This retrospective cohort study showed that septic patients with a delayed measurement for initial lactate after ICU admission had a higher risk of 28-day mortality when the initial lactate level was over 1.0 mmol/L, meaning that measuring lactates should be as soon as possible for the benefit of reducing mortality. To our knowledge, this is the first study to report that early measurement of lactate can extend to apply for the group of patients whose initial lactate was over 1.0 mmol/L. In addition, delays in the remeasurement of lactate among septic patients whose remeasurement lactate level were > 2 mmol/L were also associated with a higher risk of 28-day mortality.
Higher Lactate had been proved an indicator of adverse outcomes among critically ill both in adult and children [23–25]. Lactate levels were recommended to be used to determine who had the highest risk of mortality for sepsis, meaning that measuring lactate was very important for clinicians and nurses to care for patients. Even the updated SSC bundle strongly suggests all the sepsis patients should be tested lactate to help assist fluid resuscitation; however, the exact time of measuring initial lactate still needs to be explored. This present finding in line with the prior study that reported early Lactate measurement was associated with lower 28-day mortality than those who had a late lactate measurement[17]. One study by Han et al., who conducted a study of 1323 sepsis patients, demonstrated that one hour delayed for measurement lactate could add 2% odds of mortality in patients whose initial lactates > 2.0 mmol/L (OR = 1.02, 95%CI:1.0003-1.05)[16]. A recent observational study by Chen et al. [17] revealed that early lactate measurement was associated with a lower risk-adjusted 28-day mortality rate in septic patients with lactate levels > 2.0 mmol/L. Although our study draws similar results, this present study has some different aspects. First, our study was a multiple center study covering over 6,334 patients; second, our study also explored the association between the time to finish initial lactate and 28-day mortality for three groups (≤ 1 mmol/L;1–2 mmol/L and > 2 mmol/L). Our study indicated that performing early lactate should be extended to those patients whose first lactate was greater than l mmol/L.
Early measuring the lactate was associated with a lower rate of mortality is that this assists clinicians perform comprehensive treatment such as antibiotic therapy, vasopressor administration, and intravenous fluid as soon as possible. On the one hand, lactate’s comprehensive measurement could help treat pathogeny and maintain patients’ vital signs. Apparently, elevated lactate was reported with a higher risk of mortality among critical illness in a series study. Therefore, early measuring lactate can help stratify the risk of critical illness and promptly implement intervention to improve the outcomes. Previous studies reported that the time to antibiotic administration in delayed lactate groups was higher than those conducting early lactate measurement (median time,3.9 versus 2.0h)[16]. This present study has several clinical implications. Previous research suggests that early lactate measurement was associated with lower mortality among septic patients with lactate levels > 2.0 mmol/L[17]. Our study extends this finding to those septic patients with lactate levels of 1–2 mmol/L, meaning early measurement lactate should also apply among these groups. Once critically patients were diagnosed with sepsis, early measurement and remeasurement lactate should be essential, stratifying the high-risk group of patients. Consequently, dynamically modify treatment based on lactate level should also be conducted, eventually improving the outcome.
The latest SSC bundle recommended lactate remeasurement to reduce the risk of mortality among sepsis patients. One study recently explored the time to finish lactate remeasurement and mortality, showing that with the time delay, the risk of mortality increases among patients with lactate levels > 2.0mmol/L[17]. Our study reported similar results. However, more studies should be explored to determine the direct information for the optimal time for remeasurement in the future.
Our study had some strengths and limitations. First, this study was a real-life cohort study consisting of a large sample size of patients from multiple centers, which help to generalize this finding around the USA. Second, we have used comprehensive statical analyses such as the adjusted logistic regression model, generalized additive model (GAM), interaction analysis, and sensitive analysis to explore this association between the time to complete initial lactate and 28-day morality help elaborate these results. However, some limitations should be cautious. First, the definition for sepsis was updated with years passing. We confirmed the sepsis diagnosis based on the record of the APACHE IV dataset. Whereas the third version of sepsis criterion was defined as patients who had documented or suspected infection together with a sharp change in total SOFA score ≥ 2 points)[26]. Hence, our results were robust after we conducted a subgroup analysis based on the different sites of infection with similar results. Second, inherent to the study design’s observational nature, which lends itself subject to limitations that should be considered, including confounding by indication. Our findings are hypothesis-generating and do not imply causality. Even though we have adjusted several confounding variables, some other significant covariates such as frailty, sarcopenia, and malnutrition still influence these results[27, 28]. In addition, some variables were miss; Nonetheless, we used contemporary methods such as dummy variables to figure out the missing data to minimize potential bias. Furthermore, our study did not provide information about interventions during the initial stabilization, which may overestimate or underestimate the relationship between early lactate measurement and survival. Finally, we also acknowledge that as our participants were patients only from the eICU-CRD database, the generalizability of our study to other populations should need to explore.
Conclusion: By using data from the eICU-CRD database, our study consisted of 6334 patients with sepsis found that regardless of initial lactate 1–2 mmol/L or > 2 mmol/L, the associations between delayed in lactate measurement (hours) and risk of 28-day mortality were statistically significant. Repeating the measurement (hours) after the initial measurement as soon as possible is beneficial for patients whose remeasurement lactate was > 2 mmol/L. These findings need confirmation in other studies.