Clinical background
Deciding when to start vasopressor treatment in septic shock is not straightforward. The Surviving Sepsis Campaign recommends noradrenaline as the first-line vasopressor and includes it in the one-hour bundle, but its timing in relation to fluid therapy is contested [2]. The mean arterial pressure (MAP), used in most definitions of septic shock, does not selectively reflect vasodilation, since it is influenced by vascular tone, cardiac output and heart rate alike. The diastolic arterial pressure (DAP) reflects vascular tone to a greater degree and is not affected by systolic stroke volume in the same way as the systolic pressure. Since a lower diastolic component combined with tachycardia indicates both progressive vasodilation and compensatory sympathetic activation, the ratio of heart rate to DAP can serve as a measure of the severity of vasodilatory circulatory compromise. The instrument is intended above all to identify early those patients whose vasodilation is so pronounced that they will very probably need a vasopressor, rather than to replace established prognostic scores.
Calculating the diastolic shock index
The DSI is calculated as the ratio of heart rate to diastolic blood pressure:
A higher value indicates more severe vasodilatory circulatory compromise. The value depends on two physiological variables: the heart rate (compensatory tachycardia in response to a falling pressure and hypovolaemia) and the DAP (reflecting vascular tone during diastole). An isolated fall in DAP or isolated tachycardia has no clear prognostic significance on its own, but the combination in the form of the DSI predicts mortality [1].
The derivation study included a total of 761 patients with septic shock in two cohorts [1]. The first, preliminary cohort consisted of 337 patients in an intensive care unit at a university hospital in Cali, Colombia (January 2015 to February 2017). The second cohort consisted of 424 patients from the randomised controlled trial ANDROMEDA-SHOCK (March 2017 to April 2018), which enrolled patients at 28 hospitals in five Latin American countries. The DSI was calculated in the preliminary cohort immediately before vasopressor treatment was started, and in ANDROMEDA-SHOCK at randomisation, within four hours of the diagnosis of septic shock. Septic shock was defined by the Sepsis-3 criteria in ANDROMEDA-SHOCK, while the preliminary cohort used earlier criteria. Patients with atrial fibrillation, ventricular arrhythmias, a pacemaker, advanced cirrhosis (Child-Pugh C) or liver-related coagulopathy were excluded. Mortality at day 28 was 38.3 and 39.2 per cent and at day 90 was 43.0 and 43.9 per cent in the two cohorts.
Interpretation in practice
The DSI is a continuous measure without formal risk bands in its original derivation, but a threshold of around 2.0 has been externally validated as the optimal cut-off for identifying patients who will need a vasopressor [2]. In the Korean validation study, a DSI ≥ 2.0 was used together with a lactate ≥ 2.5 mmol/L as a two-point scale. In patients with both risk factors, early initiation of a vasopressor (within one hour) was associated with lower 28-day mortality (adjusted odds ratio 0.37, 95 per cent CI 0.14 to 0.94) [2].
| DSI value | Interpretation | Concrete action |
|---|---|---|
| < 2.0 | Lower risk of needing a vasopressor | Continued monitoring, adequate fluid therapy according to the SSC bundle |
| ≥ 2.0 | High risk of needing a vasopressor and of increased mortality | Consider early initiation of noradrenaline, and complement the assessment with lactate |
The DSI should be calculated at the time of the first hypotensive episode and, if necessary, again after an initial fluid bolus. In the Korean study, discrimination for the DSI improved somewhat after an initial fluid bolus (AUC 0.751) compared with at the time of hypotension (AUC 0.688) [2]. This suggests that the index can be used both before and after initial fluid therapy, and that a rise after fluid therapy should particularly prompt early vasopressor treatment.
An important point is that repeated DSI measurements during the early phase have dynamic value. In the derivation study, the time course of the DSI and the product of the DSI and the noradrenaline dose were significantly higher in non-survivors in both cohorts [1]. A DSI that does not fall during the first hours of resuscitation indicates persisting vasodilation and a poor prognosis.
Validation and performance
In the derivation study, the AUC for the pre-VP DSI (before vasopressor) in the preliminary cohort and the VP DSI (at the start of vasopressor treatment) in ANDROMEDA-SHOCK was equivalent to that of the SOFA score and the initial lactate for predicting mortality at day 28 and day 90 [1]. The MAP and the systolic shock index (heart rate/systolic blood pressure) performed significantly less well. When DAP and heart rate were analysed separately, the risk of death increased only when the DSI rose at the same time. An isolated fall in DAP or isolated tachycardia showed no clear association with mortality [1].
The only external validation study so far was conducted at Samsung Medical Center, Seoul, and included 1,917 patients with suspected infection and hypotension in the emergency department (2018 to 2019) [2]. The study developed a combined risk model with the DSI and lactate, in which a DSI ≥ 2.0 was the optimal cut-off for predicting the need for a vasopressor (AUC 0.688 at hypotension, 0.751 after a fluid bolus in the derivation cohort; 0.676 and 0.688 respectively in the validation cohort). Combined with lactate, the model reached an AUC of 0.741 at hypotension in the derivation cohort and 0.679 in the validation cohort [2]. Calibration was acceptable on the calibration plots in both the derivation and the validation cohort.
A post-hoc analysis of the ANDROMEDA-SHOCK data (424 patients) proposed a further development of the DSI into the VNERi (vascular norepinephrine responsiveness index), DAP/(noradrenaline dose × heart rate), which showed a stronger association with mortality than the DSI and DAP alone [3]. This variable does, however, require the patient to be on noradrenaline already and addresses a later decision problem than the DSI.
Limitations
The DSI applies to patients with septic shock or suspected infection with hypotension. It has not been validated for cardiogenic, hypovolaemic or obstructive shock, where the pathophysiology differs and the diastolic pressure does not primarily reflect vasodilation. A prerequisite for the index to be meaningful is a competent aortic valve; significant aortic regurgitation can lower the DAP independently of systemic vasodilation and thereby falsely raise the DSI.
Patients with atrial fibrillation, other supraventricular or ventricular arrhythmias and patients with pacemakers were excluded from the derivation study [1]. In such rhythms the heart rate becomes an unreliable variable and the DSI should be used with caution.
Most of the initial blood pressure measurements in the derivation study were obtained with an oscillometric cuff, not an intra-arterial catheter [1]. Oscillometric measurements at low pressures may underestimate the diastolic pressure and thereby exaggerate the DSI. Intra-arterial measurement is recommended when available.
The DSI has not been incorporated into formal Surviving Sepsis Campaign guidelines and has no support from systematic reviews or meta-analyses. The evidence rests on one primary observational study and one external validation study, both retrospective or post hoc. There is no randomised trial testing the DSI as a decision aid for triggering vasopressor treatment.
A common misuse is to interpret the DSI as a general prognostic index independent of context. It is a measure of vasodilation and works best in the early phase of septic shock, not as a monitoring tool during established vasopressor treatment, where other variables such as the VNERi may be more relevant [3].
References
- Ospina-Tascón GA, Teboul JL, Hernandez G et al. Diastolic shock index and clinical outcomes in patients with septic shock. Ann Intensive Care 2020;10(1):41. PMID: 32296976
- Kim DS, Park JE, Hwang SY et al. Prediction of vasopressor requirement among hypotensive patients with suspected infection: usefulness of diastolic shock index and lactate. Clin Exp Emerg Med 2022;9(3):176–186. PMID: 36164800
- Goury A, Djerada Z, Hernandez G et al. Ability of diastolic arterial pressure to better characterize the severity of septic shock when adjusted for heart rate and norepinephrine dose. Ann Intensive Care 2025;15(1):43. PMID: 40133652
