Skip to main content
Download PDF

Prognostic value of capillary refill time in adult patients: a systematic review with meta-analysis

Critical Care volume 27, Article number: 473 (2023) Cite this article

Abstract

Purpose

Acute circulatory failure leads to tissue hypoperfusion. Capillary refill time (CRT) has been widely studied, but its predictive value remains debated. We conducted a meta-analysis to assess the ability of CRT to predict death or adverse events in a context at risk or confirmed acute circulatory failure in adults.

Method

MEDLINE, EMBASE, and Google scholar databases were screened for relevant studies. The pooled area under the ROC curve (AUC ROC), sensitivity, specificity, threshold, and diagnostic odds ratio using a random-effects model were determined. The primary analysis was the ability of abnormal CRT to predict death in patients with acute circulatory failure. Secondary analysis included the ability of CRT to predict death or adverse events in patients at risk or with confirmed acute circulatory failure, the comparison with lactate, and the identification of explanatory factors associated with better accuracy.

Results

A total of 60,656 patients in 23 studies were included. Concerning the primary analysis, the pooled AUC ROC of 13 studies was 0.66 (95%CI [0.59; 0.76]), and pooled sensitivity was 54% (95%CI [43; 64]). The pooled specificity was 72% (95%CI [55; 84]). The pooled diagnostic odds ratio was 3.4 (95%CI [1.4; 8.3]). Concerning the secondary analysis, the pooled AUC ROC of 23 studies was 0.69 (95%CI [0.65; 0.74]). The prognostic value of CRT compared to lactate was not significantly different. High-quality CRT was associated with a greater accuracy.

Conclusion

CRT poorly predicted death and adverse events in patients at risk or established acute circulatory failure. Its accuracy is greater when high-quality CRT measurement is performed.

Introduction

Acute circulatory failure results in tissue hypoperfusion that leads to life-threatening organ dysfunction. The prognostic value of tissue hypoperfusion has generated substantial research interest. Notably, the evaluation of peripheral perfusion through capillary refill time (CRT) has gained considerable attention during the last decade. CRT measures the amount of time necessary for the skin to return to baseline color after the application of a firm pressure (Additional file 1). CRT can be measured easily at the bedside within a few seconds, and there are more rapid changes after resuscitation when compared to lactate clearance [1]; variations of CRT after a passive leg raising [2] or a fluid challenge [3] can be detected within a few seconds. Furthermore, CRT measurement is an easy-to-use, costless method that allows tissue perfusion assessment at admission as well as during ICU stay. Since its first description [4], CRT became popular in the 1980s when Champion et al. included CRT in the Trauma Score [5]. Since then, CRT has been found to be able to assess severity [6,7,8,9,10] or to guide treatments [11] in different settings. In addition, a recent randomized trial suggested that a resuscitation strategy targeting CRT normalization may reduce morbidity and mortality in septic shock patients when compared to a strategy based on lactate clearance [12, 13]. CRT was then recommended as a potential therapeutic target by international experts for critically ill patients [14]. However, the relationship between CRT and outcome are still unclear as studies have reported conflicting results [15,16,17]. The only published meta-analysis was conducted in pediatric patients [18], and it is of note that pediatric intensivists seem more convinced of the prognostic accuracy of CRT than those treating adult patients [17, 19].

We therefore conducted a systematic review of studies evaluating CRT as a prognostic factor in adult patients and performed a meta-analysis to assess the ability of CRT to predict death or adverse events in a context of acute circulatory failure or in a patients at risk of acute circulatory failure.

Methods

We conducted the study according to the Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy [20] and existing guidelines for reviews of diagnostic accuracy studies [21]. The study was reported in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA 2020) statement [22] (Additional file 3: Table S1). This systematic review was prospectively registered on PROSPERO (CRD42022297158, submitted 02/27/2022) prior to initiating data extraction.

Eligibility criteria

Clinical trials eligible for this meta-analysis were those that studied the prognostic value of CRT in a context of established acute circulatory failure or in a patient at risk of acute circulatory failure. We defined acute circulatory failure as the need for vasopressors or inotropes in combination with signs of hypoperfusion. If the inclusion criteria of the studied population aligned with this definition, the study was included and categorized as a population of patients with acute circulatory failure. We defined patients at risk of acute circulatory failure as those for whom the CRT was used as a triage method without restriction to patients in acute circulatory failure (e.g., first evaluation at the emergency department, rapid response team first evaluation, patients with trauma, etc.). The primary analysis outcome was death with no specific time frame after CRT measurement. Adverse events were defined as any unfavorable event explicitly labeled as such in the analyzed reports. This included outcomes such as admission to ICU, extended length of stay, and severe complications according to the Clavien-Dindo scale. Additionally, we accepted composite outcomes that included death as part of the definition of an adverse event. We excluded studies that concerned animals, studies not published in English language, letters, and reviews, studies that did not study the relationship between CRT and prognosis, studies in which CRT was performed with a device, and studies assessing localized perfusion such as free flap, or ischemic limb.

Search strategy

Eligible studies were identified by searching the MEDLINE, EMBASE, and Google Scholar databases from inception to February 2022 with no language restriction and using the following keywords: “Capillary refill time” or “Capillary refill.” We also screened articles in the reference section of review articles and conducted a snowballing procedure to examine the references in those review retrieved through the systematic search. No language restriction was applied to the searches.

Study selection

Two authors (MJL and AP) independently reviewed and screened the title and abstract of potentially relevant studies and determined final eligibility through examination of full texts. Disagreements that could not be resolved among the two authors through discussion were addressed by a third author (JLF). We included studies that provided information about CRT and the outcome in adults, irrespective of clinical situation. The studies had to provide the number of patients with normal and abnormal CRT and the number of patients with positive or negative outcomes (or the sensitivity, specificity, and prevalence) in each situation to calculate the number of true positives, true negatives, false positives, and false negatives. If such data were unavailable but were reported to have been collected by the authors, we emailed the corresponding author to obtain the data. A second email including co-authors was sent one month later, and in case of no reply after a month the study was excluded.

Data extraction

We used a standardized form to extract (independently collected by MJL and AP) the following variables from the selected studies: the year of publication, name of journal, the methods used to perform the CRT (site of measurement on the skin, duration of compression, mode of compression, use of a stopwatch), results, the nature of the patient’s state of shock if the patient was in shock, but also the sample size, the number of true positives, true negatives, false positives and false negatives, as well as the area under the receiver operating characteristic curve (AUC ROC). When several CRTs were performed in the first days of resuscitation, we retained the one on the first day and, if the information was given, we selected the one performed after initial resuscitation. High-quality CRT measurement was defined as those corresponding to the mean of 2 or more CRT values made using a standardized compression and a stopwatch.

Quality assessment

Two authors (MJL and AP), using the QUADAS-2 tool for assessing risk of bias in diagnostic accuracy studies, independently determined the quality of the included studies through examination of the full text. QUADAS-2 tool [23] encompasses four domains: patient selection, index test, reference standard, and flow and timing. We used the signaling questions to judge risk of bias and applicability concern. We constructed the flow diagram for the primary study, and judged bias and applicability. The risk of bias and applicability was assessed as high, low, or unclear.

Statistical analysis

We estimated the pooled AUC ROC, sensitivity, and specificity for CRT as a predictor of death or adverse events in patients with acute circulatory failure and at risk of acute circulatory failure. As we anticipated a great between-study heterogeneity, a random-effects model was used to pool effect sizes. The Mantel–Haenszel estimator was used to calculate Q and τ2. We used Knapp–Hartung adjustments [24] to calculate the confidence interval (CI) around the pooled effect. The effect size was the diagnostic odds ratio (DOR). We did not use continuity correction except to calculate individual study results in which we used a continuity correction of 0.5 in studies with zero cell frequencies. A forest plot was built to summarize the effect size of each study and pooled results.

The primary analysis was the ability of abnormal CRT to predict death in patients with acute circulatory failure. Secondary analyses included the ability of CRT to predict death or adverse events in patients at risk or with confirmed acute circulatory failure. Secondary analysis included the accuracy (AUC ROC, sensitivity, specificity, DOR) of abnormal CRT to predict death or adverse event, or to predict acute kidney injury in patients with, or at risk of, acute circulatory failure. We also estimated the accuracy of lactate as a predictor of death or adverse events in patients with acute circulatory failure and at risk of acute circulatory failure when data were available, and compared its accuracy to that of CRT. We also sought to identify explanatory factors associated with better accuracy.

We used the Spearman correlation coefficient between sensitivity and false positive rate to detect a threshold effect. We conducted several sensitivity analyses in predefined subgroups of patients, analyzed subgroup differences using the Q test, and P values of the tests were provided. We compared studies conducted in an ICU setting to those in a non-ICU setting; studies with patients in septic shock to those without patients in septic shock; studies in which patients were in acute circulatory failure to those in which patients were not in acute circulatory failure; studies in which the location of CRT was a finger to those in which this was performed at another location; studies in which CRT was performed using a method to apply pressure on the skin in a reproductive manner to those in which this was not the case; and studies describing the use of a stopwatch to measure CRT to those which did not. We also added four subgroup post hoc analyses: We compared studies with high-quality CRT measurement to those with low-quality CRT measurement; studies predicting death to those predicting adverse events; studies with a low risk of bias to those with a high risk of bias; and studies with a CRT threshold at 3 s to those using other thresholds. To evaluate the risk of bias, we used the quality assessment of diagnostic accuracy studies QUADAS-2 scale [23]. We built a scoring system, where, for each of the four domains, zero points were given for low risk, two points for high risk, and one point for unclear risk for each item of the QUADAS evaluation, and we then summed the sub-scores to calculate the QUADAS score; studies at low risk of bias were those with a score below or equal to the median of all the scores, and studies at high risk of bias studies were those with a scores strictly greater than the median of all the scores.

In a sensitivity analysis, we also investigated the causes of heterogeneity using outlier detection. We defined outliers as studies that showed an effect size that was out of the 95%CI of the effect size of the complete analysis. After excluding outlier studies, we calculated the pooled effect size on the remaining studies. Lastly, we also plotted the overall effect and I2 heterogeneity of all meta-analyses that were conducted using the leave-one-out method [25]. We performed a meta-regression based on a mixed effect model, including the same criteria as for the subgroup analyses if the P value was less than 0.5 to respect the concept of parsimony. We performed a prediction test to assess the robustness of the effect size. Results were expressed as mean (95%CI) or as mean ± standard deviation (SD). We used R version 4.0.4 (R Core Team 2017, Vienna, Austria) to perform statistical analyses. The meta [26] and meta4diag [27] packages were used. All tests were two-sided, and a p value less than 0.05 was considered significant.

Results

Characteristics of included studies

A total of 23 studies were included (Fig. 1), corresponding to 60,656 patients. These studies were published between 1994 and 2022; most of them (12/23) between 2019 and 2022. Investigations were performed in the emergency department (n = 8, 35%), in the ICU (n = 10, 43%), the operating room (n = 1, 4%), and the prehospital setting (n = 4, 17%). In 13 studies (57%), only patients with acute circulatory failure were included, and in 11 studies (48%) only those with septic shock were included. The characteristics of included studies are presented in Table 1. The mean ± SD abnormal CRT threshold value was 3.3 ± 0.8 s. The site of CRT measurement was the fingertip in 18 studies (78%), the chest in 2 studies (9%), and the knee in 3 studies (13%). A stopwatch was used in 12 studies (52%). High-quality CRT measurement was performed in 5 studies (22%). In 7 studies (30%), CRT was assessed before initial resuscitation. The mean ± SD frequency of the studied outcome (death or adverse event) was 26 ± 14%; that of death was 23 ± 14%. A summary of the sensitivity and specificity of CRT in individual studies is provided in Additional file 2: Figure S1.

Fig. 1
figure 1

Flowchart of the meta-analysis selection process. CRT Capillary refill time

Full size image
Table 1 Characteristics of included studies
Full size table

Risk of bias and applicability concerns

The overall risk of bias was high in 14/23 studies. (Individual study evaluations of the risk of bias are presented in Fig. 2, and pooled results in Fig. 3.)

Fig. 2
figure 2

Light plot QUADAS evaluation of risk of bias of each study

Full size image
Fig. 3
figure 3

QUADAS assessment of risk of bias and applicability concern. Proportions of studies with low, unclear, and high risk of bias (A) or applicability concern (B) according to each item of the QUADAS evaluation

Full size image

Primary analysis

Thirteen studies selected patients in acute circulatory failure and considered death as the outcome. In these studies, CRT was predictive of death; pooled AUC was 0.663 (95%CI [0.591; 0.756]). The pooled sensitivity was 54% (95%CI [43; 64]), and the pooled specificity was 72% (95%CI [55; 84]). The pooled DOR was 3.4 (95%CI [1.4; 8.3], P = 0.013; Table 2).

Table 2 Primary and secondary analyses
Full size table

Secondary analysis

In patients with acute circulatory failure or at risk of acute circulatory failure, CRT was also predictive of death or adverse events; the AUC was 0.69 (95%CI [0.65; 0.74]). The pooled sensitivity was 48% (95%CI [36; 61]), and the pooled specificity was 81% (95%CI [67; 90]). The pooled DOR was 4.3 (95%CI [2.6; 7.3], P < 0.001; Table 2, Fig. 4 and Fig S1.)

Fig. 4
figure 4

Diagnostic odds ratio of individual study and pooled odds ratio using a random effect model

Full size image

In patients with acute circulatory failure and at risk of acute circulatory failure (n = 11 studies), the arterial lactate level was not an accurate predictor of death; the AUC was 0.539 (95%CI [0.529; 0.549]). The pooled sensitivity was 46% (95%CI [18; 77]), and the pooled specificity was 76% (95%CI [52; 90]). The pooled DOR of an abnormal lactate to predict death or adverse events was 2.6 (95%CI [1.3; 5.2]; Table 2). After retrieving the information in reports and emailing the authors, we were able to compare CRT and lactate in 9 studies. Among the predictive ability of lactate and CRT in the 9 studies where both lactate and CRT were available, there was no significant difference between CRT and lactate to predict death (P = 0.687; Table 2).

The planned secondary analysis on acute kidney injury was not performed as only one studied reported this outcome but was among the 11 studies excluded due to the lack of data to assess the effect size.

Heterogeneity and the causes of heterogeneity

I2 and prediction interval

The between-study heterogeneity I2 value was 96% (95%CI [95; 97]; details for heterogeneity in primary, secondary analyses, subgroup analyses, and sensitivity analyses are presented in Tables 2, 3, and 4, as well as in Additional file 2: Figure S2. The prediction interval ranged from OR = 0.5 to 34.6; as this includes 1, it indicates that due to varying effects, we cannot rule out that future studies may not confirm the diagnostic ability of CRT (Additional file 2: Figure S3). The correlation between sensitivities and false positive rates suggested a threshold effect (Spearman’s correlation coefficient: 0.68, 95%CI [0.37; 0.85]).

Table 3 Influence case removed analysis
Full size table
Table 4 Subgroups analyses
Full size table

Sensitivity analyses

Subgroup analyses confirmed the significance of the effect size in all subgroups (Table 4). We then tested the effect of removing outliers from the analysis. The studies reported by Hernandez et al. [28], Coslovsky et al. [29], Darioli et al. [30], Jouffroy et al. [16], and Morocho et al. [31] (Fig. S2) showed an effect size that was out of the 95%CI of the effect size of the complete analysis. These studies were therefore considered as outliers and excluded. The analysis performed in the 18 remaining studies found a pooled AUC ROC of 0.67 (95%CI [0.57; 0.82]). The pooled sensitivity was 46% (95%CI [18; 77]), and the pooled specificity was 75% (95%CI [52; 90]). The pooled DOR was 3.1 (95%CI [2.2; 4.2], P < 0.0001), and the prediction interval OR = 1.7 to 5.5, τ2 = 0.0647 and I2 = 55% (95%CI [23; 73]) (Fig. S3). We also performed an influence analysis (Additional file 2: Figure S2) using the leave-one-out method, and no study was found to modify the meta-analysis. Finally, we performed a meta-regression; the variables with a P value less than 0.5 and hence included in the model were the following: septic shock as an inclusion criterion in the study, quality of CRT measurement, number of measurements contributing to the mean CRT value, and compression method. The model was not significant (P = 0.181), and the test for residual heterogeneity was significant (P < 0.0001). None of the covariates included in the meta-regression were found to be a significant source of heterogeneity.

Discussion

This meta-analysis showed an overall low predictive value of CRT on mortality or adverse events in adults, both in established acute circulatory failure and in patients at risk of it. Furthermore, CRT was found to be a useful parameter for assessing the patient severity in various settings.

The pooled AUC ROC curve indicated that CRT was poorly accurate, but a significant effect size was found in all the studied situations and sensitivity analysis confirmed the predictive ability of CRT in these situations. This is of little surprise, as the link between mortality and hypoperfusion is not straightforward and many competing factors could influence mortality as an outcome [32], and is supported by the AUC ROC of lactate levels to predict death that was close to that of CRT. It is also of note that there was no significant difference between the ability of CRT and lactate to predict adverse events or death, which is consistent with the equivalence or superiority of CRT as a target for therapeutic intervention [12, 13]. In this context, and owing to stress-related hyperlactatemia, as well as the numerous pitfalls in the interpretation of lactate and lactate clearance, the clinical relevance of using lactate as a potential target in shock seems to be questionable [33]. Another point is that mortality was used as the outcome criterion (reference standard) of the primary analysis. This can be considered as methodological strength as this reduces the risk of bias, but studies aiming to explore the association between perfusion variables and organ dysfunction may be more relevant than mortality [34]. Herein, we planned to study renal function yet only one report was identified; although not included in the review it was found that prolonged CRT on the sternum in 1003 patients admitted to ICU was associated with acute kidney injury [35]. This suggests that further studies could be of interest, allowing a quantitative approach to be used; for example, assessing the correlation between CRT and serum creatinine could explore a dose–response relationship, providing further evidence between skin hypoperfusion and organ hypoperfusion [36].

A limitation of the evidence included in this review is that the included studies had very heterogeneous effect size, characteristics, and designs. However, both the subgroup analyses and the meta-regression argued against the influence of heterogeneity on the results. As the prediction interval of the odds ratio included one, the inclusion of future studies in this meta-analysis may not confirm the diagnostic ability of CRT. Removing outliers led to a decrease in heterogeneity without affecting the pooled effect size. Still, this heterogeneity in effect size can be explained by the heterogeneity of the setting and CRT measurement method used in each study. Some studies took place in ICU [6], others in ED [16], and others in prehospital settings [37]; in addition, some contexts were not widely studied such as heart failure [38] or postoperative settings [8]. The method applied to assess CRT differed markedly regarding stopwatch usage, duration and amount of compression, site of measurement, threshold; in addition, many did not report this in detail and it is likely that practice varied within these studies, reflecting that reported in real-life clinical practice [17]. This is of importance as a lack of standardization increases the risk of measurement bias [39,40,41]. Limitations of the review process include the exclusion of studies not reporting sufficient data to calculate the effect size and for which the contacted authors did not provide the lacking data; nevertheless, the number of patients included in these 11 studies represented 4519 patients (data not shown) that would have represented only 7% of the total number of patients if these had been included. The choice to exclude studies reported only by abstracts, studies not published in English language, as well as unpublished studies may have also increased the risk of reporting bias, but this risk bias was reduced by prospective PROSPERO registration with a pre-specified primary and secondary analysis. Other limitations of the review process include the absence of best CRT threshold calculation as an insufficient number of thresholds for each published study were given. Also, a threshold effect was detected, reflecting heterogeneity in thresholds; a ROC curve analysis was performed because these provide an overall summary of prognostic test’s accuracy, independent of this effect [42].

Implications of the review for practice are the following. First, as high-quality CRT increased by more than fourfold, the DOR to predict mortality further efforts to standardize the measurement technique in clinical practice is warranted. This may also be the key to explain the discrepancy on reproducibility on previous studies. Second, the meta-analysis supports a statistically significant link between abnormal CRT and a poor outcome. As CRT is recognized for its ability to reflect skin blood flow [43], and considering that isolated cutaneous hypoperfusion, as seen during mild cold exposure, generally does not result in systemic consequences such as death or adverse events, the notable association between the outcome and CRT suggests that prolonged CRT may signal compromised tissue perfusion. Consequently, CRT can be considered as a warning signal of tissue hypoperfusion in patient at risk or confirmed acute circulatory failure in clinical practice.

In conclusion, this meta-analysis showed that overall the CRT poorly predicted death or adverse events in patients at risk or established acute circulatory failure. As any single variable approach the prognostic value remains low but is comparable to lactate levels. Its accuracy is greater when high-quality CRT measurement is performed, and thus, efforts should be focused on standardizing the technique in clinical practice.

Availability of data and materials

The data that support the findings of this study are available from the corresponding author, [MJL], upon reasonable request.

References

  1. Hernandez G, Pedreros C, Veas E, Bruhn A, Romero C, Rovegno M, et al. Evolution of peripheral vs metabolic perfusion parameters during septic shock resuscitation. A Clinical-Physiologic Study J Crit Care. 2012;27:283–8.

    PubMed  Google Scholar 

  2. Jacquet-Lagrèze M, Bouhamri N, Portran P, Schweizer R, Baudin F, Lilot M, et al. Capillary refill time variation induced by passive leg raising predicts capillary refill time response to volume expansion. Crit Care. 2019;23:281.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Raia L, Gabarre P, Bonny V, Urbina T, Missri L, Boelle P-Y, et al. Kinetics of capillary refill time after fluid challenge. Ann Intensive Care. 2022;12:74.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Guedel AE. Cyclopropane anesthesia. Anesthesiology. 1940;1:13–25.

    Article  Google Scholar 

  5. Champion HR, Sacco WJ, Hannan DS, Lepper RL, Atzinger ES, Copes WS, et al. Assessment of injury severity: the triage index. Crit Care Med. 1980;8:201–8.

    Article  CAS  PubMed  Google Scholar 

  6. Ait-Oufella H, Bige N, Boelle PY, Pichereau C, Alves M, Bertinchamp R, et al. Capillary refill time exploration during septic shock. Intensive Care Med. 2014;40:958–64.

    Article  CAS  PubMed  Google Scholar 

  7. Hernandez G, Bruhn A, Castro R, Pedreros C, Rovegno M, Kattan E, et al. Persistent sepsis-induced hypotension without hyperlactatemia: a distinct clinical and physiological profile within the spectrum of septic shock. Crit Care Res Pract. 2012;2012:536852.

    PubMed  PubMed Central  Google Scholar 

  8. van Genderen ME, Paauwe J, de Jonge J, van der Valk RJP, Lima A, Bakker J, et al. Clinical assessment of peripheral perfusion to predict postoperative complications after major abdominal surgery early: a prospective observational study in adults. Crit Care. 2014;18:R114.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Hernandez G, Bruhn A, Castro R, Regueira T. The holistic view on perfusion monitoring in septic shock. Curr Opin Crit Care. 2012;18:280–6.

    Article  PubMed  Google Scholar 

  10. van Genderen ME, Lima A, Akkerhuis M, Bakker J, van Bommel J. Persistent peripheral and microcirculatory perfusion alterations after out-of-hospital cardiac arrest are associated with poor survival. Crit Care Med. 2012;40:2287–94.

    Article  PubMed  Google Scholar 

  11. Ruste M, Sghaier R, Chesnel D, Didier L, Fellahi J-L, Jacquet-Lagrèze M. Perfusion-based deresuscitation during continuous renal replacement therapy: a before-after pilot study (The early dry Cohort). J Crit Care. 2022;72:154169.

    Article  CAS  PubMed  Google Scholar 

  12. Zampieri FG, Damiani LP, Bakker J, Ospina-Tascón GA, Castro R, Cavalcanti AB, et al. Effects of a resuscitation strategy targeting peripheral perfusion status versus serum lactate levels among patients with septic shock. A bayesian reanalysis of the ANDROMEDA-SHOCK trial. Am J Respir Crit Care Med. 2020;201:423–9.

    Article  CAS  PubMed  Google Scholar 

  13. Hernández G, Ospina-Tascón GA, Damiani LP, Estenssoro E, Dubin A, Hurtado J, et al. Effect of a resuscitation strategy targeting peripheral perfusion status vs serum lactate levels on 28-day mortality among patients with septic shock: the ANDROMEDA-SHOCK randomized clinical trial. JAMA. 2019;321:654–64.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Alhazzani W, Møller MH, Arabi YM, Loeb M, Gong MN, Fan E, et al. Surviving sepsis campaign: guidelines on the management of critically ill adults with coronavirus disease 2019 (COVID-19). Intensive Care Med. 2020;46:854–87.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Schriger DL, Baraff LJ. Capillary refill–is it a useful predictor of hypovolemic states? Ann Emerg Med. 1991;20:601–5.

    Article  CAS  PubMed  Google Scholar 

  16. Jouffroy R, Saade A, Tourtier JP, Gueye P, Bloch-Laine E, Ecollan P, et al. Skin mottling score and capillary refill time to assess mortality of septic shock since pre-hospital setting. Am J Emerg Med. 2019;37:664–71.

    Article  PubMed  Google Scholar 

  17. Jacquet-Lagrèze M, Wiart C, Schweizer R, Didier L, Ruste M, Coutrot M, et al. Capillary refill time for the management of acute circulatory failure: a survey among pediatric and adult intensivists. BMC Emerg Med. 2022;22:131.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Fleming S, Gill P, Jones C, Taylor JA, Van den Bruel A, Heneghan C, et al. The Diagnostic Value of Capillary Refill Time for Detecting Serious Illness in Children: A Systematic Review and Meta-Analysis. PLoS One [Internet]. 2015 [cited 2017 Sep 26];10. Available from: http://0-www-ncbi-nlm-nih-gov.brum.beds.ac.uk/pmc/articles/PMC4573516/

  19. Pickard A, Karlen W, Ansermino JM. Capillary refill time: is it still a useful clinical sign? Anesth Analg. 2011;113:120–3.

    Article  PubMed  Google Scholar 

  20. Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy [Internet]. [cited 2023 Feb 20]. Available from: https://training.cochrane.org/handbook-diagnostic-test-accuracy

  21. Devillé WL, Buntinx F, Bouter LM, Montori VM, de Vet HC, van der Windt DA, et al. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Methodol. 2002;2:9.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Whiting PF, Rutjes AWS, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155:529–36.

    Article  PubMed  Google Scholar 

  24. Knapp G, Hartung J. Improved tests for a random effects meta-regression with a single covariate. Stat Med. 2003;22:2693–710.

    Article  PubMed  Google Scholar 

  25. Cooper HM, Hedges LV, Valentine JC, editors. The handbook of research synthesis and meta-analysis. 2nd ed. New York: Russell Sage Foundation; 2009.

    Google Scholar 

  26. Schwarzer G. meta: General Package for Meta-Analysis [Internet]. 2022 [cited 2023 Jan 3]. Available from: https://CRAN.R-project.org/package=meta

  27. Riebler JG and A. meta4diag: Meta-Analysis for Diagnostic Test Studies [Internet]. 2021 [cited 2023 Jan 3]. Available from: https://CRAN.R-project.org/package=meta4diag

  28. Hernandez G, Luengo C, Bruhn A, Kattan E, Friedman G, Ospina-Tascon GA, et al. When to stop septic shock resuscitation: clues from a dynamic perfusion monitoring. Ann Intensive Care. 2014;4:30.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Coslovsky M, Takala J, Exadaktylos AK, Martinolli L, Merz TM. A clinical prediction model to identify patients at high risk of death in the emergency department. Intensive Care Med. 2015;41:1029–36.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Darioli V, Taffé P, Carron P-N, Dami F, Vallotton L, Yersin B, et al. Evaluation of the discriminative performance of the prehospital National Advisory Committee for Aeronautics score regarding 48-h mortality. Eur J Emerg Med. 2019;26:366–72.

    Article  PubMed  Google Scholar 

  31. Morocho JP, Martínez AF, Cevallos MM, Vasconez-Gonzalez J, Ortiz-Prado E, Barreto-Grimaldos A, et al. Prolonged capillary refilling as a predictor of mortality in patients with septic shock. J Intensive Care Med. 2022;37:423–9.

    Article  PubMed  Google Scholar 

  32. Daviaud F, Grimaldi D, Dechartres A, Charpentier J, Geri G, Marin N, et al. Timing and causes of death in septic shock. Ann Intensive Care. 2015;5:16.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Hernandez G, Bellomo R, Bakker J. The ten pitfalls of lactate clearance in sepsis. Intensive Care Med. 2019;45:82–5.

    Article  PubMed  Google Scholar 

  34. Le Dorze M, Legrand M, Payen D, Ince C. The role of the microcirculation in acute kidney injury. Curr Opin Crit Care. 2009;15:503–8.

    Article  PubMed  Google Scholar 

  35. Wiersema R, Koeze J, Eck RJ, Kaufmann T, Hiemstra B, Koster G, et al. Clinical examination findings as predictors of acute kidney injury in critically ill patients. Acta Anaesthesiol Scand. 2020;64:69–74.

    Article  CAS  PubMed  Google Scholar 

  36. Fedak KM, Bernal A, Capshaw ZA, Gross S. Applying the Bradford Hill criteria in the 21st century: how data integration has changed causal inference in molecular epidemiology. Emerg Themes Epidemiol. 2015;12:14.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Holcomb JB, Niles SE, Miller CC, Hinds D, Duke JH, Moore FA. Prehospital physiologic data and lifesaving interventions in trauma patients. Mil Med. 2005;170:7–13.

    Article  PubMed  Google Scholar 

  38. Rossello X, Bueno H, Gil V, Jacob J, Martín-Sánchez FJ, Llorens P, et al. Synergistic impact of systolic blood pressure and perfusion status on mortality in acute heart failure. Circ Heart Fail. 2021;14:e007347.

    Article  PubMed  Google Scholar 

  39. Kawaguchi R, Nakada T-A, Oshima T, Shinozaki M, Nakaguchi T, Haneishi H, et al. Optimal pressing strength and time for capillary refilling time. Crit Care. 2019;23:4.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Brown LH, Prasad NH, Whitley TW. Adverse lighting condition effects on the assessment of capillary refill. Am J Emerg Med. 1994;12:46–7.

    Article  CAS  PubMed  Google Scholar 

  41. Shinozaki K, Jacobson LS, Saeki K, Kobayashi N, Weisner S, Falotico JM, et al. Does training level affect the accuracy of visual assessment of capillary refill time? Crit Care. 2019;23:157.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Rosner B. Fundamentals of Biostatistics. Cengage Learning; 2015.

  43. Contreras R, Hernández G, Valenzuela ED, González C, Ulloa R, Soto D, et al. Exploring the relationship between capillary refill time, skin blood flow and microcirculatory reactivity during early resuscitation of patients with septic shock: a pilot study. J Clin Monit Comput [Internet]. 2022 [cited 2023 Jan 3]; Available from: https://0-doi-org.brum.beds.ac.uk/10.1007/s10877-022-00946-7

  44. Lechleuthner A, Lefering R, Bouillon B, Lentke E, Vorweg M, Tiling T. Prehospital detection of uncontrolled haemorrhage in blunt trauma. Eur J Emerg Med. 1994;1:13–8.

    Article  CAS  PubMed  Google Scholar 

  45. Pealing L, Perel P, Prieto-Merino D, Roberts I. CRASH-2 Trial Collaborators. Risk factors for vascular occlusive events and death due to bleeding in trauma patients; an analysis of the CRASH-2 cohort. PLoS ONE. 2012;7:e50603.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Mrgan M, Rytter D, Brabrand M. Capillary refill time is a predictor of short-term mortality for adult patients admitted to a medical department: an observational cohort study. Emerg Med J. 2014;31:954–8.

    Article  PubMed  Google Scholar 

  47. Bourcier S, Pichereau C, Boelle P-Y, Nemlaghi S, Dubée V, Lejour G, et al. Toe-to-room temperature gradient correlates with tissue perfusion and predicts outcome in selected critically ill patients with severe infections. Ann Intensive Care. 2016;6:63.

    Article  PubMed  PubMed Central  Google Scholar 

  48. Alegría L, Vera M, Dreyse J, Castro R, Carpio D, Henriquez C, et al. A hypoperfusion context may aid to interpret hyperlactatemia in sepsis-3 septic shock patients: a proof-of-concept study. Ann Intensive Care. 2017;7:29.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Lara B, Enberg L, Ortega M, Leon P, Kripper C, Aguilera P, et al. Capillary refill time during fluid resuscitation in patients with sepsis-related hyperlactatemia at the emergency department is related to mortality. PLoS ONE. 2017;12:e0188548.

    Article  PubMed  PubMed Central  Google Scholar 

  50. Serano AMN, Alonso JV, Piñero GR, Camacho AR, Benet JS, Vaquero M. Biomarkers in shock patients and their value as a prognostic tool; a prospective multi-center cohort study. Bull Emerg Trauma. 2019;7:232–9.

    Article  PubMed  PubMed Central  Google Scholar 

  51. Mongkolpun W, Bakos P, Vincent J-L, Creteur J. Monitoring skin blood flow to rapidly identify alterations in tissue perfusion during fluid removal using continuous veno-venous hemofiltration in patients with circulatory shock. Ann Intensive Care. 2021;11:59.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Bigé N, Lavillegrand J-R, Dang J, Attias P, Deryckere S, Joffre J, et al. Bedside prediction of intradialytic hemodynamic instability in critically ill patients: the SOCRATE study. Ann Intensive Care. 2020;10:47.

    Article  PubMed  PubMed Central  Google Scholar 

  53. Sebat C, Vandegrift MA, Oldroyd S, Kramer A, Sebat F. Capillary refill time as part of an early warning score for rapid response team activation is an independent predictor of outcomes. Resuscitation. 2020;153:105–10.

    Article  PubMed  Google Scholar 

  54. Amson H, Vacheron C-H, Thiolliere F, Piriou V, Magnin M, Allaouchiche B. Core-to-skin temperature gradient measured by thermography predicts day-8 mortality in septic shock: a prospective observational study. J Crit Care. 2020;60:294–9.

    Article  PubMed  Google Scholar 

  55. Magnin M, Amson H, Vacheron C-H, Thiollière F, Piriou V, Junot S, et al. Associations between peripheral perfusion disorders, mean arterial pressure and dose of norepinephrine administrated in the early phase of septic shock. Clin Exp Pharmacol Physiol. 2021;48:1327–35.

    Article  CAS  PubMed  Google Scholar 

  56. Lavillegrand J-R, Raia L, Urbina T, Hariri G, Gabarre P, Bonny V, et al. Vitamin C improves microvascular reactivity and peripheral tissue perfusion in septic shock patients. Crit Care. 2022;26:25.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

No funding source.

Author information

Authors and Affiliations

  1. Service d’anesthésie-Réanimation, Hôpital Cardiologique Louis Pradel, 59 Bd Pinel, 69500, Hospices Civils de LyonBron, France

    Matthias Jacquet-Lagrèze, Aymeric Pernollet, Delphine Chesnel, Martin Ruste, Rémi Schweizer & Jean-Luc Fellahi

  2. Faculté de Médecine Lyon Est, Université Claude Bernard Lyon 1, 8, Avenue Rockefeller, 69373, Lyon Cedex 08, France

    Matthias Jacquet-Lagrèze, Aymeric Pernollet, Delphine Chesnel, Martin Ruste, Rémi Schweizer, Bernard Allaouchiche & Jean-Luc Fellahi

  3. CarMeN Laboratoire, Inserm UMR 1060, Université Claude Bernard, Lyon 1, Lyon, France

    Matthias Jacquet-Lagrèze, Martin Ruste & Jean-Luc Fellahi

  4. Departamento de Medicina Intensiva, Facultad de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile

    Eduardo Kattan & Glenn Hernandez

  5. The Latin American Intensive Care Network (LIVEN), Santiago, Chile

    Eduardo Kattan & Glenn Hernandez

  6. Hôpital Saint-Antoine, Service de Médecine Intensive-Réanimation, Sorbonne Université, Paris, France

    Hafid Ait-Oufella

  7. Service d’anesthésie-Réanimation, Hôpital Lyon Sud, Hospices Civils de Lyon, 165 Chem. du Grand Revoyet, 69495, Pierre-Bénite, France

    Bernard Allaouchiche

Authors
  1. Matthias Jacquet-Lagrèze
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Aymeric Pernollet
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Eduardo Kattan
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Hafid Ait-Oufella
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Delphine Chesnel
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Martin Ruste
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Rémi Schweizer
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Bernard Allaouchiche
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Glenn Hernandez
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Jean-Luc Fellahi
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

MJL and AP contributed to study concept and design and interpretation of data. MJL, AP, EK, HAO, DC, MR, BA, GH, and RS were involved in acquisition and interpretation of data. MJL contributed to drafting of manuscript, statistical analysis, and study supervision. MJL, JLF, AP, EK, HAO, DC, MR, BA, RS, GH, and JLF were involved in critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript and agree to be accountable for all aspects of the work. All authors ensured that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Corresponding author

Correspondence to Matthias Jacquet-Lagrèze.

Ethics declarations

Ethical approval and consent to participate

PROSPERO registration number: 2022 CRD42022297158 Available from: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022297158

Consent for publication

Not applicable.

Competing interests

MJL is cofounder and shareholder of the DiCARTECH company that has been created to build and sell a device that measure capillary refill time.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1

. Supplementary file.

Additional file 2

. Supplementary Figures.

Additional file 3

. Table S1. PRISMA 2020 Checklist.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Jacquet-Lagrèze, M., Pernollet, A., Kattan, E. et al. Prognostic value of capillary refill time in adult patients: a systematic review with meta-analysis. Crit Care 27, 473 (2023). https://0-doi-org.brum.beds.ac.uk/10.1186/s13054-023-04751-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://0-doi-org.brum.beds.ac.uk/10.1186/s13054-023-04751-9

Keywords

Critical Care

ISSN: 1364-8535