University of Groningen
The role of troponin and albumin to assess myocardial dysfunction after cardiac surgery and
in the critically ill
van Beek, Dianne E.C.
DOI:
10.33612/diss.101333600
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
van Beek, D. E. C. (2019). The role of troponin and albumin to assess myocardial dysfunction after cardiac
surgery and in the critically ill. Rijksuniversiteit Groningen. https://doi.org/10.33612/diss.101333600
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter
Validation of
serum albumin as
a prognostic marker in
the intensive care unit:
a SICS-I sub-study
Dianne van Beek, Renske Wiersema, T. Kaufmann, Bart Hiemstra, R. Eck, F. Keus, Thomas Scheeren, Iwan van der Horst, SICS study group.
Abstract
Background: The aim of this study was to externally validate serum albumin (SA) at intensive care unit (ICU) admission as a prognostic marker for insufficient tissue perfusion, myocardial damage and mortality.
Methods: This was a post-hoc analysis of the SICS-I study, a prospective observational ICU cohort. SA levels obtained within six hours of admission were evaluated. Endpoints were stable or decreased arterial lactate levels during the first 24 hours, myocardial damage (analyzed as cardiac troponin (hs-cTnT) after 24 hours) and one-day and 90-day all-cause mortality. Correction for effect modifiers took place where appropriate.
Results: SA at admission was available in 1028 (96%) of 1075 patients of the SICS-I cohort. In multivariate analysis, higher SA was associated with stable or decreasing arterial lactate levels after 24 hours (odds ratio [OR] 1.06, 95% confidence interval [CI] 1.03-1.08, p<0.005), a lower log transformed hs-cTnT (regression coefficient -0.02, 95% CI -0.04-0.00 , p=0.039, R2 0.16) and a decrease in 90-day mortality (OR 0.97, 95% CI 0.94-0.99, p=0.010).
Conclusions: Higher SA at ICU admission is associated with a stable or decreased arterial lactate levels after 24 hours, a lower hs-cTnT after 24 hours, and a decreased 90-day mortality.
119
Validation of albumin as a prognostic marker
8
Introduction
The predictive value of serum albumin (SA) for mortality has not only been shown in a general patient population1, but also in patients admitted to an intensive care unit
(ICU)2,3. There is currently no conclusive evidence that albumin supplementation in the
ICU improves mortality.4 However, albumin supplementation has shown to reduce the
positive fluid balance often seen in ICU patients5,6;7 , and to improve the Sequential Organ
Failure Assessment (SOFA) score6. Both a positive fluid balance and the SOFA score are
clearly associated with mortality.8,9
We have previously conducted a prospective cohort study in which we found that a higher SA at ICU admission was associated with a decrease in arterial lactate levels after 24 hours.10 In a retrospective cohort study we found that lower SA levels were associated with
a higher cumulative amount of myocardial damage after cardiac surgery.11
The aim was to externally validate SA at ICU admission as a prognostic marker for arterial lactate levels after 24 hours, myocardial damage, and one-day and 90-day all-cause mortality.
Methods
Study population
The study population consisted of patients included in the Simple Intensive Care Studies - I (SICS-I), a single-center, prospective observational cohort study designed to evaluate the diagnostic and prognostic value of combinations of clinical and hemodynamic variables in critically ill patients (NCT02912624)12. Inclusion criteria were acute ICU admission and an
expected stay of >24 hours. Patients were excluded if their ICU admission was elective, if acquiring research data interfered with clinical care, or if informed consent was absent. The local institutional review board approved the original study (M15.168207).
Variables
Variables either obtained for the purpose of the main study or as part of standard care were used for the analyses. Data was extracted from electronic patient records. All clinical examination variables were obtained by trained researchers, not involved in patient care. Cardiac Index was determined using transthoracic echocardiography. A detailed protocol on data collection has been published elsewhere13.
120
Chapter 8
All first measurements of SA, plasma creatinine, C-reactive protein (CRP), cardiac troponin (hs-cTnT), and arterial lactate after ICU admission were collected (limited to a maximal time interval of six hours after admission). Reference ranges were 35-50 g/L for SA, 50-110 µmol/L (males) and 50-90 µmol/L (females) for creatinine, < 5.0 mg/L for CRP, and <2.2 mmol/L for lactate. The hs-cTnT was obtained by a high-sensitive troponin T test with a reference range < 50 ng/L. Additionally, the first measured arterial lactate and hs-cTnT on the second day of ICU admission were registered (a time interval of 18 to 30 hours after ICU admission was deemed acceptable).
Outcomes
The delta lactate arterial lactate level (Δ lactate) was determined by comparing the arterial lactate at the second day of ICU stay to the level at ICU admission. All patients were classified into either one of two categories depending on whether the arterial lactate level had decreased or stayed stable, or whether it had increased. The troponin (hs-cTnT) used to assess myocardial damage was the first measured hs-cTnT on day two. The 90-day mortality dates were obtained from the municipality registry, and first 90-day and 90-90-day mortality were calculated.
Statistical analysis
Baseline characteristics were presented as proportions, a mean with standard deviation (SD) or a median with interquartile ranges (IQR) where appropriate. All analyses were conducted on a complete case basis analysis as all variables in our multivariable models had fewer than 5% missing values.14 Values were presented with 95% confidence interval.
For all analyses significance was set at p< 0.05. Log transformation of hs-cTnT after 24 hours was performed to ensure normal distribution of the data. If univariable analysis was significant, subsequent multivariable analyses were performed with correction for potential effect modifiers. Potential effect modifiers were selected based on clinical knowledge and are presented per analyses in table 1.
We used logistic regression to validate the results with a binominal outcome, and linear regression as appropriate. For linear regression models, normality of the residuals was assessed with kernel density plots and multicollinearity was checked with variance inflation factors. For logistic regression models, calibration was checked with calibration plots and Hosmer-Lemeshow tests and discrimination was evaluated with area under the receiver operating characteristic (AUC ROC)-curves.
121
Validation of albumin as a prognostic marker
8
Table 1. Potential effect modifiers per analyses.
Potential effect modifiers: Δ arterial lactate level 1. Gender
2. Age
3. Type of patient: acute surgery / planned surgery/ medical 4. APACHE II score
5. CRP at ICU admission
Potential effect modifiers: troponin after 24 hours 1. Gender
2. Age
3. Creatinine level at ICU admission 4. DM: yes / no
5. Low cardiac index (below 2.2 L/min/m2)
6. Type of patient: acute surgery / planned surgery/ medical 7. APACHE II score
Potential effect modifiers: mortality 1. Gender
2. Age 3. BMI
4. Type of patient: acute surgery / planned surgery/ medical 5. APACHE II score
6. Troponin at ICU admission 7. Lactate at ICU admission
Results
In total, 1075 patients were included within the SICS-I cohort. SA at ICU admission was available in 1023 patients (95%). The mean SA at ICU admission was 29.8 g/L (SD ±7.6 g/L) and most patients were admitted for a medical reason (n=713, 66%). Eight patients (1%) were lost to follow-up due to emigration or residence in another country. The baseline characteristics of the entire cohort and per category of analyses are shown in table 2. Arterial lactate level
The Δ lactate was available in 1003 patients (93%). The arterial lactate level decreased or remained stable during the first day in 732 patients (73%). Higher SA at admission was associated with a stable or decreasing lactate level (odds ratio [OR] 1.05, 95%CI 1.03-1.07, p<0.005). This result remained robust after correction for effect modifiers (OR 1.06, 95%CI 1.03-1.08, p<0.005). The AUC ROC was 0.63 (95%CI 0.59-0.68), Hosmer Lemeshow Goodness of Fit 13.16, p=0.11.
122
Chapter 8
Table 2. Baseline characteristics. Values shown as number (percentage) unless otherwise indicated SICS-I (n=1075) Lactate available (n=1003) Troponin available (n=776) SA and mortality available (n=1023) Age in years 62 (15) 62 (14) 62 (14) 62 (15) Male 674 (63%) 663 (63%) 489 (63%) 642 (63%) BMI (kg/m2) 26.9 (5.5) 26.8 (5.4) 26.8 (5.2) 26.9 (5.5) Admittance Medical 713 (66%) 666 (656%) 502 (65%) 673 (66%) Acute surgery 316 (29%) 297 (30.0%) 245 (32%) 305 (30%) Planned surgery 46 (4%) 40 (4%) 29 (4%) 45 (4%) APACHE II score 21.4 (7.2) 21.6 (7.2) 22.1 (7.2) 21.6 (7.2) SA (g/L) 29.8 (7.6) 30.0 (7.6) 30.5 (7.5) 29.8 (7.6) Lactate (mmol/L) 1.9 [1.1, 3.6] 1.9 [1.1, 3.6] 1.9 [1.1, 3.7] 1.9 [1.1, 3.6] CRP (mg/L) 25 [4, 121] 25 [4, 122] 20 [3, 106] 25 [4, 122] Creatinine 120 (111) 120 (112) 119 (113) 121 (112) DM 217 (20%) 203 (20%) 160 (21%) 209 (20%)
Mean (standard deviation), or median [interquartile ranges], APACHE: Acute Physiology and Chronic Health Evaluation, SA: serum albumin, CRP: c-reactive protein, BMI: body mass index, DM: diabetes mellitus
Myocardial damage
hs-cTnT after 24 hours of ICU admission was available in 776 patients (72%). hs-cTnT increased during the first day of ICU admission in 392 (61%) patients. Higher SA was associated with a lower log transformed hs-cTnT after 24 hours of ICU admission (regression coefficient -0.03, 95% CI-0.04/-0.01, p<0.005, R2 0.02). This linear association remained after correction for
effect modifiers (regression coefficient -0.02, 95% CI-0.04/-0.00, p=0.039, R2 0.16)
Mortality
In total, 49 (4.8%) patients died during the first day of ICU admission and 298 (28%) patients died within 90-days. SA was associated with one-day mortality in univariate but not in multivariate analyses (table 3). In contrast, the association remained significant in multivariate analysis for 90-day all-cause mortality (AUC ROC 0.78 95% CI 0.74-0.82, Hosmer Lemeshow Goodness of Fit 8.72, p=0.37).
Table 3. Logistic regression for SA at ICU admission and mortality.
Unadjusted Adjusted**
OR (95% CI) p OR (95% CI) p
One-day mortality 0.96 (0.93-0.99) 0.048* 0.97 (0.92-1.02) 0.252
90-day mortality 0.97 (0.95-0.99) 0.001* 0.97 (0.94-0.99) 0.010* *indicates a significant result. ** adjusted for: gender, age, BMI, type of admission, APACHE II score, lactate at ICU admission, troponin at ICU admission.
123
Validation of albumin as a prognostic marker
8
Discussion
In this cohort of critically ill patients, SA at ICU admission was associated with a stable or decreasing arterial lactate level after 24 hours, a lower Tn after 24 hours, and a lower 90-day mortality. These results serve as an external validation of our previously published results.10,11
Insufficient tissue perfusion ultimately results in formation of lactate from anaerobic metabolism.15 The association between arterial lactate level after one day of ICU
admission and mortality has been demonstrated.16 In addition, the use of arterial lactate
measurements to guide resuscitation has been shown to be associated with an improved patient outcome.17 The importance of a quick reduction of arterial lactate levels in ICU
patients has also been demonstrated, as patient outcome may already improve with a reduction of arterial lactate during the first 24 hours of ICU stay.18
The highly sensitive and specific biomarker hs-cTnT is generally used to detect myocardial damage.19 In a previous study we demonstrated that SA levels were associated with
the cumulative amount of myocardial damage after cardiac surgery.11 In this study we
demonstrated that SA has a negative linear association with hs-cTnT after 24 hours of ICU stay. We hypothesize that a lower hs-cTnT level after 24 hours is clinically relevant, as it suggests that on-going myocardial damage is no longer worsening and may even indicate an improvement.
We found that higher SA at ICU admission was associated with lower odds for 90-day mortality. It has previously been demonstrated that SA <25g/L was independently associated with 28-day mortality in ICU patients.20 However, a clear effect of albumin
supplementation on mortality has not yet been demonstrated in ICU patients.4 This may
be explained by the fact that SA is merely a marker for severity of illness and not a causal factor. Another explanation could be that previous studies were underpowered to detect a small effect on mortality.21
Strength of this study is that it is an external validation of previously published research.10,11
In prognostic research it is important that the results of your study are validated in a distinct cohort of patients.22,23 We validated our results
This study has several limitations. First of all, this study is a post-hoc analysis and was therefore limited to previously collected data. For instance, we were unable to determine the area under the curve of cTnT, since not for all patients serial measurement of hs-cTnT was not available. Second, selection of time related variables was based on calendar
124
Chapter 8
date and not on the exact number of hours since ICU admission. This means that the exact time between for instance laboratory values is not completely identical between all included patients. Third, we conducted a complete case analysis and not all patients could be included in all analyses due to missing data. Although this poses a risk of bias, we believe that our sub-cohorts are still sufficiently large to be generalizable. Finally, although we extensively corrected for potential effect modifiers, we cannot exclude the possibility for residual confounding in unmeasured and unknown factors.
Conclusions
In conclusion, a higher SA upon ICU admission was associated with a decrease in arterial lactate levels, a reduction in hs-cTnT and a lower mortality. This early biomarker could become a valuable indicator to inform clinicians on the short-term (24-hour) prognosis of patients admitted to the ICU.
125
Validation of albumin as a prognostic marker
8
References
1. Goldwasser P, Feldman J. Association of serum albumin and mortality risk. J Clin Epidemiol. 1997;50(6):693-703.
2. Pan S-W, Kao H-K, Yu W-K, et al. Synergistic impact of low serum albumin on intensive care unit admission and high blood urea nitrogen during intensive care unit stay on post-intensive care unit mortality in critically ill elderly patients requiring mechanical ventilation. Geriatr Gerontol Int. 2013;13(1):107-115.
3. McCluskey A, Thomas AN, Bowles BJ, Kishen R. The prognostic value of serial measurements of serum albumin concentration in patients admitted to an intensive care unit. Anaesthesia. 1996;51(8):724-727.
4. Roberts I, Blackhall K, Alderson P, Bunn F, Schierhout G. Human albumin solution for resuscitation and volume expansion in critically ill patients. In: Roberts I, ed. Cochrane Database of Systematic Reviews. Chichester, UK: John Wiley & Sons, Ltd; 2011.
5. Caironi P, Tognoni G, Masson S, et al. Albumin replacement in patients with severe sepsis or septic shock. N Engl J Med. 2014;370(15):1412-1421.
6. Dubois M-J, Orellana-Jimenez C, Melot C, et al. Albumin administration improves organ function in critically ill hypoalbuminemic patients: A prospective, randomized, controlled, pilot study. Crit Care Med. 2006;34(10):2536-2540. 7. Vincent JL, De Backer D, Wiedermann CJ. Fluid
management in sepsis: The potential beneficial effects of albumin. J Crit Care. 2016;35:161-167. 8. Raith EP, Udy AA, Bailey M, et al. Prognostic
Accuracy of the SOFA Score, SIRS Criteria, and qSOFA Score for In-Hospital Mortality Among Adults With Suspected Infection Admitted to the Intensive Care Unit. JAMA. 2017;317(3):290. 9. Claure-Del Granado R, Mehta RL. Fluid overload
in the ICU: evaluation and management. BMC Nephrol. 2016;17(1):109.
10. van Beek DEC, Königs MHH, Kuijpers YAM, van der Horst ICC, Scheeren TWL. Predictive value of serum albumin levels on noradrenaline and fluid requirements in the first 24 h after admission to the Intensive Care Unit — A prospective observational study. J Crit Care. 2018;47:99-103. 11. van Beek DEC, van der Horst ICC, de Geus AF, Mariani MA, Scheeren TWL. Albumin, a marker for post-operative myocardial damage in cardiac surgery. J Crit Care. 2018;47:55-60.
12. Hiemstra B, Eck RJ, Koster G, et al. Clinical examination, critical care ultrasonography and outcomes in the critically ill: cohort profile of the Simple Intensive Care Studies-I. BMJ Open. 2017;7(9):e017170.
13. Hiemstra B, Koster G, Wiersema R, et al. The diagnostic accuracy of clinical examination for estimating cardiac index in critically ill patients: the Simple Intensive Care Studies-I. Intensive Care Med. January 2019:1-11.
14. Jakobsen JC, Gluud C, Wetterslev J, Winkel P. When and how should multiple imputation be used for handling missing data in randomised clinical trials - a practical guide with flowcharts. BMC Med Res Methodol. 2017;17(1):162.
15. Levy B, Desebbe O, Montemont C, Gibot S. Increased aerobic glycolysis through B2 simulation is a ommon mechanism involved in lactate formation during shock states. Shock. 2008;30(4):417-421.
16. Mahmoodpoor A, Shadvar K, Saghaleini SH, et al. Which one is a better predictor of ICU mortality in septic patients? Comparison between serial serum lactate concentrations and its removal rate. J Crit Care. 2018;44:51-56.
17. Jansen TC, van Bommel J, Schoonderbeek FJ, et al. Early Lactate-Guided Therapy in Intensive Care Unit Patients. Am J Respir Crit Care Med. 2010;182(6):752-761.
18. Jansen TC, van Bommel J, Mulder PG, et al. Prognostic Value of Blood Lactate Levels: Does the Clinical Diagnosis at Admission Matter? J Trauma Inj Infect Crit Care. 2009;66(2):377-385.
126
Chapter 8
19. Thygesen K, Alpert JS, Jaffe AS, et al. Third universal definition of myocardial infarction. Eur Heart J. 2012;33(20):2551-2567.
20. SAFE Study Investigators, Finfer S, Bellomo R, et al. Effect of baseline serum albumin concentration on outcome of resuscitation with albumin or saline in patients in intensive care units: analysis of data from the saline versus albumin fluid evaluation (SAFE) study. BMJ. 2006;333(7577):1044-1044.
21. Caironi P, Langer T, Gattinoni L. Albumin in critically ill patients. Curr Opin Crit Care. 2015;21(4):302-308.
22. Altman DG, Vergouwe Y, Royston P, Moons KGM. Prognosis and prognostic research: validating a prognostic model. BMJ. 2009;338:b605. 23. Moons KGM, Royston P, Vergouwe Y, Grobbee DE,
Altman DG. Prognosis and prognostic research: what, why, and how? BMJ. 2009;338:b375.
