Impact of a Capillary Refill Time–Guided Personalised Hemodynamic Resuscitation Protocol in Septic Shock in Low- and Middle-Income Countries: A Post Hoc Analysis of the ANDROMEDA-SHOCK-2 Trial
The capillary refill time-guided resuscitation protocol showed consistent treatment effects in septic shock patients across different income settings, but the clinical significance of these findings requires careful consideration.
Where it sits
this study against the rest of the vasopressin corpusSummary and findings
The study assessed the treatment effect of capillary refill time–guided personalised hemodynamic resuscitation in patients with early septic shock across low- and middle-income countries. The primary outcome was a hierarchical composite of 28-day mortality, duration of vital organ support, and hospital length of stay. The analysis included 1,467 patients from 86 ICUs.
Abstract
<title>Abstract</title> <p> <bold>PURPOSE:</bold> To assess the treatment effect of capillary refill time–guided personalised hemodynamic resuscitation (CRT-PHR) in patients with early septic shock in low- and middle-income countries (LMICs). <bold>METHODS:</bold> Post hoc analysis of the ANDROMEDA-SHOCK-2 randomised clinical trial, conducted in 86 ICUs across 19 countries. Centres were classified according to World Bank income groups as LMICs or high-income countries (HICs). The primary outcome was a hierarchical composite of 28-day mortality, duration of vital organ support, and hospital length of stay, analysed using a stratified win ratio. <bold>RESULTS:</bold> Among 1,467 patients, 652 (44.4%) were treated in LMIC centres and 815 (55.6%) in HIC centres. Patients enrolled in LMICs were younger, had higher lactate levels, more frequent abnormal CRT, and greater baseline vasopressin use (59.4% vs 29.5%). Treatment-effect estimates were similar across income groups, with win ratios of 1.16 (95% CI 1.14–1.18) in LMICs and 1.18 (95% CI 1.17–1.20) in HICs, with no evidence of treatment-effect modification by country income level. Within LMICs, baseline lactate >4 mmol/L showed the strongest exploratory signal of heterogeneity (win ratio 1.57 vs 0.93; interaction <italic>p</italic> =0.002). Vasopressin use was associated with a numerically larger treatment-effect estimate (1.39 vs 1.12), but was not statistically significant. Centre-level analyses suggested heterogeneity by hospital and ICU size (interaction <italic>p</italic> =0.081 for both). <bold>CONCLUSION:</bold> Despite the marked differences in baseline severity, vasopressor use, and resuscitation practice, CRT-PHR produced consistent and significant treatment effects in both LMIC and HIC settings. Within LMICs, hyperlactatemia a centre-level heterogeneity identified contexts of differential benefit, underscoring limitations of national income classification as a proxy for local critical-care capacity. </p>
Background
This paper addresses the effectiveness of capillary refill time–guided personalised hemodynamic resuscitation (CRT-PHR) in patients experiencing early septic shock, particularly in low- and middle-income countries (LMICs). Prior research has indicated that resuscitation protocols can vary significantly based on local healthcare resources and practices. Understanding the impact of CRT-PHR in diverse settings is crucial for improving outcomes in septic shock management.
Methods
This study is a post hoc analysis of the ANDROMEDA-SHOCK-2 randomized clinical trial, which included 1,467 patients across 86 ICUs in 19 countries. The primary outcome was a hierarchical composite of 28-day mortality, duration of vital organ support, and hospital length of stay, analyzed using a stratified win ratio. Patients were classified based on whether they were treated in LMICs or high-income countries (HICs).
Results
The primary endpoint showed a win ratio of 1.16 (95% CI 1.14–1.18) for patients in LMICs and 1.18 (95% CI 1.17–1.20) for those in HICs, indicating a consistent treatment effect across income levels. Notably, patients in LMICs had higher baseline lactate levels and greater vasopressin use (59.4% vs 29.5%). The interaction p-value for baseline lactate >4 mmol/L was 0.002, indicating significant heterogeneity in treatment effect.
Interpretation
The findings suggest that CRT-PHR can produce significant treatment effects in both LMIC and HIC settings, although the clinical significance of the win ratios may vary. The study's results align with prior literature indicating variability in treatment outcomes based on local healthcare capacity. However, the lack of statistical significance in the vasopressin treatment-effect estimate and the post hoc nature of the analysis introduce confounding factors that limit the conclusions.
Key findings
- 1,467 patients enrolled, n=1,467.
- 44.4% treated in LMIC centres, n=652.
- Win ratio of 1.16 (95% CI 1.14–1.18) in LMICs.
- Win ratio of 1.18 (95% CI 1.17–1.20) in HICs.
- Baseline lactate >4 mmol/L showed win ratio of 1.57 vs 0.93, interaction p=0.002.
- Vasopressin use associated with treatment-effect estimate of 1.39 vs 1.12, not statistically significant.
Limitations
- Post hoc analysis may introduce confounding factors.
- Findings may not be generalizable due to varying critical care capacities.
- Statistical significance not achieved for vasopressin treatment effect.
- Single-site and multi-country variability not fully accounted for.