressure for its calculation.

References

1. Allgöwer M, Burri C. Schockindex. Dtsch Med Wochenschr 1967. PMID: 5299199
2. Koch E, Lovett S, Nghiem T et al. Shock index in the emergency department: utility and limitations. Open Access Emerg Med 2019. PMID: 31616192
3. Akpinar G, Bulut SP, Ozsemerci HA et al. Sepsis prognosis in the emergency department: assessment of shock indices, SOFA, qSOFA, and lactate. Int J Emerg Med 2026. PMID: 42310527